Somatosensory Research Methods 1071630679, 9781071630679

Citation preview

Neuromethods 196

Nicholas Paul Holmes Editor

Somatosensory Research Methods

NEUROMETHODS

Series Editor Wolfgang Walz University of Saskatchewan Saskatoon, SK, Canada

For further volumes: http://www.springer.com/series/7657

Neuromethods publishes cutting-edge methods and protocols in all areas of neuroscience as well as translational neurological and mental research. Each volume in the series offers tested laboratory protocols, step-by-step methods for reproducible lab experiments and addresses methodological controversies and pitfalls in order to aid neuroscientists in experimentation. Neuromethods focuses on traditional and emerging topics with wide-ranging implications to brain function, such as electrophysiology, neuroimaging, behavioral analysis, genomics, neurodegeneration, translational research and clinical trials. Neuromethods provides investigators and trainees with highly useful compendiums of key strategies and approaches for successful research in animal and human brain function including translational “bench to bedside” approaches to mental and neurological diseases.

Somatosensory Research Methods Edited by

Nicholas Paul Holmes School of Sport, Exercise and Rehabilitation Sciences, University of Birmingham, Birmingham, UK

Editor Nicholas Paul Holmes School of Sport, Exercise and Rehabilitation Sciences University of Birmingham Birmingham, UK

ISSN 0893-2336 ISSN 1940-6045 (electronic) Neuromethods ISBN 978-1-0716-3067-9 ISBN 978-1-0716-3068-6 (eBook) https://doi.org/10.1007/978-1-0716-3068-6 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface to the Series Experimental life sciences have two basic foundations: concepts and tools. The Neuromethods series focuses on the tools and techniques unique to the investigation of the nervous system and excitable cells. It will not, however, shortchange the concept side of things as care has been taken to integrate these tools within the context of the concepts and questions under investigation. In this way, the series is unique in that it not only collects protocols but also includes theoretical background information and critiques which led to the methods and their development. Thus, it gives the reader a better understanding of the origin of the techniques and their potential future development. The Neuromethods publishing program strikes a balance between recent and exciting developments like those concerning new animal models of disease, imaging, in vivo methods, and more established techniques, including, for example, immunocytochemistry and electrophysiological technologies. New trainees in neurosciences still need a sound footing in these older methods in order to apply a critical approach to their results. Under the guidance of its founders, Alan Boulton and Glen Baker, the Neuromethods series has been a success since its first volume was published through Humana Press in 1985. The series continues to flourish through many changes over the years. It is now published under the umbrella of Springer Protocols. While methods involving brain research have changed a lot since the series started, the publishing environment and technology have changed even more radically. Neuromethods has the distinct layout and style of the Springer Protocols program, designed specifically for readability and ease of reference in a laboratory setting. The careful application of methods is potentially the most important step in the process of scientific inquiry. In the past, new methodologies led the way in developing new disciplines in the biological and medical sciences. For example, Physiology emerged out of Anatomy in the nineteenth century by harnessing new methods based on the newly discovered phenomenon of electricity. Nowadays, the relationships between disciplines and methods are more complex. Methods are now widely shared between disciplines and research areas. New developments in electronic publishing make it possible for scientists that encounter new methods to quickly find sources of information electronically. The design of individual volumes and chapters in this series takes this new access technology into account. Springer Protocols makes it possible to download single protocols separately. In addition, Springer makes its print-on-demand technology available globally. A print copy can therefore be acquired quickly and for a competitive price anywhere in the world. Saskatoon, SK, Canada

Wolfgang Walz

v

Preface On the hottest ever day in the United Kingdom—at least by the time, in 2019—sixty-eight somatosensory researchers packed themselves into Lecture Theatre A1 in the School of Psychology at the University of Nottingham. A two-day workshop funded by the Experimental Psychology Society—“Research IN Touch”—began. As this volume’s Editor began giving his Introduction, it became clear that the air conditioning was not working. The researchers from all over the UK, and several from continental Europe and further afield, shifted in their seats. Their skin temperature was rising, they were sweating, itching, feeling damp. Some began to experience discomfort. The first act of this interdisciplinary somatosensory meeting was a homeostatic one, to seek an air-conditioned room! The somatosensory systems—those mediating our senses of touch, force, movement, temperature, itch, wetness, pain, pleasure, hunger, thirst, and all other bodily feelings—are critically important to our survival. These systems are both difficult to study and remarkably under-studied in the history of psychology and the neurosciences. Yet the somatosensory systems were among the very first to be studied in the modern era of neuroscience which began, arguably, in the early 1800s (Chapters 1 and 5, this volume). The first workshop of the fledgling “Research IN Touch” network was followed by an invitation to edit a volume for the Springer NeuroMethods series. A short anecdote from the workshop will suffice to explain why I was excited by and immediately accepted this invitation. At one of the coffee breaks on the first day, I was chatting with a young researcher, who said: “Ah yes, I know your work – you’re a methods person.” This comment struck me, and it still does. I don’t see myself as “a methods person,” although I am, like many others, obsessed with improving my methods so that I can try to answer important and interesting questions. This volume is a two-hundred-thousand-word love letter to the methods that somatosensory psychologists, engineers, and neuroscientists use to study the most interesting systems of the brain. It has been a great pleasure to work with its 49 authors from 12 countries. Many of these researchers attended the workshop in Nottingham, but many others were invited. It has taken me far too long—3 years—to put this volume together, and I take all the credit for this delay. In the rest of this Introduction, I summarize the fantastic contributions that make this volume a valuable resource for any researcher who is interested in the somatosensory system. The volume begins with five chapters on the more mechanical or discriminative aspects of somatosensation: tactile detection, discrimination, and orientation perception (Holmes and Tame`, Chapter 1); force perception and generation (Kilteni, Chapter 2); proprioception and kinesthesis (Kavounoudias, Blanchard, Landelle, and Chancel, Chapter 3); tactile motion perception (Seizova-Cajic´, Fuchs, and Brooks, Chapter 4); and tactile distance perception (Longo, Chapter 5). Together, these chapters cover the fundamental methods of studying discriminative touch and proprioception. In some cases, these methods have been used for nearly two centuries—throughout the entire modern history of neuroscience—and they remain useful today. But discriminative touch and proprioception, although perhaps the most studied, are just two of the critical functions of somatosensation.

vii

viii

Preface

Affective touch, pain, wetness, itch, and interoception are the vital somatosensory functions described in the next five chapters. Only recently discovered, the somatosensory nervous system contains receptors and nerves that seem to be specialized for stroking, grooming, and affiliative touch. The methods required to study these systems are described by Haggarty, Makdani, and McGlone in Chapter 6. Fundamental to survival and clinically extremely important, methods for the study of pain are covered in Chapter 7 by Valentini, Vaughan, and Clauwaert. I then challenge any reader to get through Chapter 8 without becoming intensely aware of their crawling, itching skin; the power of pruritus and how to study it are illustrated in detail there by Holle and Lloyd. Another remarkably unstudied domain of somatosensation is wetness. Merrick, Ackerley, and Filingeri explain in Chapter 9 how there is no receptor for wetness perception, and dedicated methods are required to study it. In Chapter 10, Crucianelli and Morrison offer an insightful “interoceptive take on touch”—bringing together touch, temperature, pain, and pleasure to study the interoceptive system as a whole. Reading these first ten chapters, I was struck by the massive individual variability that researchers have to deal with when studying the somatosensory systems. In other sensory systems, we are very well aware of individual differences in sensitivity, as well as what can go wrong. People wear glasses or corrective lenses for visual differences; about 10% of men and fewer women have color vision anomalies. Older adults often wear hearing aids, while only young children can hear the highest-pitch sounds. We know, colloquially, of “super-tasters,” wine sommeliers, and food experts. The COVID-19 respiratory disease pandemic has shown millions of people worldwide what it feels like to lose their senses of smell, taste, and flavor. But what do we really know, colloquially or scientifically, about individual differences in touch, proprioception, motion, pain, pleasure, affection, temperature, itch, wetness, and interoception? This volume argues that there is still a lot to be discovered about individual differences in somatosensation. It is to these individual differences that Chapters 11, 12, 13, and 14 turn their attention. Puts and Cascio, in Chapter 11, ask how we can study touch in atypical populations of children and young people, particularly those with attentional, social, or communicative difficulties. In Chapter 12, the other side of development—aging—is tackled by Loomes, Roberts, and Allen, who examine how touch may decline as we grow old. An important aspect of individual differences is highlighted by studying somatosensory illusions, covered by Seizova-Cajic´, Zopf, Riemer, and Fuchs in Chapter 13. Our nervous systems mediate our perception of the world; illusions show us situations in which this mediation may fail. This chapter describes two methods for studying how people perceive the ownership and position of their own hands. To complete our account of individual differences in perception, Brooks, Kristja´nsson, and Unnthorsson describe their method of using somatosensory systems to replace or substitute for sensory losses in other modalities such as vision or hearing (Chapter 14). They note that the body provides an ideal surface upon which to recreate other sensory capacities, and it is thanks to the brain’s remarkable plasticity and adaptability that these substitutions are possible. The final six chapters of the volume are devoted to the brain. This ordering was deliberate. Open any random textbook on “perception” or “sensation,” and, when you have flipped through the first 60–80% of the book that deals with vision and hearing (the sole exception is Frank Geldard’s excellent The Human Senses, published in 1953 and again in 1972), you will often find that the somatosensory system is described, initially and

Preface

ix

primarily, in terms of its receptors in the body, the nervous pathways that lead to the brain, and the gross organization of the primary somatosensory cortex. This is all the editor remembers from his own psychology and neurobiology teaching. But, as the first 14 chapters of this volume testify, somatosensation is much more complex than just which types of receptors dwell in the body, or which of several crossed and uncrossed nervous pathways lead from the periphery to the central nervous system, or how the map of the body in the brain is distorted. Now that we understand the richness and complexity of the many submodalities that comprise somatosensation, we can return to ask: how can the brain possibly do this? The remarkable and often heroic work of electrophysiologists is described in Chapters 15 and 16. Ackerley and Watkins describe the painstaking methods of microneurography that are required to stimulate and record from individual nerve fibers in the human body (Chapter 15). By doing so, they can study the properties of individual receptors and afferent ¨ ztu¨rk, Deveciog˘lu, Vardar, fibers subserving multiple submodalities. In Chapter 16, O Duvan, and Gu¨c¸lu¨ provide a comprehensive account of the methods they have used to study the peripheral and central mechanisms of tactile processing in rodents, covering the generation of mechanical stimuli, the care and treatment of their experimental animals, and the recording of multiple responses at multiple levels of the nervous system. We stay with mammalian somatosensation in Chapter 17, where Panniello, Limal, and Kohl provide a high-resolution guide to recording calcium signals in the primary somatosensory “barrel” cortex of the mouse—arguably one of the most important model systems in neuroscience. Brain imaging in humans—this time using functional magnetic resonance imaging—is the topic of Chapter 18, in which Puckett and Sanchez Panchuelo describe high-resolution and high-field mapping of tactile representations in the human primary somatosensory cortex. One of the oldest neuroimaging techniques to study the somatosensory brain is described in Chapter 19. Electroencephalography (EEG) can be used to study somatosensory processing at very early stages, at high temporal resolution, and without invasive procedures. Vibell, Gillmeister, Sel, Haggarty, Van Velzen, and Forster team up to lay out their experiences of using EEG to study perception, attention, and multisensory integration in humans. The volume concludes in Chapter 20 with a survey of and guidance on using non-invasive methods of stimulating the human somatosensory brain electromagnetically. Brain stimulation methods such as transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (TES) have been available for decades; however, authors Tame` and Holmes describe how they can and should be taken much greater advantage of in the study of human somatosensation. It has been a privilege to read and edit the work of so many scientists from so many disciplines. I hope that through this volume, through our research network, and through future meetings, they will stay in touch and continue to discover how the somatosensory systems make us human.

Reference Geldard FA (1953) The Human Senses. Wiley, New York

Contents Preface to the Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

v vii xiii

DISCRIMINATIVE TOUCH, PROPRIOCEPTION, AND KINAESTHESIS

1 Detection, Discrimination & Localization: The Psychophysics of Touch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicholas Paul Holmes and Luigi Tame` 2 Methods of Somatosensory Attenuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Konstantina Kilteni 3 Muscle Tendon Vibration: A Method for Estimating Kinesthetic Perception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anne Kavounoudias, Caroline Blanchard, Caroline Landelle, and Marie Chancel 4 Creating Tactile Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tatjana Seizova-Cajic´, Xaver Fuchs, and Jack Brooks 5 Measuring Tactile Distance Perception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matthew R. Longo

3 35

55

71 95

PART II AFFECTIVE TOUCH, PAIN, WETNESS, ITCH, AND INTEROCEPTION 6 Affective Touch: Psychophysics, Physiology and Vicarious Touch Perception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Connor J. Haggarty, Adarsh Makdani, and Francis McGlone 7 Qualia, Brain Waves, and Spinal Reflexes: The Study of Pain Perception by Means of Subjective Reports, Electroencephalography, and Electromyography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elia Valentini, Sarah Vaughan, and Amanda Clauwaert 8 The Many Challenges of Human Experimental Itch Research . . . . . . . . . . . . . . . . Henning Holle and Donna M. Lloyd 9 Experimental Framework and Methods for the Assessment of Skin Wetness Sensing in Humans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Charlotte Merrick, Rochelle Ackerley, and Davide Filingeri 10 Skin-Mediated Interoception: The Perception of Affective Touch and Cutaneous Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laura Crucianelli and India Morrison

xi

109

129 161

181

199

xii

Contents

PART III

INDIVIDUAL DIFFERENCES, DEVELOPMENT, AND ILLUSIONS

11

Atypical Development of Tactile Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicolaas A. J. Puts and Carissa J. Cascio 12 Measuring Touch Sensitivity in an Aging Population . . . . . . . . . . . . . . . . . . . . . . . . Aldrin R. Loomes, Roberta Roberts, and Harriet A. Allen 13 Somatosensory Illusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tatjana Seizova-Cajic´, Regine Zopf, Martin Riemer, and Xaver Fuchs 14 Sensory Substitution: Visual Information via Haptics. . . . . . . . . . . . . . . . . . . . . . . . ´ rni Kristja´nsson, and Runar Unnthorsson Jack Brooks, A

PART IV 15

227 251 267

287

THE SOMATOSENSORY NERVOUS SYSTEM

16

Microneurography: Recordings from Single Neurons in Human Peripheral Nerves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 Rochelle Ackerley and Roger Holmes Watkins Electrophysiological Techniques for Studying Tactile Perception in Rats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 ¨ ztu ¨ rk, I˙smail Deveciog˘lu, Bige Vardar, Sevgi O ¨ c¸lu ¨ Fikret Taygun Duvan, and Burak Gu

17

Imaging Somatosensory Cortex in Rodents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mariangela Panniello, Severin A. C. Limal, and Michael M. Kohl 18 Imaging Somatosensory Cortex: Human Functional Magnetic Resonance Imaging (fMRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander M. Puckett and Rosa M. Sanchez Panchuelo 19 Electroencephalography of Touch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jonas Vibell, Helge Gillmeister, Alejandra Sel, Connor J. Haggarty, Jose Van Velzen, and Bettina Forster 20 Neurostimulation in Tactile Perception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luigi Tame` and Nicholas Paul Holmes

373

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

483

397 431

451

Contributors ROCHELLE ACKERLEY • Aix Marseille Univ, CNRS, LNC (Laboratoire de Neurosciences Cognitives – UMR 7291), Marseille, France HARRIET A. ALLEN • School of Psychology, University of Nottingham, Nottingham, UK CAROLINE BLANCHARD • Division of Clinical Neuroscience, School of Medicine, University of Nottingham, Nottingham, UK JACK BROOKS • University of Otago, Dunedin, New Zealand; Deakin University, Melbourne, VIC, Australia CARISSA J. CASCIO • Department of Psychiatry and Behavioral Sciences, Vanderbilt University Medical Center, Nashville, TN, USA MARIE CHANCEL • Brain, Body, and Self Laboratory, Karolinska Institute, Stockholm, Sweden AMANDA CLAUWAERT • Department of Experimental-Clinical and Health Psychology, Ghent University, Ghent, Belgium LAURA CRUCIANELLI • Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden I˙SMAIL DEVECIOG˘LU • Department of Biomedical Engineering, C ¸ orlu Faculty of Engineering, Tekirdag˘ Namık Kemal University, Tekirdag˘, Turkey FIKRET TAYGUN DUVAN • Catalan Institute of Nanoscience and Nanotechnology—ICN2, Barcelona, Spain DAVIDE FILINGERI • THERMOSENSELAB, Skin Sensing Research Group, School of Health Sciences, University of Southampton, Southampton, UK BETTINA FORSTER • Department of Psychology, City, University of London, London, UK XAVER FUCHS • Department of Psychology, University of Salzburg, Salzburg, Austria; Centre for Cognitive Neuroscience, University of Salzburg, Salzburg, Austria HELGE GILLMEISTER • Centre for Brain Science, Department of Psychology, University of Essex, Wivenhoe Park, Colchester, UK BURAK GU¨C¸LU¨ • Institute of Biomedical Engineering, Bog˘azic¸i University, I˙stanbul, Turkey CONNOR J. HAGGARTY • Human Behavioral Pharmacology Lab, Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL, USA HENNING HOLLE • Department of Psychology, University of Hull, Hull, UK NICHOLAS PAUL HOLMES • School of Sport, Exercise & Rehabilitation Sciences, University of Birmingham, Birmingham, UK; School of Psychology, University of Nottingham, Nottingham, UK ANNE KAVOUNOUDIAS • Laboratory of Cognitive Neurosciences, Aix-Marseille University, CNRS, UMR7291, Marseille, France KONSTANTINA KILTENI • Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden MICHAEL M. KOHL • Institute of Neuroscience and Psychology, University of Glasgow, Glasgow, UK; Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK

xiii

xiv

Contributors

´ RNI KRISTJA´NSSON • University of Iceland, Reykjavik, Iceland; National Research A University Higher School of Economics, Moscow, Russia CAROLINE LANDELLE • McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, QC, Canada SEVERIN A. C. LIMAL • Institute of Neuroscience and Psychology, University of Glasgow, Glasgow, UK; Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK DONNA M. LLOYD • Division of Psychology, Communication and Human Neuroscience, School of Health Sciences, University of Manchester, Manchester, UK MATTHEW R. LONGO • Department of Psychological Sciences, Birkbeck, University of London, London, UK ALDRIN R. LOOMES • School of Psychology, University of Birmingham, Birmingham, UK ADARSH MAKDANI • School of Psychology, Liverpool John Moores University, Liverpool, UK FRANCIS MCGLONE • School of Psychology, Liverpool John Moores University, Liverpool, UK; Institute of Psychology, Health & Society, University of Liverpool, Liverpool, UK CHARLOTTE MERRICK • THERMOSENSELAB, School of Design and Creative Arts, Loughborough University, Loughborough, UK INDIA MORRISON • Center for Social and Affective Neuroscience, CSAN, Linko¨ping University, Linko¨ping, Sweden ¨ ZTU¨RK • Department of Neurobiology, School of Medicine, Duke University, Durham, SEVGI O NC, USA MARIANGELA PANNIELLO • Institute of Neuroscience and Psychology, University of Glasgow, Glasgow, UK; Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK ALEXANDER M. PUCKETT • School of Psychology, University of Queensland, Brisbane, QLD, Australia NICOLAAS A. J. PUTS • Department of Forensic and Neurodevelopmental Sciences, Institute of Psychiatry, Psychology, and Neuroscience, King’s College, London, UK MARTIN RIEMER • Biological Psychology and Neuroergonomics, Technical University Berlin, 10623, Berlin, Germany ROBERTA ROBERTS • School of Psychology, University of Birmingham, Birmingham, UK ROSA M. SANCHEZ PANCHUELO • Sir Peter Mansfield Imaging Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK TATJANA SEIZOVA-CAJIC´ • Faculty of Medicine and Health, The University of Sydney, Sydney, NSW, Australia ALEJANDRA SEL • Centre for Brain Science, Department of Psychology, University of Essex, Wivenhoe Park, Colchester, UK LUIGI TAME` • School of Psychology, University of Kent, Canterbury, UK RUNAR UNNTHORSSON • University of Iceland, Reykjavik, Iceland ELIA VALENTINI • Department of Psychology and Centre for Brain Science, University of Essex, Colchester, UK JOSE VAN VELZEN • Department of Psychology, Goldsmiths University of London, London, UK BIGE VARDAR • Department of Electrical and Electronics Engineering, Faculty of Engineering and Natural Sciences, I˙stanbul Bilgi University, I˙stanbul, Turkey SARAH VAUGHAN • School of Psychology, University of Chester, Chester, UK

Contributors

xv

JONAS VIBELL • Department of Psychology, University of Hawai’i at Ma ¯noa, Honolulu, HI, USA ROGER HOLMES WATKINS • Aix Marseille Univ, CNRS, LNC (Laboratoire de Neurosciences Cognitives – UMR 7291), Marseille, France REGINE ZOPF • Macquarie University, Sydney, NSW, Australia; Lab for Autonomic Neuroscience, Imaging and Cognition (LANIC), Department of Psychosomatic Medicine and Psychotherapy, Jena University Hospital, Jena, Germany

Part I Discriminative Touch, Proprioception, and Kinaesthesis

Chapter 1 Detection, Discrimination & Localization: The Psychophysics of Touch Nicholas Paul Holmes and Luigi Tame` Abstract Detecting and discriminating touches on your fingertip and other highly sensitive body parts has been a paradigm in somatosensory science since the birth of psychophysics in the nineteenth century. By isolating a body part and applying discrete stimuli over many repetitions, the limits of somatosensation and bodily perception can be discovered. This chapter will focus on two methods of studying discriminative touch in the temporal and spatial domains: vibrotactile perception and spatial acuity. Different psychophysical approaches and experimental designs will be described and evaluated in terms of their validity, efficiency, and reliability. Practical and participant-specific difficulties will be noted. Vibrotactile and spatial acuity methods offer relatively cheap and reliable measures of somatosensation, often suitable for undergraduate student projects. Yet care and experimentation is required to ensure that the experimental design is adequate, and the data collection is sufficient to answer your theoretical question. Key words Detection, Discrimination, Vibrotactile, Grating orientation task, Cutaneous, Fingers, Behavior, Two-point discrimination

1

Introduction Neuroscience had barely begun, psychological laboratories did not exist, and the word psychophysics had not been coined when Ernst Weber began his research on tactile perception in the 1800s. Over several years, both at the University of Leipzig, and likely on the shores of Alpine lakes, Ernst and his brother Wilhelm probed each others’ bodies thousands of times, and recorded the results. In doing so they created the first psychophysical studies of touch, published (in Latin) in 1834 [1]. In this remarkably modern work, Weber describes the relative sensitivity of different body parts, the discrimination of two points on the skin, the differing spatial sensitivity along different body axes (Chapter 5, this volume), and between different body parts, spatial summation, effects of handedness, practice and fatigue, the role of attention and eye-movements, the interaction of tactile and muscle senses

3

4

Nicholas Paul Holmes and Luigi Tame`

(Chapter 3, this volume), of surface area and size, of surface area and temperature (Chapter 9, this volume), and more besides. Perhaps most importantly, in his systematic studies of touch, pressure, and temperature, and with his “just noticeable differences”, he laid the ground for his colleague Gustav Fechner [2] to formalize the discipline of psychophysics. Following the methods and the successes of psychophysics in the 1800s, psychological laboratories were set up around the world in the late 1800s and early 1900s. While the study of touch has been at the heart of psychology since its inception, its influence has waxed and waned during the twentieth century, with changing psychological paradigms and laboratories. Here we focus on detecting and discriminating touches to the skin. 1.1 Detection of Touches and Vibrations

The skin contains mechanical touch receptors of multiple kinds, but these are not uniformly distributed across the body. The detection of simple, point-like touches on the skin therefore varies with the body area (e.g., back versus foot), with the type of skin (e.g., sweaty or glabrous versus hairy), as well as more finely across the skin surface, depending on receptor density and type (Chapter 5, this volume). Systematic study of detection thresholds for these simple stimuli was advanced by von Frey’s “aesthesiometer” [3]—a series of different-diameter hairs (cat, human, horse) attached to wooden handles. When pressed vertically into the skin, the hairs begin to bend at a given force. After bending, the force stays approximately constant. For each point on the skin, the pressure threshold is the lowest diameter of hair which can reliably be felt. We do not detect pressure. When immersing some skin in water, we only feel the pressure gradient at the boundary of air and water. Our sense of pressure comes from pressure-induced deformation of the skin—stretch. Weber realized this with his experiments on human hair—keeping the skin and muscles of the scalp static while pulling on a hair prevented perception of the direction of motion [1]. When a vibrating probe, solenoid, or membrane is placed onto the body, the skin is repeatedly stretched and relaxed at the vibration frequency. Von Be´ke´sy described the traveling waves of skin deformation that are set in motion by a vibrating probe [4]. A single probe at the wrist might set up traveling waves along the skin that reach as far as the shoulder, or further. Our detection and localization of touches on the skin must therefore take account of this fact; mechanisms of spatial and temporal inhibition allow us to localize touches to a particular body part, but they can also create illusions [5], referred sensations, and other forms of sensory “funneling” [4, 6] (Chapter 13, this volume), and other tactile [7] and haptic [8] illusions. The apparently simple ability to detect touches on the skin is further complicated by the surface area and duration of stimulation, and by interactions between sub-modalities of somatosensation—

Detection, Discrimination & Localization

5

temperature in particular. In an intense series of studies, using adaptation, masking, skin temperature manipulation, and systematic investigation of the relationship between stimulus intensity and frequency, twentieth-century tactile psychophysics tested hypotheses about the number and kind of independent tactile “channels” in the skin, much from studying vibration perception [4, 9– 12]. The resulting “four channels” of cutaneous mechanosensation map approximately onto four general classes of cutaneous mechanoreceptors (Chapters 15, 16, 17, and 18, this volume). The rapidly adapting (RA) Pacinian (P) channel has the lowest thresholds of all, with optimal sensitivity at around 200 Hz. Of the other three non-Pacinian (NP) channels, one is rapidly adapting (RAII), and two are slowly adapting (SA). Any individual cutaneous experience is likely generated by activity in all four channels [13]. Further, “touch-blends” such as wetness or stickiness [14] derive from interactions between these four cutaneous mechanoreceptor channels, and the additional somatosensory channels of warming, cooling, irritation, and pain (Chapters 6, 7, 8, and 9, this volume). This volume as a whole may do some justice to the complexity of somatosensory experience. This chapter focuses first on simple methods to study the perception of discrete mechanical vibrations, and second on discrimination methods for assessing tactile spatial resolution (acuity). 1.2 Tactile Discrimination and Tactile Spatial Resolution

Weber devised and developed the classic “two-point” discrimination method, in which a pair of compass points or some other contraption is pressed onto a participant’s skin with different separations between the two points [1]. The participant is required to say whether they feel “two” distinct points or just “one”. “Catch” trials are added in which just one point is stimulated. Weber argued that the method produces a reliable index of the size and orientation of the “sensory circles” of the skin, what we might call “receptive fields” (or perhaps “dermatomes”) now (Chapter 5, this volume) and of tactile spatial acuity. Weber, his students, and his followers used this two-point method for many decades—and some still do today—as a measure of tactile acuity across the body surface. But, as pointed out by Boring [14], in studies as early as 1915, the two-point method was shown not to provide a reliable measure of tactile spatial acuity [15]. The first problem is that the perception of two points on the skin does not simply or discretely change from “two” to “one” as the points are brought closer together. Instead, there is a range of feelings—for example, an elongated or extended surface—and the distinction between “two” and “one” is blurred. To make a decision, participants must set a personal “criterion”— an arbitrary point to divide their personal spectrum of sensation into two categories. People differ in how they set this criterion, both between participants as stable individual differences, and

6

Nicholas Paul Holmes and Luigi Tame`

between individual tests over time or under differing experimental conditions, so the measure of spatial resolution is compromised. Good experimental design can minimize this problematic criterion (Subheading 3.2). A second problem is that the difference between “two” and “one” is not simply due to the spatial resolution of the skin. These stimuli also differ in “length” or “magnitude” and participants can learn to discriminate between one and two points at distances which should be impossible, given the known size of the cutaneous receptive fields, and the density of mechanoreceptors that are hypothesized to underlie tactile acuity. The flaws in the two-point method were well-known since at least 1915, but the decisive study came 66 years later. In a series of experiments in animals and humans, Johnson and Philips [16] trained participants over many hours, and found that people could reliably discriminate one and two points separated by 0 mm. This ability cannot be due to the spatial resolution of cutaneous receptive fields. Rather, it must have been due to the discrimination of some other non-spatial stimulus property such as magnitude. Neurophysiological data from the monkey peripheral nerve supported this claim [17]. Forty years after these results, many researchers still claim that the traditional two-point method provides a measure of tactile spatial resolution (e.g., ref. 18). Improvements can be made upon the standard two-point technique by adopting criterion-free methods, for example using two-interval forced-choice designs, in which a single point is presented in one (randomized) interval, and two points in another. This forces participants to say in which interval the two- (or one-) point stimulus occurred. This deals with the criterion problem, but not with the magnitude problem. Other researchers have instead revised the task into a “two-point orientation discrimination” task, in which two points at a fixed distance are first presented, for example, along the principal axis of a finger, then presented again, orthogonally, across the axis. The participant’s task is to say which of the two intervals was “along”. Alternatively, in a single-interval design, to say “along” or “across” for each stimulus. This simple manipulation deals effectively with the criterion problem, and likely also deals with the stimulus magnitude problem [19]. One alternative method for tactile spatial resolution, widely regarded as more reliable than two-point discrimination, is the grating orientation task [16]. This uses a series of small domes (e.g., ref. 20), each of which has a square-wave grating of, at minimum, 0.35 mm width (i.e., a wavelength of 0.7 mm). In a single trial, participants are presented with one or two gratings, and their task is to discriminate the orientation of the grating—typically “along” or “across” the finger. The grating width might start at 3 or 4 mm, and descend according to a staircase, or a constant range

Detection, Discrimination & Localization

7

of widths might be presented to the participant. The measure of tactile spatial acuity is then the smallest grating width whose orientation can be discriminated better than chance, or the grating width that gives ~75% correct discrimination, depending on the threshold algorithm used. We recently assessed the effects of different experimental designs (1- versus 2-interval), participant sex, hand used, hand posture, and hand visibility on the grating orientation task [21]. Task, sex, and hand posture all significantly affected tactile spatial resolution, but we found no effects of left versus right hand, or of hand visibility. 1.3

2

Summary

In the rest of this chapter, we will describe what we have learned from using these two methods—detecting and discriminating vibrations, and discriminating the orientation of gratings. We set up our equipment on a low budget, often constructing it ourselves. While this may not be not ideal for the precise and reliable measurement of the limits and neurophysiological mechanisms of cutaneous perception, we have found that care over the experimental design, the training of participants, and the collection of sufficient data was often more critical to us as experimental psychologists than the quality or precision of the apparatus.

Materials

2.1 Presenting Pressure Stimuli (Vibrations) to the Fingertip

The following list is in approximate order of complexity and/or stimulus quality. Figure 1 shows a range of hardware options.

2.1.1

Perhaps the simplest, cheapest method of presenting vibrations to the skin is to use the electromagnetic membrane provided by a loudspeaker. The speakers could be used directly, or the membrane taken out and customized. The advantages are that the hardware is cheap and readily available and can be plugged straight into a standard computer’s audio port. Loudspeakers can provide a sensitive stimulus operating at low voltages. The disadvantages are that you may need an additional amplifier, the loudspeakers may have a limited frequency range—they’re optimized for our auditory, not our tactile systems—they may be large and resonant, with unwanted spread of vibration, and precise measurement and control of the amplitude presented to the skin may be difficult.

Loudspeakers

2.1.2 Bone-Conducting Hearing-Aid Vibrator

A convenient “off-the-shelf” option that we have used over many years, thanks to passing through Charles Spence’s laboratory in the early 2000s, is a bone-conducting vibrator produced by Oticon (Fig. 1d). For stimulus frequencies of around 100–1000 Hz, this

8

Nicholas Paul Holmes and Luigi Tame`

Fig. 1 Some of the vibrotactile stimulators typically used to investigate tactile performance. In the left side of the image there is a depiction of the device and on the right side the module(s) that is(are) used to stimulate the body part of interest. (a) “Tapper” or solenoid stimulators that produce a mechanical stimulation via a small electromagnet. Supplied by Dancer Design. (b) Piezoelectric stimulator that produces a mechanical stimulation via voltage-dependent bending of a piezoelectric chip and can be made EEG, MEG, and MRI compatible. Supplied by QuaeroSys. (c) Electrical ring stimulators that produce an electrical stimulation of the skin or nerves, for example, the fingers and the digital nerve. Supplied by Digitimer. (d) Oticon boneconducting stimulator driven by a standard computer sound card and headphone socket that produces a mechanical stimulation of the skin

device seems to provide quite reliable amplitudes of stimulation, and we have been able to measure very good-looking frequencyamplitude curves for detection thresholds (Fig. 2). Below about 100 Hz, these vibrators do not work well. The advantages are the small size—highly suited to fingertips—they’re purpose-built for applying vibrations to the skin and can be plugged directly into a low voltage signal source (e.g., a PC, or via an amplifier). Moreover, if the stimulator is securely attached to the skin, they can produce

Detection, Discrimination & Localization

9

Fig. 2 Vibrotactile detection thresholds as a function of frequency. Data are shown from four participants, each of whom completed between one and four repetitions of a two-interval task to detect a single 500 ms sinusoidal vibration on the pad of their index fingertip. Target frequency varied from 70 to 1400 Hz (x-axis), and thresholds varied from around 0.02 to 0.71 (y-axis, log scale, arbitrary units). Within an individual participant, thresholds varied over a 10- to 30-fold range, with lowest thresholds in the Pacinian range, at 150–250 Hz. The data confirmed that our Oticon vibrators worked well at frequencies from about 100 to 1000 Hz, and that our stimuli could be presented effectively at very low amplitudes

very similar stimulation regardless of the position in space of the body part stimulated, for example whether the hand is turned palm down or palm up. The disadvantages are the limited frequency range and lack of flexibility for stimulating different body areas. Note 4.1 discusses differences in sensitivity across vibrators and time. 2.1.3

Solenoid

Likely the most controllable and reliable simple stimulus for cutaneous mechanoreception is the solenoid or “tapper”. This can provide a single discrete tap, a series of discrete taps, or can be controlled using continuously varying amplitudes, for example in sinusoidal waves. Advantages include being very cheap, readily available, low voltage, easily controllable, and programmable. Dancer Design (Fig. 1a) provides a range of solenoids suitable for skin stimulation. The disadvantages are that they may not be as sensitive as a loudspeaker of the same size and they are “sticky”— there is some inertia which needs to be overcome at very low stimulus amplitudes, so small or weak solenoids may be unable to measure absolute threshold at the fingertip. In practice, when using a staircase method for finding detection thresholds, the output of a small solenoid at very low intensities is likely binary—tap or no tap—rather than the continuously varying amplitude required for precise thresholds. This concern likely applies to most tactile stimulators, but we first noticed it for solenoids.

10 2.1.4

Nicholas Paul Holmes and Luigi Tame` Piezoelectric Chips

Piezoelectric chips use a non-magnetic means of moving the skin. When an electric current is passed through, the piezoelectric crystal deforms with an amplitude proportional to the voltage. This deformation can be used to drive a probe that contacts the skin. One major advantage of piezoelectric chips is they can be made compatible with magnetic resonance imaging (MRI) and magnetoencephalography (MEG) methods, and can be versatile, with an extremely precise control of stimulus timing, on the order of a few milliseconds. The clear disadvantages are the need for a more sophisticated control apparatus to control the high voltages required, which makes this a more expensive option. QuaeroSys provides a scalable system with up to 16 tactors (Fig. 1b). Their hardware is a mechanical tactile stimulator that uses several pin matrices and other devices which can apply mechanical forces. The system consists of a master device which communicates with a personal computer via USB, and several stimulation modules which are slaves of the master device. The master device has separate ports for input and output trigger signals, enabling precise synchronization between stimulation and measurement systems. For more complex experiments, the system can be upgraded with up to 16 cards for stimulation or trigger input or output purposes. The timing accuracy is around 0.5 ms. We have successfully used this hardware both in fMRI [22] and MEG [23] settings (see also Chapter 18, this volume).

2.1.5 Pneumatic (Air-Puff) Stimulation

We have limited experience with pneumatic apparatus [24– 28]. However, we have used the MRI-compatible air-puff stimulator “dodecapus” [29]. This device is driven by an air compressor that can be positioned in the scanner control room, and provides input to a 9-way solenoid manifold valve (“S” Series Valve; Numatics Inc., Highland, MI) that is controlled by transistortransistor logic (TTL) pulses. For fMRI experiments, several plastic air tubes from the manifold valve pass through waveguides into the scanner room, where they connect to a block mounted beside the body area to be stimulated. The block serves as a rigid base for several flexible tubes with nozzles (Loc-Line Inc., Lake Oswego, OR), flexibly arranged to direct ~50 ms air puffs (at an input air pressure of 3.5 bar) at various locations on the skin. Notably, such an approach does not require continuous contact with the skin surface.

2.1.6

Other systems, such as hydraulic [30], are available, and all systems come with their own advantages and disadvantages. The cost and scope of your project is likely to be the main factor driving your choices. If you plan to run only a small number of experiments, for example during a student project, then a cheap and simple setup which plugs into a standard computer can be made to work well,

Summary

Detection, Discrimination & Localization

11

particularly for more “cognitive” experiments. But if your goal is precise measurement of absolute detection thresholds, precise manipulation of stimulus size, amplitude, temperature, or frequency, or careful neurophysiological or neuroimaging work, then more sophisticated hardware and software will be needed. Most of our psychophysical work to date has been of the former kind, and we detail in this chapter how well we’ve been able to make this work, even for relatively precise measurements of the threshold. 2.2 Presenting Gratings to the Fingertip 2.2.1

Gratings

Depending on your target area of skin, you will need a range of grating widths, from 0.35 mm if your target is the lip or fingertip, to over 10 mm if you target large and relatively insensitive portions of the skin. If you are testing special populations—children, older adults, or people with relevant medical conditions—then you will also need an expanded range of gratings. For work on healthy adult fingertips, we have used gratings with widths of 0.35, 0.5, 0.75, 1.0, 1.25, 1.5, 1.75, 2.0, 2.25, 2.5, 2.75, and 3.0 mm (Fig. 3). A set of acrylic gratings can be purchased via medical or scientific equipment suppliers (search for “JVP domes”). These stimuli are named after Johnson, van Boven and Phillips [15–17], and are the standardized set used in many published reports. If you are on a lower budget, then you can design and 3D print your own domes, or send away for them to be 3D printed. This is the approach we took, printing at ShapeWays.com (see Note 4.2). Our “mushroom” designs and .STL files which are ready for sending to a 3D printer can be viewed in free computer-aided design software such as MeshLab, and are available here: https://osf.io/da893/.

Fig. 3 Grating orientation discrimination task (GOT) and apparatus. (a) Wooden box (45 cm wide, 50 cm deep, 25 cm high) housing two ethernet-controlled stepper motors, an acrylic disc (38 cm diameter) holding 12 × 2 3D-printed plastic domed gratings with ridge widths of 0.35, 0.5, 0.75, 1.0, 1.25, 1.5, 1.75, 2.0, 2.25, 2.5, 2.75, 3.0 mm and diameter 2.5 cm at their widest point. (b) Top view of the grating wheel. Gratings were arranged in ascending order of grating widths, alternating across and along (clockwise from ~7 o’clock in Fig. 3b). A laser pointer is used as a light-gate timing signal. (c) Two gratings photographed in position on the wheel next to a schematic finger at approximately the same scale. When rotated into position, the gratings run across and along the long axis of the finger. (Reprinted in part from French et al. [21])

12 2.2.2

Nicholas Paul Holmes and Luigi Tame` Presentation

As with all tactile stimuli, presentation can be manual or automated, depending on the needs of the experiment and the availability of equipment and expertise. Automated presentation is desirable to reduce experimenter-associated variability, however for some applications, robot-delivered touch may feel less natural and less social than human-delivered touch (Chapter 6, this volume), and also generate a different response [31]. If experimenter-applied manual presentation is used, the stimuli are simply pressed into the skin of the participant. The experimenter should aim to minimize variability in the skin location, the angle, the force [32], the duration and, critically, the orientation. JVP advises placing a mark on the target location, but additional constraints could be added. The skin target could be placed in the aperture of a flat surface, force could be monitored by an in-series transducer, or controlled by in-series spring damping, and a metronome or other timing device could be used. We have used a laser pointer and laser-detecting diode just below the skin target both to monitor stimulus duration and to trigger peripheral hardware during experiments. To check and calibrate stimuli before experiments, we can also place a simple filament force transducer between the skin and the stimuli to estimate the likely onset of the stimulus. Automated control is desirable. Several systems can be created from existing designs (e.g., the TAPS device in ref. 33), or you could design your own. The requirements are for a system first to hold and move into place a range of grating stimuli at two or more orientations, and second to press and release the grating into the skin. We opted to mount 24 gratings (2 orientations each of 12 widths) in a large plastic (acrylic) disk. The disk was mounted onto one National Instruments ethernet-controlled integrated stepper motor (ISM-7411) that provided rotation, to move the required grating into place, underneath an aperture. The rotation motor was mounted onto a second stepper motor which tilts the disk up and down, to bring the grating into contact with the finger. The two motors are controlled using National Instruments’ LabView and SoftMotion. By dividing the disk’s 360° into 24,000 steps, we can give position, velocity, and acceleration commands to rotate the disk between the 24 gratings relatively smoothly, resulting in very reliable positioning. Because of the weight of the disk and motor, however, tilting the disk up and down occurs less smoothly. Using a laser-pointer and diode, and an in-series force transducer, we were able to calibrate the tilting motor to reduce the variability (range) of stimulus onsets to around 25 ms, while the range of stimulus durations achieved was ~300–350 ms. For behavioral experiments, this temporal variability seems acceptably small, but we have not yet been successful in using brain or nerve stimulation to disrupt grating discrimination performance (Chapter 20, this volume), which likely requires greater temporal precision.

Detection, Discrimination & Localization

13

Software

The majority of our vibration perception work has used Matlab, PsychToolBox 3, and the QUEST adaptive staircase procedure (our Matlab code and toolbox are here: https://osf.io/ea3fd/). Any programming language should work, as long as it can output the required stimulus waveforms to the peripheral hardware. This could be via a computer audio output, serial port, parallel port, USB port or, for example, a dedicated National Instruments card. Control of stepper motors and other complex moving hardware for grating orientation experiments likely requires more sophisticated software (see Chapters 2, 3, and 4, this volume). The LabView code we use is here: https://osf.io/ea3fd/.

2.4 Calibration and Quality Control

Never assume that your stimulus will be presented as you had intended! Wherever possible, some form of measurement, calibration, and monitoring of the stimulus actually presented is desirable. At minimum this should be done before data collection begins. Ideally, it would occur during all experiments, so long as it does not interfere with the stimuli or task. For vibrotactile experiments, the stimulus waveform can be fed into a data-acquisition device (e.g., National Instruments, Cambridge Electronic Design, BioPac, AD Instruments, or a basic oscilloscope) to check on stimulus timing and frequency. For tactile waveforms, a sampling frequency of 8000 Hz has been sufficient—at minimum, the signal must be sampled at twice the maximum frequency of the stimulus. To confirm the stimulus that was actually presented through the device, a sensitive microphone can be placed above the vibrator and fed into a data acquisition system. An intermediate level of precision can be obtained via a cheap accelerometer chip stuck onto the vibrating surface. The output of this chip can be fed into the data acquisition device. Voltages in several different dimensions can then be converted to accelerations and distances. These chips are small, cheap, and sensitive. The first author has destroyed several chips by careless touch, static electricity, and transient voltages when inserting batteries—include a physical “off” switch when wiring-up your chip! For the ultimate stimulus amplitude measurement, we once approached colleagues in Physics to borrow a laser interferometry setup to measure the presented vibration more precisely. While this remains a goal, we did not follow-through, as none of our scientific questions have yet required the micrometerlevel precision that such calibration methods could provide. For the grating orientation task, our approach has been to use a laser-pointer and a light-detecting diode to provide stimulus timing information during experiments. This is critical for triggering other peripheral hardware like transcranial magnetic stimulation (TMS, Chapter 20, this volume), and force monitoring during experimental design and setup. Laser and force outputs are fed into the data acquisition device and co-registered with any other stimuli or responses.

2.3

2.4.1 Stimulus Timing and Amplitude

14

Nicholas Paul Holmes and Luigi Tame`

2.5 Environmental Conditions

3

Anecdotal evidence from our laboratory suggests that the humidity and temperature of the testing room can be critical to the ability of some participants to perceive low-amplitude vibrations on their skin. We have no useful data on this, so we can only warn researchers to take this into consideration [34]. Possible participant-specific issues are considered in Note 4.3.

Methods

3.1 Experimental Tasks

We have found that the choice of experimental task can be critical, for example, to whether and how primary somatosensory cortex seems to be involved [35, 36], and the effectiveness of brain stimulation (Chapter 20, this volume). In our work we have focused on two main tasks to date (detection and discrimination), but many others are possible—for ideas, please see the rest of this volume. Our understanding (and definitions) of these tasks is given here.

3.1.1

In “detection” tasks, a single target stimulus (or no stimulus, on “catch” trials) is presented to the participant’s skin, and their task is to respond according to whether or not they detect the presence of the stimulus. Different experimental designs are possible (Subheading 3.2), but the two most common would be a single and a two-interval task. In a single interval task, half of the trials should contain no stimulus and half contain a stimulus; the participant should respond “yes” or “no” to target presence. Another design involves two sequential temporal intervals where one interval contains a stimulus and one contains nothing; the participant should say “first” or “second” to identify which interval contained the target stimulus. Both single and two-interval designs could involve a “pedestal” manipulation—a constant stimulus is presented, and the participant’s task is to detect a small increment or change in the stimulus at a certain point [37]. We have also run experiments where another body location receives a distractor or “masker” stimulus, presented in every interval [38]. The participant’s task is always to focus on the target location and ignore the irrelevant distractors. We think of “detection” tasks as ones in which the participant only needs to perceive that something is different or present in trials or intervals with a target—it doesn’t matter what it feels like or where it is, they just have to know that something happened on their body. The result of a detection task could be a threshold—the lowest possible intensity that the participant can detect (Fig. 2), a proportion correct (e.g., for a series of stimuli presented near threshold), or a value of the signal detection parameter d-prime. For us, detection is the most fundamental, simplest tactile task.

Detection

Detection, Discrimination & Localization

15

3.1.2

Discrimination

In what we think of as “discrimination” tasks, participants are presented with two (or more) alternative stimuli, each of which is readily detectable (i.e., well above threshold). The participant’s job is to say what kind of a stimulus is presented. The two tactile stimuli could differ in many ways—different intensities (amplitudes), different frequencies, different locations, orientations, sizes, different patterns or waveforms—but the key things for us is that they are physically different, that there is always a correct response, and that the stimuli are easily detectable on their own. As for detection tasks, discrimination tasks can be presented in a single interval (e.g., is this a “strong” or a “weak” stimulus?), in which case we could call it a “classification” task, or two intervals (e.g., in which interval was the “strong” stimulus presented?).

3.1.3

Localization

A simple task which we have not used much is localization—participants need to say where on the body a stimulus was presented [39]. This task could be combined with the “detection” or “discrimination” tasks above—the stimulus to-be-localized could be below, near, or well-above the detection threshold. Because stimulus location on the body is manipulated, these tasks need to take into account the very different tactile sensitivities and abilities of different body parts, and even of different locations within a body part (Chapter 5, this volume). Since we are less familiar with this task, in the rest of this chapter we will focus on detection, discrimination, and spatial orientation tasks.

3.2 Experimental Design

In this section, we review the typical choices that need to be made in designing simple tactile detection and discrimination tasks.

3.2.1 One Interval or Two?

In single interval designs, the participant only gets one opportunity to experience the stimulus (or the absence of it), and must make an absolute judgment—“Is it there?” “Is it horizontal?” A two-interval design gives participants two opportunities, allowing relative judgments to be made—“Was the first one stronger than the second?” Performance is generally better for these two-interval, relative judgments. One clear advantage of single-interval designs is time—the total experimental duration can be approximately halved relative to two-interval designs, dealing much better with fatigue and attentional fluctuations. A second advantage is simplicity—it is very easy to explain to participants how to make absolute judgments about single events—if you feel it, press the left button, if you don’t press the right. Press the left one if it’s strong, the right if it’s weak. Our participants, typically from psychology courses, find single-interval tasks quite straight-forward. But the main disadvantage of single interval designs, in our experience, is that participants’ response criteria become critical.

16

Nicholas Paul Holmes and Luigi Tame`

While participants understand very clearly and intuitively that they should press one button if they feel something, participants seem to differ wildly in their expectations about what it means to “feel something” and/or their ability to feel it. When running a tactile detection experiment and training participants, it is an almost-daily occurrence that the participant says: “I can’t feel anything”. When the experimenter places their finger on the vibrator, the stimulus is bright, clear, and strong! Partly, this is due to between-participants variation in absolute detection thresholds (e.g., the effects of age and sex [40]), but we believe it is mostly due to task comprehension, training, and/or instruction (see Note 4.4). With a longer, more laborious two-interval design, participants are able to focus on only what is different between the two temporal intervals. While the participant may say “I feel nothing clearly” after a single interval, they can much more easily be coached into saying “OK, something was different in the first interval” or “it was slightly stronger in the second interval” or “the first one felt a bit more vertical”. That difference gives participants something to focus on and learn from. These differences between one- and two-interval designs are important. Here are two examples. First, in the classic “blindsight” phenomenon, patients with damage to their visual pathways deny being able to see a stimulus, yet can guess better than chance in a two-alternative forced choice design. This dissociation is partly due to the experimental design (one- vs. two-interval), partly to patients’ response criteria (“I see nothing” vs. “something was different before”), and of course partly to how the visual brain works [41]. Second, the first author had his sight tested by the local opticians recently. They used a two-interval psychophysical staircase design to measure the prescription for his new glasses. For 30 years his sight has been tested in this way—there must be good reason! Other designs are possible. One design commonly used in visual studies is a “delayed match to sample” design for stimulus discrimination (e.g., ref. 42). The first interval contains one of two possible stimuli, and the second interval contains both possible stimuli. The participants have to say which of the two stimuli was presented in the first interval. (The intervals can also be reversed.) One advantage may be that the participant does not have to remember what the two possible target stimuli are when they make their discrimination response. They need to remember what they just felt, then recognize the matching one. This contrasts with both single-interval designs in which the target stimulus (or an internal representation of it) must be kept in mind over multiple experimental trials; and with two-interval designs in which the participant must keep both intervals in memory for a short time before making

Detection, Discrimination & Localization

17

a response. These task demands may be critical in different experiments or interventions [35] (Chapter 20, this volume). 3.2.2 Yes/No, Magnitude, Forced-Choice, or Confidence?

Closely related to the issue of how many intervals to present is: what question do I ask the participant? This of course depends mostly on the research question, but exactly how you frame the question to the participants can have important effects on their responses [43, 44]. Yes/No If a single interval design is used, the most simple question is perhaps: “Did you feel anything?” to which the answer could be given verbally, manually, or using foot pedals: “Yes” or “No”. This provides participants with a very intuitive question and answer which likely matches their everyday experience of feeling things happen on or in their body. The problems with this approach are that different participants set different response criteria—how much of a feeling do they need to justify the response “Yes”? Are they very conservative, mostly saying “No”, or very liberal, mostly saying “Yes”? If your research question concerns participants’ conscious perception or awareness of a stimulus, then Yes/No might be the best response options to measure. If you are more interested in neurophysiological thresholds and less in participants’ conscious perception, we would advise against using Yes/No responses unless the response criterion is carefully accounted for. Magnitude Instead of asking for a binary “Yes/No” response, participants could be prompted to rate the intensity of a single stimulus, or to rate its magnitude on any other dimension (e.g., frequency or pleasantness). We have not used this response option, but in studies of non-discriminative touch, including affective, pain, pruritic, temperature, and wetness submodalities, it is an essential tool (Chapters 6, 7, and 8, this volume). Same/different With two-interval designs, or designs in which there is a known standard stimulus, participants can be asked to respond whether a comparison stimulus is the same as or different from the known standard. This can be done with two temporal intervals (requiring a small memory demand) or with two stimuli presented together. A related design is the temporal order judgment (TOJ) task in which participants receive two stimuli which begin either at the same or at different times [45]. An advantage of the same/different design is that participants are required to respond only to any difference that they perceive between the two stimuli. Our intuition is that this design will be more sensitive to participants discriminating these differences than an alternative design, in which the relative magnitude of the stimuli must be compared (e.g., first/second and higher/lower). The consequent disadvantage of the same/different design is that you cannot know

18

Nicholas Paul Holmes and Luigi Tame`

which stimulus the participant felt had the greater magnitude. Worse, if the stimulus parameter that you are manipulating behaves non-linearly (e.g., vibrotactile frequency and perceived intensity— Note 4.5), then a difference above the standard magnitude will not correspond to the same difference measured below the standard magnitude. The resulting threshold or just noticeable difference will therefore be some kind of average of a supra-standard difference and a sub-standard difference. One additional problem with same/different designs is that, just as in Yes/No designs, participants need to set a personal criterion—what counts as “different”? [46]. We have seen this cause particular problems in temporal order judgment (simultaneity judgment) tasks—someone with a very liberal criterion who says “different” to almost every stimulus pairing will appear to have a very sharp simultaneity tuning curve; a participant with a conservative criterion may produce a broadly tuned simultaneity curve. Yet, the perceptual abilities of these two participants may be identical. You cannot afford to ignore the effects of response criteria in sensory psychophysics, or any similar decision-making task. Higher/lower Perhaps more informative is to ask the participant whether one stimulus is higher or lower along some dimension than another stimulus, either an implicit standard (e.g., from memory), or an explicit standard (e.g., presented in the first interval). Typical questions could be: “Is the comparison (or second) stimulus stronger or weaker than the standard (or first) stimulus?” Or: “Is it higher frequency?” (though see Note 4.5) Or: “Does it last longer than the first stimulus?” These example questions ask only about these basic parameters of vibrotactile intensity, frequency, and duration. Similar questions could be asked about shape, size, angle, roughness, speed, or any other more complex stimulus parameters (Chapters 2, 3, 4, 5, 6, 7, 8, 9, and 10, this volume). First/second One of our most-used questions is to ask the participant: in which interval was the target stimulus? This works well for both detection [38] and discrimination tasks [35]. If your experiment requires multiple kinds of judgment or task, keeping the same task design throughout can be helpful and intuitive to the participant. One advantage of the first/second question is that the target can then be defined in any way you like—the mere presence of a stimulus (“detection”), the presence of a particular constant stimulus (e.g., weak vs. strong) relative to varying comparison stimuli, or to make any other judgment such as when was the higher-frequency stimulus, or which was the most intense (see Note 4.5). Confidence Most experimental designs ask just one question about the target stimulus, as detailed above. But a richer account of

Detection, Discrimination & Localization

19

participants’ perception could be gained by adding a “confidence” judgment. Instead of only giving binary responses: “yes” or “no” or “same” or “different”, participants could be encouraged to rate the confidence of their responses, on a scale [47, 48]. We recently piloted using this form of response along with a forced-choice location discrimination task (index vs. middle finger), asking participants to provide a judgment on what they felt: “nothing”, “something”, “mostly clear”, and “clear”. We have not completed this experiment, but this form of experimental design may be critical in distinguishing between so-called “numbsense” (a tactile equivalent of “blindsight”) and simple task-dependent changes in response criterion [35, 49]. 3.2.3 Manual, Pedal, or Vocal Responses?

Experimenters should make the responses to their experimental question as simple as possible, ensuring that giving the responses does not interfere with the tactile stimulus or the psychological process studied. If tactile stimuli are presented to one or two fingers, it might be best to use foot pedals to collect the responses. In a recent experiment, we piloted using 2 feet to respond to one aspect of the task (target location—index vs. middle finger), and the fingers of the non-stimulated hand to provide a confidence rating about the first response (Subheading 3.2.2). This was extremely confusing and was abandoned immediately, opting instead to collect a foot response followed by a vocal confidence rating. Our main advice here is to pilot-test the experiment on yourself and, critically, on some naı¨ve participants. Responding to the experimental question should be as simple and intuitive as possible—any extra attention that the participant requires to give responses is attention lost to focus on the stimuli and task. A related issue is whether to require participants to respond as quickly as possible (emphasizing response speed), as accurately as possible (emphasizing response correctness), or both. This depends mostly on the experimental question and outcome measure (e.g., reaction times versus thresholds), but there is always a trade-off [50]. Neither participants nor experimenters want to stay in the lab all day, so being perfectly accurate at a task is rarely an option, and is also often uninformative due to ceiling effects. But we also want to ensure that participants’ brains are processing all relevant aspects of the stimulus, so we also don’t only want the fastest-possible responses if they come at the cost of many errors. Our recommendation is to record and report both reaction times and errors, to instruct participants consistently, but to acknowledge the constraints of the testing session and the social context of the experimental situation [51].

20

Nicholas Paul Holmes and Luigi Tame`

3.2.4 Constant or Varying Stimulus Magnitude?

One major problem in tactile, and all sensory, research is to calibrate stimulus magnitude for different participants. People’s ability to detect and discriminate, and their evaluation of sensory stimuli can vary over a remarkably wide range, as much as ten times or more in our experience (e.g., ref. 52). These massive betweenparticipant differences in sensitivity can require quite different ranges of stimuli to be presented. The challenge is to select the correct range of stimuli for each participant in as short a time as possible so that the main experimental questions can be addressed. Despite many years of trying, this remains very difficult. Three approaches are described: constant stimuli, varying stimuli, and a compromise between the two. In the classic psychophysical method of constant stimuli, the same fixed set of stimuli is presented to the participant a set number of times, and all participants make their judgments on the same set. For example, vibrations of seven different amplitudes, varying from below the threshold of the most sensitive person, to above the threshold of the least sensitive person, are presented ten times each. From these 70 trials, a curve can be fit through the proportion of stimuli detected on the y-axis as a function of the vibration amplitude on the x-axis. What counts as “chance” or belowthreshold performance will depend on the experimental design— 0 in a single interval yes/no design, or 0.5 in a two-interval or two-alternative forced-choice design. Different curves or distributions can be fit to the data (e.g., linear, sigmoid, Weibull), and different measures can be extracted for each participant (threshold, bias, just-noticeable difference, slope), depending on the experimental question. The advantages of using constant stimuli are that the same stimuli are used for all participants, and the data can provide estimates of both the mean and the variability of participants’ performance. The disadvantages are that the range of stimuli may not be appropriate for some participants, that task difficulty varies with sensitivity so that some people may find the task trivially easy and others impossibly difficult, and that many more trials may be required. An alternative approach is to vary the stimulus magnitude depending on the participant’s responses to previous stimuli. In general these are called “adaptive” approaches, and often use a “staircase” method to set stimulus magnitude on each trial— going up and down in magnitude after each trial or set of trials. We have most experience with the QUEST (quick estimation of threshold) method [53], which is an example of a PEST technique (parameter estimation by sequential testing). The QUEST algorithm is implemented in Matlab and PsychToolBox [54]. It works first by creating a statistical distribution of the parameter being estimated (e.g., a detection threshold), then by suggesting an informative stimulus magnitude for the next trial. After the trial is run, it updates the distribution, given the stimulus magnitude

Detection, Discrimination & Localization

21

Fig. 4 Example adaptive staircase thresholding procedure using QUEST. In each of 48 trials in the block (x-axis), the QUEST algorithm estimates a mean (circle) and standard deviation (error bars) of the underlying distribution that represents a participant’s threshold. The threshold (y-axis, arbitrary units, A.U.) could be anything, for example, the amplitude of a vibration to be detected, the frequency difference between two vibrations to be discriminated, or the duration of a gap within an otherwise continuous vibration. As the trials progress, the QUEST algorithm obtains more information about the participant’s performance, so it is able to increase the mean and standard deviation (following an error) or decrease the mean and standard deviation (following a correct response). In general, we have used between 30 and 50 trials per staircase, but the staircase could also be ended by using a minimum value for the standard deviation (e.g., 0.05 might be a good criterion in this example)

actually presented and the response obtained (0 for incorrect, 1 for correct). QUEST requires several parameters to set-up the initial distribution, most importantly an approximate mean and standard deviation of the parameter to be estimated. After around 30 iterations of this suggest-test-update algorithm, in our experience QUEST can provide a quick and quite reliable estimation of vibrotactile thresholds (Fig. 4). The advantages of such adaptive approaches are the flexibility to cover a wide range of participant sensitivities, the relatively low number of trials required, and the considerable advantage that all participants will find the task equally difficult—the algorithm finds the stimulus magnitude that leads, for example, to around 75–85% correct performance. The disadvantages with QUEST are that it can be very sensitive to errors made early in the sequence of trials, and performs poorly when reaching the floor or ceiling of the range of magnitudes being tested (see Note 4.6).

22

Nicholas Paul Holmes and Luigi Tame`

A good compromise between constant and adaptive approaches that we and many others have taken advantage of is to use an adaptive method to set the stimulus magnitude range for an individual participant, then to test the main experimental hypotheses using a constant stimulus set. For example, participants can be trained on a two-interval task using QUEST to find an approximate vibration detection threshold, then this threshold can be used to set the intensity or intensities of stimuli in the main experiment (e.g., ranging from 0.5 to 5 times the detection threshold, or at a constant of 2 times detection threshold). In tasks with discrete stimulus magnitudes such as the grating orientation discrimination task, QUEST could be used to find the narrowest-discriminable grating width, and the experiment could then be run with a range of gratings, for example from two gratings below threshold to five gratings above threshold. The advantages and disadvantages of this approach are the same as for the pure constant and pure adaptive methods. One additional disadvantage of this hybrid approach is that thresholds can change substantially over time. Whatever stimulus parameters are set during the training phase may no longer be appropriate during later experimental phases. In this case, short experiments and repeated training and calibration may be required. We would strongly recommend calibrating at the start of every session if participants return for multiple experiments. 3.2.5 Stimulus Waveform: Square-Wave, Sinusoidal, or Broadband?

In theory, any complex waveform can be presented to the skin as an amplitude- or frequency-modulated stimulus like a vibration. We have only used three types—square wave “taps”, sinusoidal “tones”, and random “white noise”. The stimulus presented will be determined largely by the specific research question, so we make only a few points here. The hardware, as well as the skin itself and the somatosensory receptors, will filter the stimulus waveform. Different skin compliance, thickness, and location will filter the stimulus differently. If the precise shape of the stimulus is important, additional sensitive measurement will be needed. Likewise, if the precise frequency of stimulation is important, more care will be needed than to present, for example, simple square-wave taps to the skin. When passed through a Fourier analysis, a square-wave signal can be described as comprising many different frequencies superimposed upon each other—a tap is a broadband stimulus. To present a cleaner signal at a particular frequency (e.g., 200 Hz, Fig. 2), we would aim for a sinusoidally varying stimulus that contains no stepped transients. The waveform should rise and fall smoothly (e.g., over 5–10 ms) and be composed of half or full wavelengths that taper to zero intensity at the start and end (e.g., a 200 Hz stimulus should be presented for durations that are multiples of 2.5 ms, a 20 Hz stimulus for multiples of 25 ms). Our toolbox provides some waveform-generation functions (https:// osf.io/8r2mx/.

Detection, Discrimination & Localization 3.2.6 Task Difficulty and Duration

23

The tactile perception tasks that we have studied can be difficult, attention-demanding, and tiring. Typically, only around threequarters of our participants are able or willing to complete the tasks. This can be frustrating for participants and experimenters. Some general advice for tactile detection and discrimination tasks is: • If possible, aim to study fewer participants for a greater amount of time—get better measurements from each participant rather than only trying to increase the sample size of your study. Statistical power depends on both sample size and effect size; increasing N will not help if the measurements become more variable (i.e., if the effect size decreases). • Include lots of time for training on the task and establishing thresholds—as much as half an hour might be needed, even for simple tasks. Expect long experiments with multiple sessions. • Check that participants really understand the task; it is surprising how long participants can continue to do a task even if they don’t feel anything or don’t understand what they are doing. Include attention-checks and monitor task performance frequently. • Always include some trials which are “easy” to remind participants of the target and task. • Psychophysical curves will need a wide range of difficulty, from extremely easy to impossible, to provide good estimates of the desired parameters. • Adaptive staircases, particularly in detection tasks, can be frustrating for participants who may feel almost nothing on their skin during the critical trials near their detection threshold or other limen. The best participants will spend more time feeling almost nothing! They may need encouragement to stay on task (see Note 4.7). • Overall, participants perform well, are well-motivated, and enjoy the tasks when they achieve 80–90% correct performance. Below 70% can feel to the participant like they are simply guessing, while over 90% may lead to ceiling effects—both are undesirable. • If the study requires lots of data or multiple repetitions of conditions over multiple sessions (we advise in general that it should), each individual session should include training, thresholding, counterbalancing, and whichever baseline or comparison conditions are required to answer all your scientific questions. If the data obtained from each session can stand alone, participants who are unable to return for later sessions will not need to be dropped from the study.

24

Nicholas Paul Holmes and Luigi Tame`

3.3 Experimental Procedures

Finally, we present guidance for the general procedures that we have followed in our work.

3.3.1 Vibrotactile Detection and Discrimination

The participant is seated when they arrive and remains seated throughout to prevent large changes in heart rate, unless that may help to wake them up between sessions. General instructions are given and consent taken. The task will be explained verbally, very carefully, using printed diagrams of the trial sequence, stimuli, and responses: “There will be two short intervals separated by flashes of light, here and here [pointing to two LEDs]. Between the flashes there will be a one second gap. In the middle of that gap, there may be a short vibration—‘bzzz’—or there may be nothing. Your task is to say whether there was a vibration or not [or: when the vibration was—first or second], by lifting the pedal under your left foot for ‘no’ and right foot for ‘yes’. Keep both feet pressed down at all times, except when making your response. If you made the correct response, there will be a short pause before the next trial. If you made a mistake, both lights will flash twice. OK?” Further advice could be given, such as: “Try not to think too much, just stay still, really pay attention to your finger, and make a guess about the vibration if you have to. Don’t worry if you make a mistake—in fact we want you to make a mistake about once every 5 or 6 vibrations. We are interested in how well you can perceive the vibrations, especially when it’s really difficult.” A typical experiment might then proceed as follows: • Position stimulus and check participant can feel it; secure the stimulus if needed. Avoid placing the stimulus so that the participant can change the force or pressure. Use a weighted see-saw or other device to minimize changes in force. • Short blocks of 4–10 practice trials with clearly supra-threshold stimuli until participants can repeatedly get >80% correct and clearly understand the task. • Detection thresholding task using QUEST ~2–3 repeats of 30–40 trial staircases. Repeat if necessary until good, stable performance. Start QUEST above their likely threshold. • Discrimination task using method of constant stimuli or QUEST. Usually a single condition per block, so run all conditions at least once. If repeating conditions, run in reverse order. Counterbalance wherever possible. • If additional baseline conditions or repeated thresholds are required, either intersperse them with the main conditions, or run them between repetitions.

Detection, Discrimination & Localization

25

• Keep individual sessions to around 40 min or up to an hour at most. About 6–8 blocks of around 5 min each, with 1 or 2 min break between is reasonable for one session. • Schedule additional sessions at least 20 min later. Repeat the practice, thresholding, and baseline conditions at least once for each day of testing. 3.3.2

Spatial Acuity

Procedures for the grating orientation task are similar. Instruct the participant: “On each trial you will feel a plastic grating stimulus press into your fingertip. Your task is to say whether the gratings run across or along your fingertip. If you feel them running horizontally, across your fingertip, lift the left pedal; if you feel them running vertically, along your fingertip, lift the right pedal. If you are unsure, just make your best guess—it will always be either across or along”. Additional instructions might be to keep as still as possible, not change posture, to keep the finger, hand, and arm aligned with the body, the eyes and the head—evidence from our lab shows that orientation perception is better when the hand is aligned with the head and body [21]. A typical experiment might run like this: • Training on the task, using the largest grating stimuli (e.g., 3 mm) until performance is close to perfect. • Find the grating orientation threshold using the full range of gratings (0.35–3 mm). With 12 gratings and 2 orientations, a full set requires 24 trials, but reliable thresholds are likely to require at least 8, and as many as possible, repetitions per condition. We recommend at least 2 blocks of 48 trials to find the approximate threshold. • For the main experiment (e.g., Fig. 5), either repeat the full range of stimuli for each experimental condition, or select a constant set of at least five gratings around the threshold level.

4

Notes

4.1 Tactile Hardware Is Not Uniform Over Time or Across Devices

When using the Oticon vibrators over many years and experiments, we have found that stimulus amplitude seems to differ greatly over time and between individual vibrators (although this may also just be the experimenter’s sensitivity!). We have four Oticons in the lab. Without a precise method of measuring and calibrating stimulus amplitude, the best we can do is number them 1 through 4, and use the same vibrator throughout an experiment, or at least throughout a session or for one participant. Alternatively, if it is necessary to use all tactors on the same participants for stimulating different body parts, varying their locations across participants can control for any confounding effects of specific stimulators. This advice is recommended also for all other electrical and mechanical devices.

26

Nicholas Paul Holmes and Luigi Tame`

Fig. 5 Grating orientation discrimination performance. Data show individual participants in gray, and the mean and 95% confidence intervals in black. In this experiment, 12 participants performed a block of 96 trials to estimate their grating orientation threshold before the main experiment (which involved brain stimulation). The 96 trials were equally divided and pseudorandomly presented across 12 grating widths from 0.35 to 3 mm—using the method of constant stimuli. Each grating could be presented “across” (horizontally) or “along” (vertically) their index fingertip, with four repetitions per grating width and orientation. The group data in black show a good, relatively smooth psychophysical curve, with performance being significantly better than chance with 1 mm gratings (as the confidence intervals do not include zero). The individual data, however, are very variable, suggesting that 8 repetitions per grating width are insufficient. Later experiments in this series (not shown) suggested that 6 blocks of 48 trials (i.e., 24 repetitions per grating, taking about 30 min per participant) produced reasonably good psychophysical curves for two individual participants

Again, without good measurements of between-session variability in stimulus amplitude or quality, it is best to collect all relevant data within a single session. If the experiment has a baseline condition, make sure it is counterbalanced with other conditions within each session, and use each session’s baseline data as a comparison or, for example, to calculate tactile thresholds in decibels [38, 55]. Since the vibrator output likely changes as it warms with extended skin contact, thresholds measured at the start of a session may not be the same as at the end—counterbalancing within each experimental session is recommended. Finally: always be aware that, without precise measures of stimulus amplitude and quality, detection and

Detection, Discrimination & Localization

27

discrimination thresholds are somewhat arbitrary, giving relative rather than absolute measures of sensitivity. 4.2 3D Printing Tactile Stimuli

For the grating discrimination task, we found that 3D printing of the smallest gratings (0.35 and 0.5 mm) was unsuccessful at first. This was due to the material used to print the stimuli. With such small gaps between the grating ridges, the printing material formed “bridges” between the gaps, degrading the grating waveform. Instead, we needed to select a higher-quality material to prevent these “bridges” forming. Because the smallest gratings are made of different materials, we should therefore avoid using these stimuli. In retrospect, we should have paid more money to print all gratings in the highest-quality material available. With 3D printing now more widely available at a relatively low cost, we can see great potential for sharing the designs and material for tactile stimuli. An open repository of tactile stimuli may improve reproducibility across laboratories.

4.3 ParticipantSpecific Environmental Considerations

Anecdotal evidence from trying to measure participants’ detection and discrimination thresholds over many years suggests that changes in temperature and humidity of the room can greatly impact participant performance. Changes in participants’ body temperature, blood pressure, and vasodilation likely also influence participants’ performance within an experimental session. More constant participant-specific issues include whether the participant smokes cigarettes (nicotine seems to damage finger sensitivity), plays a musical instrument (calluses), does lots of manual work (thicker skin, larger fingers), or has a peripheral condition such as Reynaud’s disease. Temperature and humidity are known to be important factors that can affect tactile performance. It may be important that these parameters are monitored in experiments. This is critical when, for example, a researcher wants to perform a behavioral tactile study that has previously been done in a laboratory room (about 24 °C) or within the fMRI scanner room (18 °C) [56]. In some cases, the participant simply won’t feel the tactile stimulus, and not much can be done to change that—thank them for their time and apologize that the equipment was not sensitive enough for them. In other cases, such as when a participant arrives into a warm laboratory from a wintry English day, or into a cold airconditioned laboratory from a hot Italian summer’s day, their sensitivity and ability to do the task may change a great deal during the session.

28

Nicholas Paul Holmes and Luigi Tame`

4.4 Task Comprehension and Training

By far the most important factor in tactile threshold experiments, and the one that has costs us the most in lost time and lost data, is the participant’s comprehension of the task. The experimenter must be empathetic, patient, and very careful to explain exactly what the participant is asked to do. Very likely the participant will not have done this kind of task before, and it is not always intuitive in the way that many psychological experiments are. The experimenter should watch very carefully whatever the participant says or does, correcting and prompting where necessary and praising good performance. If there is any doubt in the participant’s mind at all, the experimenter should take time to reassure them, and to correct their understanding where possible. For about 10% of participants we have found it impossible to get the task to work—either the participant just can’t feel the stimulus, or just cannot comprehend the task. It is genuinely perplexing! In one memorable case, the first author was testing a participant who was performing at about 30% correct in a two-interval task, when chance performance was 50%— the participant was consistently choosing the wrong stimulus. First, the experimenter checked that the stimulators were plugged in the right way, and that the response pedals were correctly positioned. All was OK. The participant explained their understanding of the task and it turned out they had learned the incorrect stimulusresponse mapping during practice. NH said: “Oh, don’t worry, I can see that you are doing the task correctly, so all you need to do is swap the responses over”. Despite three or four attempts, we were unable to get the participant to learn the opposite stimulusresponse mapping, and performance continued at 20–30% correct! We thanked the participant and resolved to write this anecdote into a methods chapter. When the task is difficult, for example in thresholding experiments, participants need to be encouraged to guess, and to be comfortable in going with their “gut feelings”. Some participants seem to have real difficulty making these sorts of guesses, and find it quite strange that you are asking them to say, for example, which finger was a vibration on when they didn’t feel anything. They should be encouraged to guess with no particular pattern, just to choose whichever response they think it might be. These tasks do not produce good data if participants use strategies, have response biases, or try to remember which responses they’ve given on previous trials. Participants should be encouraged not to over-think it, to free their mind, to treat each stimulus as independent, and to guess as best as they can, but not to worry about mistakes. Finally, when designing the experiment, the researcher should try as hard as possible to “break” their experiment—what would happen if the participant tried to cheat, or always gave the same response, or misinterpreted the instructions? Imagine being the

Detection, Discrimination & Localization

29

worst possible participant, and make sure you know how that behavior would look in your experiment. Change your experiment to make sure cheating is not possible, or if it is, that it results in chance performance, or otherwise impossible performance that can safely be excluded from the dataset post-hoc. 4.5 Frequency and Intensity Interact

Just like in the auditory system, there are non-linear interactions between the perception of frequency and amplitude of vibrations in touch (e.g., refs. 4, 57). The perceived intensity of a stimulus depends on the frequency. At our peak sensitivity of around 200– 250 Hz (Fig. 2), a supra-threshold stimulus will feel strong. Maintaining the amplitude, but increasing or decreasing the frequency away from 250 Hz will result in two changes—in perceived frequency but also a decrease in perceived intensity. We made the mistake of running a frequency discrimination experiment where the standard stimulus was 200 Hz, and each comparison stimulus was of a higher frequency, set using QUEST. We tested the experiment on ourselves and it worked well—our frequency discrimination thresholds were around 40–60 Hz, so during experimental blocks we were often comparing, for example, a 200 Hz with a 250 Hz stimulus. This task also worked well for well-trained participants in our series of TMS studies [35] (Chapter 20, this volume). When we started testing typical, non-expert participants, however, their discrimination thresholds began at much higher levels of around 100–200 Hz, so they were regularly comparing 200 and 300 Hz or 200 and 400 Hz stimuli. Their task was to say which stimulus was of “higher” frequency. But since stimuli over 300 Hz also feel less intense than 200 Hz stimuli, several participants were unable to separate out the changes in frequency and intensity. For naı¨ve participants, this was a poor choice of task. The lesson is to pilot-test on both expert and naı¨ve participants.

4.6 Adaptive Staircases and Experimenter Expertise

For an experienced experimenter who can adjust the QUEST algorithm [53], and carefully instruct the participant during an experiment, QUEST can provide a powerful psychophysical tool that can be adapted to any experimental question. But without care and expertise, we have found that these adaptive approaches can easily fail. Imprecise instructions from the experimenter, one or two early mistakes by the participant, or a range of stimulus magnitudes that is either too small or too large can result in a confusing or frustrating task for the participant. Adaptive designs require a hands-on and patient experimenter who is willing to re-start a block of trials, re-instruct the participant, or abort the experiment if it can’t be adapted to the participant. We would not recommend them for short duration or single-experiment student projects.

30

4.7

Nicholas Paul Holmes and Luigi Tame`

Attention

Detecting a near-threshold somatosensory stimulus requires an unusual state of perception and attention (e.g., ref. 58) that is very hard to describe to a participant, so the experimenter should take time to learn how best to explain it. It can also be very tiring for the participant. The first author finds, as a participant, that entering a sort of meditative state helps: visually fixate a speck of dirt or small feature on the table, don’t look directly at the site of tactile stimulation, but do focus all your attention on that one body part. Try to keep breathing and other movements regular. While the stimulus is easy to detect, try to get into a rhythm of cue-target-responsewait. . . cue-target-response-wait. . . Try not to think about much at all. As the stimulus intensity decreases down the staircase, what started out feeling like a long, continuous vibration can begin to feel like a short or discontinuous flutter, or perhaps only a single touch, as if a tiny insect moved a single leg across a ridge in your fingerprint. In this intense meditative state, you will become aware of vibrations traveling across the floor and the furniture in the testing room. Your heartbeat and its pulse in your finger will soon become the strongest stimulus you are aware of; it will often coincide with the target and you will feel nothing but your own pulse. Your vision may narrow down to a tunnel around your fixation point and you may hear the blood rushing through your ears. Life near the limen can be strange.

References 1. Weber EH (1978) E.H. Weber on the tactile senses. Academic Press 2. Fechner GT (1860) Elemente der psychophysik [Elements of psychophysics]. Breitkopf und H€artel, Leipzig 3. Bradman MJG, Ferrini F, Salio C, Merighi A (2015) Practical mechanical threshold estimation in rodents using von frey hairs/semmesweinstein monofilaments: towards a rational method. J Neurosci Methods 255:92–103. https://doi.org/10.1016/j.jneumeth.2015. 08.010 4. von Be´ke´sy G (1967) Sensory inhibition. PhD thesis, Princeton University, Princeton, NJ 5. Sadibolova R, Tame` L, Walsh E, Longo MR (2018) Mind the gap: the effects of temporal and spatial separation in localization of dual touches on the hand. Front Hum Neurosci 12:55. https://doi.org/10.3389/fnhum. 2018.00055 6. Hidaka S, Tame` L, Longo MR (2021) Tactile interactions in the path of tactile apparent motion. Cognition 209:104569. https://doi. org/10.1016/j.cognition.2020.104569

7. Hayward V (2008) A brief taxonomy of tactile illusions and demonstrations that can be done in a hardware store. Brain Res Bull 75(6): 742–752. https://doi.org/10.1016/j. brainresbull.2008.01.008 8. Lederman SJ, Jones LA (2011) Tactile and haptic illusions. IEEE Trans Haptics 4(4): 273–294. https://doi.org/10.1109/ToH. 2011.2 9. Geldard FA (1953) The human senses. Wiley, New York, pp 11–584 10. Bolanowski SJJ, Gescheider GA, Verrillo RT, Checkosky CM (1988) Four channels mediate the mechanical aspects of touch. J Acoust Soc Am 84(5):1680–1694. https://doi.org/10. 1121/1.397184 11. Craig JC (1985a) Tactile pattern perception and its perturbations. J Acoust Soc Am 77(1): 238–246. https://doi.org/10.1121/1. 392265 12. Verrillo RT (1985) Psychophysics of vibrotactile stimulation. J Acoust Soc Am 77(1): 225–232

Detection, Discrimination & Localization 13. Saal HP, Bensmaia SJ (2014) Touch is a team effort: interplay of submodalities in cutaneous sensibility. Trends Cogn Sci 37(12):689–697. https://doi.org/10.1016/j.tins.2014.08.012 14. Boring EG (1942) Sensation and perception in the history of experimental psychology. Appleton-Century, New York, pp 15–644 15. Craig JC, Johnson KO (2000) The two-point threshold: not a measure of tactile spatial resolution. Curr Dir Psychol Sci 9(1):29–32. https://doi.org/10.1111/1467-8721.00054 16. Johnson KO, Phillips JR (1981) Tactile spatial resolution. I. Two-point discrimination, gap detection, grating recognition, and letter recognition. J Neurophysiol 46(6):1177–1191. https://doi.org/10.1152/jn.1981.46.6.1177 17. Phillips JR, Johnson KO (1981b) Tactile spatial resolution. II. Neural representation of bars, edges, and gratings in monkey primary afferents. J Neurophysiol 46(6):1192–1203. https://doi.org/10.1152/jn.1981.46.6.1192 18. Onishi H, Nagasaka K, Yokota H, Kojima S, Ohno K, Sakurai N, Kodama N, Sato D, Otsuru N (2022) Association between somatosensory sensitivity and regional gray matter volume in healthy young volunteers: a voxel-based morphometry study. Cereb Cortex:bhac188. https://doi.org/10.1093/cercor/bhac188 19. Tong J, Mao O, Goldreich D (2013) Two-point orientation discrimination versus the traditional two-point test for tactile spatial acuity assessment. Front Hum Neurosci 7:579. https://doi.org/10.3389/fnhum.2013. 00579 20. Tremblay F, Backman A, Cuenco A, Vant K, Wassef AM (2000) Assessment of spatial acuity at the fingertip with grating (JVP) domes: validity for use in an elderly population. Somatosens Mot Res 17(1):61–66. https://doi.org/ 10.1080/08990220070300 21. French B, Di Chiaro NV, Holmes NP (2022) Hand posture, but not vision of the hand, affects tactile spatial resolution in the grating orientation discrimination task. Exp Brain Res 240:2715–2723. https://doi.org/10.1007/ s00221-022-06450-3 22. Tame` L, Braun C, Lingnau A, Schwarzbach JV, Demarchi G, Hegner YL, Farne` A, Pavani F (2012) The contribution of primary and secondary somatosensory cortices to the representation of body parts and body sides: an fMRI adaptation study. J Cogn Neurosci 24(12): 2306–2320. https://doi.org/10.1162/jocn_ a_00272 23. Tame` L, Pavani F, Papadelis C, Farne` A, Braun C (2015) Early integration of bilateral touch in the primary somatosensory cortex. Hum Brain

31

Mapp 36(4):1506–1523. https://doi.org/10. 1002/hbm.22719 24. Bro¨nnimann B, Meier ML, Hou M, Parkinson C, Ettlin DA (2016) Novel air stimulation MR-device for intraoral quantitative sensory cold testing. Front Hum Neurosci 10: 335. https://doi.org/10.3389/fnhum.2016. 00335 25. Servos P, Zacks J, Rumelhart DE, Glover GH (1998) Somatotopy of the human arm using fMRI. Neuroreport 9(4):605–609. https:// doi.org/10.1097/00001756199803090-00008 26. Huang R, Chen C, Tran AT (2012) Mapping multisensory parietal face and body areas in humans. Proc Natl Acad Sci U S A 109(44): 18114–18119. https://doi.org/10.1073/ pnas.1207946109 27. Chen C, Kreutz-Delgado K, Sereno MI, Huang R (2017) Validation of periodic fMRI signals in response to wearable tactile stimulation. NeuroImage 150:99–111. https://doi. org/10.1016/j.neuroimage.2017.02.024 28. Huang R, Sereno MI (2007) Dodecapus: an MR-compatible system for somatosensory stimulation. NeuroImage 34(3):1060–1073. https://doi.org/10.1016/j.neuroimage. 2006.10.024 29. Tame` L, Tucciarelli R, Sadibolova R, Sereno MI, Longo MR (2021) Reconstructing neural representations of tactile space. NeuroImage 229:117730. https://doi.org/10.1016/j. neuroimage.2021.117730 30. Broser PJ, Braun C (2012) Hydraulic driven fast and precise nonmagnetic tactile stimulator for neurophysiological and MEG measurements. IEEE Trans Biomed Eng 59(10): 2852–2858. https://doi.org/10.1109/ TBME.2012.2212191 31. Valentini M, Kischka U, Halligan PW (2008) Residual haptic sensation following stroke using ipsilateral stimulation. J Neurol Neurosurg Psychiatry 79(3):266–270. https://doi. org/10.1136/jnnp.2007.120279 32. Gibson GO, Craig JC (2006) The effect of force and conformance on tactile intensive and spatial sensitivity. Exp Brain Res 170(2): 172–181. https://doi.org/10.1007/s00221005-0200-1 33. Goldreich D, Wong M, Peters RM, Kanics IM (2009) A tactile automated passive-finger stimulator (TAPS). J Vis Exp 28:e1374. https:// doi.org/10.3791/1374 34. Liao X, Li Y, Hu J, Ding X, Zhang X, Ying B, Takatera M, Sukigara S, Pan F, Sun Y, Wang Y, Feng W, Zhu C, Li Q, Wu X (2018) Effects of contact method and acclimation on

32

Nicholas Paul Holmes and Luigi Tame`

temperature and humidity in touch perception. Can J 14:1605–1615. https://doi.org/10. 1177/0040517517705628 35. Tame` L, Holmes NP (2016) Involvement of human primary somatosensory cortex in vibrotactile detection depends on task demand. NeuroImage 138:184–196. https://doi.org/ 10.1016/j.neuroimage.2016.05.056 36. Tame` L, Braun C, Holmes NP, Farne` A, Pavani F (2016) Bilateral representations of touch in the primary somatosensory cortex. Cogn Neuropsychol 33(1–2):48–66. https://doi.org/ 10.1080/02643294.2016.1159547 37. Vardar Y, Gu¨c¸lu¨ B, Basdogan C (2018) Tactile masking by electrovibration. IEEE Trans Haptics 11(4):623–635. https://doi.org/10. 1109/TOH.2018.2855124 38. Tame` L, Moles AP, Holmes NP (2014) Within, but not between hands interactions in vibrotactile detection thresholds reflect somatosensory receptive field organization. Front Psychol 5:174. https://doi.org/10.3389/fpsyg.2014. 00174 39. Medina S, Tame` L, Longo MR (2018) Tactile localization biases are modulated by gaze direction. Exp Brain Res 236(1):31–42. https:// doi.org/10.1007/s00221-017-5105-2 40. Gescheider GA, Bolanowski SJJ, Hall KRL, Hoffmann KE, Verrillo RT (1994) The effects of aging on information-processing channels in the sense of touch: I. Absolute sensitivity. Somatosens Mot Res 11(4):345–357. h t t p s : // d o i . o r g / 1 0 . 3 1 0 9 / 08990229409028878 41. Harris JA, Thein T, Clifford CWG (2004) Dissociating detection from localization of tactile stimuli. J Neurosci 24(14):3683–3693. https://doi.org/10.1523/JNEUROSCI. 0134-04.2004 42. Schmidt TT, Blankenburg F (2018) Brain regions that retain the spatial layout of tactile stimuli during working memory – a “tactospatial sketchpad”? NeuroImage 178:531–539. https://doi.org/10.1016/j.neuroimage. 2018.05.076 43. Tame` L, Bumpus N, Linkenauger SA, Longo MR (2017) Distorted body representations are robust to differences in experimental instructions. Atten Percept Psychophys 79(4): 1204–1216. https://doi.org/10.3758/ s13414-017-1301-1 44. Tame` L, Linkenauger SA, Longo MR (2018) Dissociation of feeling and belief in the rubber hand illusion. PLoS One 13(10):e206367. https://doi.org/10.1371/journal.pone. 0206367

˜ o´n E (2014) Using time to inves45. Heed T, Azan tigate space: a review of tactile temporal order judgments as a window onto spatial processing in touch. Front Psychol 5:76. https://doi.org/ 10.3389/fpsyg.2014.00076 46. Constable MD, Welsh TN, Huffman G, Pratt J (2019) I before U: temporal order judgements reveal bias for self-owned objects. Q J Exp Psychol 72(3):589–598. https://doi.org/10. 1177/1747021818762010 47. Baumgarten TJ, Schnitzler A, Lange J (2016b) Prestimulus alpha power influences tactile temporal perceptual discrimination and confidence in decisions. Cereb Cortex 26(3):891–903. https://doi.org/10.1093/cercor/bhu247 48. Grund M, Forschack N, Nierhaus T, Villringer A (2021) Neural correlates of conscious tactile perception: an analysis of bold activation patterns and graph metrics. NeuroImage 224: 1 1 7 3 8 4 . h t t p s :// d o i . o r g / 1 0 . 1 0 1 6 / j . neuroimage.2020.117384 49. Hurme M, Railo H (2022) Promise and challenges for discovering transcranial magnetic stimulation induced “numbsense”—commentary on Ro & Koenig (2021). Conscious Cogn 98:103265. https://doi.org/10.1016/j.con cog.2021.103265 50. Bruyer R, Brysbaert M (2011) Combining speed and accuracy in cognitive psychology: is the inverse efficiency score (IES) a better dependent variable than the mean reaction time (RT) and the percentage of errors (PE)? Psychol Belg 51(1):5–13 51. Orne MT (1962) On the social psychology of the psychological experiment, with particular reference to demand characteristics and their implications. Am Psychol 17(11):776–783. https://doi.org/10.1037/h0043424 52. Croy I, Bierling A, Sailer U, Ackerley R (2021) Individual variability of pleasantness ratings to stroking touch over different velocities. Neuroscience 464:33–43. https://doi.org/10. 1016/j.neuroscience.2020.03.030 53. Watson AB, Pelli DG (1983) QUEST: a Bayesian adaptive psychometric method. Percept Psychophys 33(2):113–120. https://doi.org/ 10.3758/BF03202828 54. Brainard DH (1997) The psychophysics toolbox. Spat Vis 10(4):433–436. https://doi. org/10.1163/156856897X00357 55. D’Amour SAO, Harris LR (2014) Contralateral tactile masking between forearms. Exp Brain Res 232(3):821–826. https://doi.org/ 10.1007/s00221-013-3791-y 56. Rusconi E, Tame` L, Furlan M, Haggard P, Demarchi G, Adriani M, Ferrari P, Braun C, Schwarzbach JV (2014) Neural correlates of

Detection, Discrimination & Localization finger gnosis. J Neurosci 34(27):9012–9023. https://doi.org/10.1523/JNEUROSCI. 3119-13.2014 57. LaMotte RH, Mountcastle VB (1975) Capacities of humans and monkeys to discriminate between vibratory stimuli of different frequency and amplitude: a correlation between

33

neural events and psychophysical measurements. J Neurophysiol 38:539–559 58. Whang KC, Burton H, Shulman GL (1991) Selective attention in vibrotactile tasks: detecting the presence and absence of amplitude change. Percept Psychophys 50(2):157–165. https://doi.org/10.3758/bf03212216

Chapter 2 Methods of Somatosensory Attenuation Konstantina Kilteni Abstract Self-generated touch feels less intense (and less ticklish) than an externally generated touch of the same intensity. Since the early 1970s, researchers in somatosensation sought to understand the origins and the principles of this phenomenon. This chapter will focus on two methods of studying the perceived intensity of self-generated touch in relation to externally generated touch: the force-matching task and the forcediscrimination task. The relative merits of each method as well as their associated practical difficulties will be discussed. Key words Somatosensory attenuation, Force perception, Intensity, Ticklishness, Force-matching, Force-discrimination, Psychophysics, Internal forward models, Prediction, Behavior

1

Introduction A mechanoreceptor in our skin cannot distinguish if its input results from our self-generated movement or from the environment. We can touch our foot with our hand and activate the receptor to the same extent as a dangerous spider would when crawling into our shoe, to pick a dramatic example. Still, the nervous system shows a remarkable capacity in differentiating between these two sources of information. An efficient way to perform this distinction is to suppress the perception of the self-generated touch on the foot caused by our hand movement, with respect to that of the touch caused by the spider. By doing so, the salience of the externally generated touch, that is the touch from the spider, is increased. To achieve this, the nervous system uses information about our hand movement, to compute when and where to expect the touch on the foot [1–5]. The term somatosensory attenuation refers to the phenomenon that touches produced by our voluntary movements feel less intense compared to touches of the exact same intensity, at the exact same body location, but generated by another person or machine, or a spider.

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_2, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

35

36

Konstantina Kilteni

The first experimental study on somatosensory attenuation was triggered by the question “why can’t we tickle ourselves?” and was conducted in the early 1970s [6]. Researchers constructed a tickling device on top of which the participants were asked to place the sole of their bare foot. Participants moved a handle on the device to apply strokes along their foot soles (self-generated touch condition). In another condition, the experimenter moved the handle to stroke the participants’ foot soles. Despite the touches in the two conditions being matched in terms of pressure and frequency, all participants (10/10) reported that the strokes controlled by the experimenter felt more ticklish than the self-generated strokes. This result led the researchers to conclude that the efferent signals corresponding to the hand movement contributed to the cancellation of the tickling sensation on the foot. Studies on tickling continued in the following years replicating the finding that self-generated touches feel less ticklish than externally generated ones, also in the hand and arm area [3, 7]. In their experiment, Blakemore and colleagues [3] used a setup that consisted of two robots, one for each of the participants’ hands. Participants held one robot with their left hands and performed sinusoidal movements. A second robot imitated the trajectory of the first robot that was guided by the participants’ left hands, and applied the sinusoidal stroke to the palm of their right hands. This setup dissociated the movement of the left hand from the resulting touch on the right palm. Moreover, the distance between the two robots was minimal, creating the illusion of touching one hand with the other through a rigid object. By introducing perturbations between the movement of the participants’ left hands and the resulting touch on their right palms the authors showed that the perceived ticklishness of a self-generated touch increases, and reaches that of an external touch when a large delay (e.g., 300 ms) or a spatial deviation (e.g., shifting the phase by 90°) is used. These results suggest that an increase in the discrepancy between what sensations are expected on the basis of the movement, and the actual sensations, also increases the perceived ticklishness. The authors concluded that the ticklishness of self-generated touches is attenuated as long as the predicted somatosensory consequences of our movement match the somatosensory feedback. Since then, research on somatosensory attenuation has largely abandoned the study of ticklishness, probably due to difficulties in defining and quantifying ticklishness. Instead, efforts now focus on the attenuation of the perceived intensity of self-generated touches, which was significantly advanced with two psychophysics tasks introduced in the early 2000s—the force-matching task [5] and the force-discrimination task [8].

Somatosensory Attenuation

1.1 Behavioral Methods of Somatosensory Attenuation

37

The force-matching task was first presented in 2003 by Shergill and colleagues [5] as a psychophysical method to quantify the attenuation of the perceived intensity of self-generated touches. Before its introduction, studies on somatosensory attenuation, particularly those testing tickling perception [3, 7, 9], used subjective reports: participants were asked to rate the perceived intensity or ticklishness of a stimulus applied to their body on a rating scale, where the lower limit of the scale corresponded to “Not (intense/ticklish) at all” and the higher limit corresponded to “Extremely (intense/ ticklish)” or “Very (intense/ticklish)”. In the force-matching task, participants receive an externally generated force on the pulp of their left index finger. Immediately afterward, they are asked to match it with a self-generated force, by pressing their right index finger against their left index finger. The method offers a quantitative measure of somatosensory attenuation across different levels of stimulus intensity, while at the same time avoiding the drawbacks of self-reports, such as task compliance effects. Since 2003, the task has been repeatedly used by many researchers in the field [1, 4, 10–18]. The force-discrimination paradigm was introduced in 2005 by Bays and colleagues [8] as an alternative method to study somatosensory attenuation of self-generated taps. In this task, participants receive two consecutive taps on their left index finger. One tap is triggered by a voluntary movement of the right index finger, thus self-generated, while another tap is triggered in the absence, or independently, of movement, and is therefore considered to be of external origin. Participants are then asked to discriminate which tap, the self-generated or the external one, felt stronger. In relation to the force-matching task, the short duration of the stimuli (taps rather than forces) has allowed experimenters to introduce delays between the participants’ movements and the received taps, and thus study the temporal dependencies of the phenomenon. Similarly to the force-matching task, the force-discrimination task has also been used in several studies [19–21]. Somatosensory attenuation has also been investigated with complementary methods; for example, electroencephalography (EEG) [12], magnetoencephalography (MEG) [22], structural and functional magnetic resonance imaging (sMRI, fMRI) [14, 16, 23–28], and positron emission tomography (PET) [29]. See Chapters 17, 18, and 19, this volume, for more on these methods. Using fMRI, it has repeatedly been observed that selfgenerated touch yields reduced responses in the secondary somatosensory cortex and the cerebellum compared to external touch [14, 23, 24]. This activity pattern is in agreement with earlier behavioral findings showing, for example, that self-generated touch feels systematically weaker than externally generated touch [3, 5, 8, 21]. Further insights were created when combining the force-matching task with fMRI. A recent study showed that

38

Konstantina Kilteni

stronger attenuation of self-generated forces in the force-matching task was associated with stronger functional connectivity between the cerebellum and the somatosensory cortex during self-generated touches compared to externally generated touches [14]. Although the scope of this chapter is not to provide an exhaustive review of experimental findings on somatosensory attenuation, these two psychophysics tasks have generated a number of fascinating key findings, five of which will be outlined below. First, somatosensory attenuation is a robust perceptual phenomenon. Using the force-matching task, attenuation was detected in a staggering 98%, or 315 out of 322 people, across an age range of 18–88 years [16]. Second, the phenomenon relies on predictive processes—on predicting the touches that will result from the movement—rather than on postdictive processes—on other sensory or motor information that is available at or after receiving the touch. Specifically, using the force-discrimination task, it was shown that participants attenuated a tap on their left index finger when they expected a contact with their right index finger, even if at the end, the fingers unexpectedly did not make contact [21]. Third, a touch that is highly predictable in terms of time and space but does not occur through a voluntary movement does not get attenuated. Using the force-discrimination task, it was shown that participants do not attenuate a tap on their left index finger when this is produced by the passive (involuntary) movement of the right index finger [19], or when it is simultaneously presented with a tap on the right index finger [8]. These results suggest that motor prediction is necessary for somatosensory attenuation. (See also refs. 30, 31 for similar conclusions in anticipatory postural adjustments and auditory attenuation respectively.) Fourth, a self-generated touch is attenuated when it matches the somatosensory prediction in terms of space. Using the forcematching task, Bays and Wolpert [1] showed that when a horizontal distance was introduced between the two hands, participants matched the external forces more accurately; that is, they showed less attenuation of their self-produced forces. This finding was replicated a few years later [4]. Why would the distance between the hands matter? A distance between the two hands makes it less probable that the experienced touch on the left index finger results from the movement of the right hand. In contrast, when there is no distance between the hands, a direct contact of the two fingers is very likely. More recent studies showed that the decrease in attenuation due to the hands’ distance can be overcome when holding a tool on the right hand to apply touch on the left index finger [4], or when experiencing the illusion that the right hand is located closer to the left index finger than it physically is [13]. Fifth, a self-generated touch is attenuated when it matches the somatosensory prediction in terms of time. Using the force-

Somatosensory Attenuation

39

discrimination task, two studies showed that participants perceived a delayed self-generated touch to be more intense than a non-delayed self-generated touch [8, 20]. Interestingly, after exposing participants to a fixed delay of 100 ms between the movement of the right index finger and the resulting tap on the left index finger, the non-delayed touch felt stronger (its attenuation decreased) and the delayed touch felt weaker (its attenuation increased) [20]. This suggests that the tactile expectations we have about our movements can adjust to changes in the way our body interacts with the environment. Before discussing the methodological details of the forcematching and force-discrimination tasks (Sect. 3), we first discuss the proposed mechanism of the phenomenon, as well as its wider implications across different populations and species. 1.2 Insights from Computational Motor Control

Computational accounts for sensorimotor control proposed that the key mechanism for the attenuation of self-generated touches is sensorimotor prediction [32]. When we move, our brain predicts the sensory consequences of our movement. For example, when we reach to grasp a cup, our brain predicts the visual input (we should see our hand closer to the cup with our fingers around it), the proprioceptive input (we should feel certain proprioceptive input from our new postural configuration) and the somatosensory input (we should receive touch on our fingers from grasping the cup). This prediction is generated by an internal model called a forward model that uses a copy of the motor command (an efference copy) to predict how the state of our body (e.g., the position of the hand or the angles of the joints) will change after the movement [32–36] (Fig. 1). Once the sensory feedback becomes available, the brain integrates it with the prediction of the forward model to estimate the new state of the body. This new estimated state is based both on sensory prediction and sensory input, and it is a more accurate estimation than if based on sensory input alone [32, 33, 35], which is typically noisy and subject to delays in transduction and transmission [37]. Let’s now apply this framework to reaching to touch and tickle our foot. Our brain can use the efference copy to predict the somatosensory consequences of our hand movement on our foot [34, 36, 38]. The predicted sensory consequences can then be “removed” from the actual somatosensory consequences on our foot once they are received. This leads to the attenuation (or cancellation) of self-generated touches, and makes the system sensitive in detecting touches that are not self-generated but caused by the environment; for example, the touch caused by the spider crawling on our foot [1, 2, 32, 39]. As discrepancies are introduced between the predicted and the received somatosensory consequences, the attenuation is reduced, the perceived intensity and

40

Konstantina Kilteni

Fig. 1 Somatosensory attenuation and internal forward models. As we move our hand to touch our foot, the forward model uses the copy of the motor command (efference copy) and the currently estimated state of our body (e.g., the position of the hand and the foot) to predict the future sensations on the foot (and the hand). Once the movement is executed, the predicted sensory consequences (including visual, proprioceptive, and somatosensory) are combined with the received sensory input to estimate the new state of the body parts. Selfgenerated touch is attenuated through this somatosensory prediction signal that is available during voluntary movement. See also refs. 1, 3, 35, 39

ticklishness increase [1, 3, 8, 20], and the sensation is attributed to external causes [40–42]. Here one should mention that, in practice, there is no complete cancellation of the perceived intensity of self-generated touches. If this were the case, then we would not perceive anything at all every time we touch our body. In reality, we do perceive our selfgenerated touches, but at a lower intensity than the one they have. Therefore, although the term “cancellation” is used in the literature, and this might indeed apply for the perception of decreased ticklishness of self-generated touches, the word attenuation is more accurate when we refer to the perception of intensity. 1.3 Clinical Applications

Since somatosensory attenuation is considered to result from the integration of somatosensory signals and predictions about the somatosensory consequences of the movement (Fig. 1), researchers have used somatosensory attenuation paradigms to assess whether this integration differs in clinical populations compared to healthy ones. In a force-matching task, patients with functional movement disorders were shown not to distinguish between self-generated and externally generated forces and thus, they attenuated their self-generated forces less than healthy controls [43]. Using the same task, patients on medication with Parkinson’s disease were shown to exhibit levels of attenuation comparable with healthy controls. Nevertheless, across the population of patients, the degree of somatosensory attenuation was negatively related to disease severity, and positively related to dopamine dosage, suggesting that dopamine might increase the weight of the sensorimotor predictions with respect to the sensory signals [15].

Somatosensory Attenuation

41

Somatosensory attenuation has been extensively discussed in relation to schizophrenia. Using the force-matching task, Shergill and colleagues [44] showed that schizophrenic patients attenuated their self-generated touches to a lesser extent than healthy matched controls. These findings were corroborated in a subsequent study showing that the difference in the responses of the secondary somatosensory cortex during self-generated touch compared to external touch was smaller in schizophrenic patients than in healthy controls. Moreover, the extent of somatosensory attenuation in schizophrenic patients was negatively related to the severity of their hallucinations [45]. In further support, healthy individuals with higher levels of delusional ideation showed less attenuation of their self-produced forces in the force-matching task [18]. These results from the force-matching task agree with results from studies using self-reports; psychiatric patients with auditory hallucinations and/or passivity experiences perceived their self-generated touches to be as ticklish and intense as external touches [46], while healthy subjects with high schizotypal traits perceived their self-generated touches as more ticklish than subjects with low schizotypal traits [47]. Together, these results enforced the view that many of the symptoms that patients with schizophrenia have (e.g., delusions of control) might be due to deficits in their predictive mechanisms [48–51]. Finally, the force-matching task was also used to show that somatosensory attenuation increases as a function of age, possibly as a strategy of the organism to compensate for age-related reductions in sensory precision. Specifically, it was observed that older participants (65+ years) attenuated their self-generated forces to a greater extent than younger adults (aged 18–39 years) [16]. Moreover, by combining these findings with MRI, the authors further showed that this increase in somatosensory attenuation was related to lower gray matter volume in motor-related brain areas (i.e., the pre-supplementary motor area). 1.4 Beyond Human Somatosensation

It is fundamental for our survival to efficiently identify the source of our sensory input (self or environment) in order to trigger the appropriate actions. It is therefore not surprising that strategies similar to that of somatosensory attenuation can be found in other species besides humans [39, 52–55]. For example, in crickets, an auditory interneuron that bursts in phase with the animal’s wing movement was detected to inhibit (both pre-synaptically and postsynaptically) central auditory processing. Through this inhibition, crickets prevent the desensitization of their auditory system during self-chirping, and maintain its sensitivity to sounds caused by environmental events, such as their predators [56, 57]. A similar strategy was shown in mice. The animals were trained to associate specific tone frequencies to their locomotion. After learning these associations, auditory cortical responses to self-generated sounds

42

Konstantina Kilteni

were suppressed. Importantly, this cancellation was specific to the frequency of the tones that the animals predicted through their movements, and not present for externally generated sounds (e.g., when the animals were at rest) [58]. Another example is the weakly electric fish that communicates and senses its surrounding environment through electrical signals [59]. In order to differentiate between self-generated and externally generated electrical signals, a negative image of the predicted electrosensory consequences is generated and added to the signal at the first stage of central processing, allowing the animal to detect and respond only to externally generated discharges [53, 60]. In agreement, experiments in primates showed that the responses of neurons in the vestibular nucleus are canceled during active head movements (i.e., for self-generated vestibular information) compared to passive head movements. This distinction helps the animals to maintain their head and body posture and ensures that vestibulocollic and vestibulospinal reflexes are elicited only when appropriate [52, 53, 61–63]. Taken together, despite differences across species and sensory systems, attenuation of self-generated sensory signals seems to constitute a generic strategy of biological organisms to differentiate the self from the others using motor signals.

2

Materials The equipment used for the force-matching task and the forcediscrimination task is relatively simple. Although some setups use commercially available haptic robotic devices [12, 43], most studies use custom-made hardware. This includes an electric motor with its controller, a lever with a probe that is attached to the motor, a force sensor that is attached to the lever which serves to measure the forces applied by the motor to the passive index finger, and a slide potentiometer or a joystick for the control condition (Fig. 2). A second force sensor (identical to the first one) is used to record the forces applied by the active index finger. In case the participants give non-verbal responses in the force-discrimination task, a foot pedal or response button can be used. Custom-made software is then used to control the motor output; for example, through implementing a proportional– integral–derivative (PID) controller. The forces applied by the active index finger on the force sensor or the displacement of the potentiometer wiper can be used to adjust the force output of the motor on the participants’ left index finger.

Somatosensory Attenuation

43

Fig. 2 Basic setup for force-matching and force-discrimination task. Participants receive forces through a probe that is attached to a lever (L) that is controlled by a DC motor (M ). The forces applied to the pulp of the participants’ left index fingers are measured by the probe sensor (F1). In the force-matching task, the participant can reproduce the force that they had just felt on the left index finger by using the right index finger to press a second force sensor (F2) that controls the force output of the lever (F1) on the left index finger. Alternatively, they can directly press their right index finger against the probe force sensor (F1). In the forcediscrimination task, the participant can trigger one of the two taps experienced on the left index finger and measured by the probe sensor (F1) by tapping with their right index finger the second force sensor (F2)

3

Methods

3.1 Force-Matching Task

In a typical trial of the force-matching task, the participant receives a constant externally generated force (presented force) on the pulp of their left index finger, by a probe controlled by a motor (Fig. 2) (see Note 4.1). The probe can already be in contact with the finger, increasing the applied force from a baseline level. A sensor placed inside the probe measures the applied force. Immediately after the presentation of the external force, the participants are asked to produce a self-generated force (matched force) that matches the intensity of the presented force. The presented force typically lasts 3 s [1, 4, 14], and the participants are given 3 s (the same time as the duration of the presented force) to produce the matched force. In some studies, the presented force is ramped up and down for 0.25 s at the beginning and the end of the presentation period, to avoid overshoots in the application of the external force from the controller [16]. Auditory “go” and “stop” signals can be used to notify participants when the presented force starts and stops, as well as when to start and stop reproducing the presented force (matched force). In every trial, participants receive an external force of different intensity. The levels of the presented forces typically vary between 0.5 and 4 N, although one study tested forces with intensity up to 10 N [11]. Each level of the presented force is repeated several times (e.g., 6–16), and the matched forces of each participant within a temporal window (usually the last 500 ms) are averaged across the same force level.

44

Konstantina Kilteni

Typically, the participants produce the matched forces in two ways: either by pressing with their right index finger or by using the slider. The two ways respectively correspond to the press condition (also called direct in some studies [15, 16]) and the slider condition. In the slider condition, the participants move the wiper of the potentiometer with their right hand. When one moves the wiper of the slider from one side to the other, its output voltage changes. Therefore, the slider can linearly control the force output on the participants’ fingers. The two sides of the slider are set to correspond to predefined force outputs. For example, the lower limit can be set to a minimum force output (e.g., 0 N) and the upper limit to the highest force output set by the experimenter (e.g., 5 N). Before the beginning of each trial, the wiper of the slider is set to the lower limit. Instead of a slider, some studies have used a joystick to control the force output [1, 5]. Note 4.2 discusses strategies to train participants to use the slider. In the press condition, the participants press their right index finger against their left index finger through the probe [5, 16]. Alternatively, they can press their right index finger against a force sensor that is placed above, but not in contact with, the probe. This sensor then controls the force output of the lever [1, 4, 13, 14]. Although both versions have shown robust somatosensory attenuation, the second version is somewhat preferred because the self-generated force is applied through the motor, as in the slider condition, which allows for a better comparison between the experimental and the control conditions respectively. As mentioned above, the participants are given 3 s to produce the matched force so that it matches as closely as possible the presented force. Before the experiment, the participants should be explicitly encouraged to finely tune their responses within the given period (3 s) as much as they think, since their latest estimation will be considered as their response. Moreover, the experimenter needs to ensure that the participants have not released the sensor (press condition) or returned the wiper to the lowest value (slider condition) before the end of the 3 s. Participants can “cheat” in the force-matching task and perform the task based on sensory cues other than the somatosensory ones. For example, a participant can use visual information about the pressure applied to their finger during the presented force (skin deformation of the finger, skin color changes due to changes in blood flow, change in the position of the probe or lever) or auditory cues created either by the motor or the probe when applying the presented force. The participant can then use this information to produce a matched force that matches these visual or auditory consequences. To avoid this, the experimenter can hide the participants’ left hand and related pieces of the equipment using a screen or a cloth, and the participants can wear headphones, for example.

Somatosensory Attenuation

45

Fig. 3 Example of force-matching data from ref. 14. Forces generated by the participants (matched forces) as a function of the externally generated forces (presented forces) (mean ± SE across 28 participants). The dotted line represents theoretically perfect performance, and the colored lines are the fitted regression lines for each condition. The position of the markers has been horizontally jittered for visualization purposes. There is significantly greater overestimation in the press than in the slider condition (paired t-test, t(27) = 13.57, p < 0.001)

Finally, no feedback should be provided to the participants concerning their performance, neither during the training nor during the task, as this would greatly bias the participant and interfere with the task. The main pattern of results from the force-matching task is that participants overestimate the required forces in the press condition (Fig. 3); that is, they match an externally generated force with a selfgenerated force of higher intensity. This overestimation suggests that the participants attenuate their self-generated forces; for example, when a participant produces 2 N to match the externally generated 2 N, he or she perceives the self-generated 2 N at a weaker intensity and must therefore increase the intensity of the self-produced force to compensate for the somatosensory attenuation. Interestingly, this overestimation is not observed in the slider condition. Although there is a clear mapping between the slider value and the resulting force, participants produce accurate selfgenerated forces. The absence of attenuation when using the slider suggests that the motor command generated to move the wiper of the slider does not allow the prediction of touch and thus its attenuation. The slider condition is therefore considered a control condition since it assesses the participants’ somatosensory perception in the absence of any attenuation effects.

46

Konstantina Kilteni

3.2 ForceDiscrimination Task

In the force-discrimination task, the participants receive two taps on the pulp of their left index finger—the test tap and the comparison tap—and they are asked to indicate which tap feels stronger, the first or the second. They can indicate it either verbally [19] or using foot pedals [20] or response buttons [8, 21]. The taps have a short duration (e.g., 80–100 ms) and the delay between their presentation varies to avoid the anticipation of the second tap (e.g., between 800 and 1500 ms). The intensity of the test tap is fixed (e.g., 2 N) while the intensity of the comparison tap varies from trial to trial (e.g., between 1 and 3 N). Typically, two conditions are tested to assess somatosensory attenuation. In the baseline or external condition, participants receive the two taps on their left index finger while their right arm remains relaxed. This condition is considered a baseline because it assesses the participants’ force perception in the absence of any movement. In the active or self condition, participants actively move their right index finger to tap the force sensor that consequently triggers the test tap on their left index finger (Fig. 2). Note 4.3 presents additional instructions for the force-discrimination task and Note 4.4 discusses strategies and training to limit participant tapping variability. For each condition, the participants’ responses are fitted with psychometric curves (Fig. 4). The parameter of interest is the point of subjective equality (PSE) which represents the intensity at which the test tap feels as strong as the comparison tap. The PSE parameter quantifies the degree of somatosensory attenuation. However, a second parameter, the just noticeable difference (JND) might also be of interest to the researcher, since it reflects the participant’s sensitivity for the force discrimination and could be used to assess whether the task was more difficult to perform in one condition than in another.

3.3 Comparing the Tasks: Advantages and Disadvantages

Both the force-matching task and the force-discrimination task show advantages with respect to the subjective ratings used in early studies on somatosensory attenuation. Subjective ratings might not be able to reflect small changes in the participants’ perception, or might be susceptible to task compliance effects. Different participants might have different internal rating scales, but this should not be problematic for within-subject designs. However, the same participant might change their scale during the experiment; this can be due to post-perceptual cognitive processes. For example, when participants are asked to report how intense the sensation feels now, they might give a different response after having experienced more intense stimuli compared to before. In addition, the labels assigned to the extremes of the scale (e.g., not intense at all to extremely intense) can be interpreted differently by different participants, or can be misunderstood (e.g., not perceptible to painful, respectively).

Somatosensory Attenuation

47

Fig. 4 Example of force-discrimination data from ref. 19. Plots generated using the mean point of subjective equality (PSE) and the mean just noticeable difference (JND) across the 30 participants for the self and the external condition. The PSE corresponds to the intensity of the comparison tap at which the test tap feels equally strong (probability of comparison tap being perceived as stronger than test tap = 0.5). The attenuation is seen as a significant decrease in the mean PSE in the self condition from the external condition (Wilcoxon signed rank test, V = 6, p < 0.001)

Both the force-matching and the force-discrimination tasks give more meaningful data than subjective ratings. For the forcematching task, the participants’ responses are typically fitted with linear regression models, and two parameters can be extracted—the intercept and the slope. These two parameters can be used to assess whether the participants show an attenuation effect across all different force levels (change only in intercept) or depending on the force level (change also in slope), and/or whether the participants show changes in their sensitivity to the presented forces (change in slope). The case for the force-discrimination task is similar. The PSE and the JND can be used to assess whether the experimental condition yields a change in the perceived intensity of the stimulus (PSE), and/or a change in the force discrimination capacity (JND). In principle, the force-matching task and the force-discrimination task both require more time (i.e., trials) than collecting subjective reports, although this strictly depends on the number of repetitions used by the experimenter. However, the outlined advantages of the two psychophysical methods and the disadvantages of using subjective reports might compensate for the additional time required to collect the psychophysics data.

48

Konstantina Kilteni

One specific disadvantage of the force-discrimination task might be the time required to collect the data in each condition, with respect to the force-matching task. This can be an issue when the experimenter is interested in comparing multiple conditions. When using fixed intensity values for the comparison tap, 10 repetitions per force level has provided sufficiently good fits [19, 20]. In addition, researchers might decrease the step size around the expected value of the PSE. For example, if the test tap has intensity of 2 N, a researcher could sample responses around this value in steps of 0.25 N or less (e.g., 1.75, 2.00, 2.25 N) while increasing the step size when going further from this value (e.g., 1.0, 1.5, 2.5, 3.0 N) for the comparison tap. Besides the step size, the experimenter might also choose to adjust the number of repetitions according to the step size, so that he or she collects more responses for intensities of the comparison tap around the value of the test tap, and fewer responses for intensities in the tails of the curve (e.g., at 1 and 3 N). Alternatively, the experimenter can use an approach that adapts to the participants’ responses. For example, the researcher can determine the intensity of the comparison tap for the next trial by fitting the responses from all previous trials using a maximum likelihood procedure and limiting the range of intensities to a part of the fitted curve [8] and Chapter 1, this volume.

4 4.1

Notes Comfort

In the setups used for the force-matching and force-discrimination tasks, the tested hand of the participants needs to be placed palm up so that the forces are applied on the pulp of the index finger. This posture might produce sensations of discomfort to some participants or even lead to reports of painful sensations in the shoulder or wrist during the experiment. If uncomfortable sensations are experienced, it is highly likely that the participant will move his/her hand during the experiment in order to change posture. Changing the position of the hand during the experiment might have two drawbacks. First, the forces might not be applied correctly by the controller, resulting in force overshoots. Second, the area of the finger that is tested will change during experimentation. It is therefore very important to ensure that the participants adopt a comfortable posture prior to the experiment, which they can maintain throughout the experiment. This can be further improved by placing sponges or vacuum pillows under the wrist and forearm, taking breaks so that participants can stretch, and reducing the duration of the experimental conditions.

Somatosensory Attenuation

49

4.2 Instructions in the Force-Matching Task

As with every experiment, it is fundamental that the participants understand the task they are asked to execute. With respect to the slider condition, it is important that the participants understand how the slider works and explore its full range before the experiment. In my experience, a few participants might avoid moving the wiper within the upper range of the slider because they anticipate experiencing painful sensations. Therefore, it is important to train participants in using the slider before the start of the experiment, by asking them to move the wiper all over the range, either in an ascending or descending order, with both a slow and a fast velocity. Moreover, to control for any time or habituation effects due to learning the devices or the task, it is important to counterbalance the order of the two conditions (press and slider) across participants.

4.3 Instructions in the ForceDiscrimination Task

In the force-discrimination task, it is important to clarify that the participants have to judge the intensity of the two taps experienced on the passive index finger and not between taps on the active and passive finger. In addition, it is also important to emphasize that they should not aim to balance their responses—that is, to give the same number of “first” and “second” responses. Participants might believe that half of the answers should be “first” and half should be “second” and this could lead them to switch to “first” responses when they perceive that they gave too many “second” responses for example. Moreover, participants might experience difficulties in choosing between the two taps especially when the intensities of the taps are very close or the same. To avoid confusion during the experiment, the experimenter can instruct the participants beforehand to guess in this particular case. Furthermore, if a participant has a bias to select the first (or the second) tap in general, or whenever unsure, the baseline condition can capture and remove this in its comparison with the self condition. Finally, in order to focus their attention, participants can be asked to fixate their gaze during the task on a spot that is away from the motor and finger.

4.4 Intensity of the Active Tap

It is advisable to control the intensity of the participants’ active tap so that it remains stable across the experiment and (if possible) matches the intensity of the test tap. Although it has been shown not to be a critical factor [1], a correspondence between the intensity of the tap of the right index finger and the intensity of the tap simultaneously received on the left index finger can strengthen their causal relationship. For example, a participant could question the self-generated nature of the test tap if she applied a very light tap with her right index finger in one trial and a very strong tap on another trial, but she received the same test tap on her left index finger in both trials. One way to do so is to train participants to tap with a certain force intensity prior to the experiment. For example, in the study of Bays and colleagues [8], participants were trained to tap a force similar to the test tap (2.7 N) prior to the experiment.

50

Konstantina Kilteni

During the experiment, trials in which the active tap greatly deviated from that value (outside the range [1.75–3.50 N]) were rejected and repeated. Depending on the training, the participants, and the exclusion criteria, this method might result in the rejection and repetition of many trials that can further extend the duration of the experimental session. For example, in a pilot experiment, we trained participants to tap with 2 N force, and we rejected and repeated those trials in which the participants tapped outside a conservative range of intensities (1.5–2.5 N). By doing so, we ended up rejecting almost half of the trials and thus we stopped the experiment. This is mainly because participants do not seem to have a genuine insight into what a force of 2 N feels like. Moreover, even with extensive training, the experience of different levels of forces throughout the experiment makes it easy to confuse the desired force intensity for the active tap. As an alternative, the experimenter might not reject the trials but give verbal feedback to the participants to tap more strongly or weakly in the next trial. Although this does not extend the duration of the experiment, it can interfere with the participants’ performance if the feedback from the experimenter is too frequent. For example, the participants might focus more on applying correct forces with their right index finger than on discriminating the two forces on their left index finger. Another solution is that the experimenter instructs the participants to perform a familiar movement; for example, to “tap as when tapping the screen of a smartphone”. This produces taps of around 1–1.5 N. This instruction is well understood by participants, requires less feedback from the experimenter, and can produce more stable taps.

References 1. Bays PM, Wolpert DM (2008) Predictive attenuation in the perception of touch. In: Haggard EP, Rosetti Y, Kawato M (eds) Sensorimotor foundations of higher cognition. Oxford University Press, pp 339–358 2. Blakemore S-J, Wolpert D, Frith C (2000) Why cant you tickle yourself? Neuroreport 11:R11–R16. https://doi.org/10.1097/ 00001756-200008030-00002 3. Blakemore SJ, Frith CD, Wolpert DM (1999) Spatio-temporal prediction modulates the perception of self-produced stimuli. J Cogn Neurosci 11:551–559. https://doi.org/10.1162/ 089892999563607 4. Kilteni K, Ehrsson HH (2017) Sensorimotor predictions and tool use: hand-held tools attenuate self-touch. Cognition 165:1–9. https://doi.org/10.1016/j.cognition.2017. 04.005

5. Shergill SS, Bays PM, Frith CD, Wolpert DM (2003) Two eyes for an eye: the neuroscience of force escalation. Science 301:187. https:// doi.org/10.1126/science.1085327 6. Weiskrantz L, Elliott J, Darlington C (1971) Preliminary observations on tickling oneself. Nature 230:598–599. https://doi.org/10. 1038/230598a0 7. Claxton G (1975) Why can’t we tickle ourselves? Percept Mot Skills 41:335–338. https://doi.org/10.2466/pms.1975.41. 1.335 8. Bays PM, Wolpert DM, Flanagan JR (2005) Perception of the consequences of self-action is temporally tuned and event driven. Curr Biol 15:1125–1128. https://doi.org/10.1016/j. cub.2005.05.023

Somatosensory Attenuation 9. Harris CR, Christenfeld N (1999) Can a machine tickle? Psychon Bull Rev 6:504–510. https://doi.org/10.3758/BF03210841 10. Kilteni K, Andersson BJ, Houborg C, Ehrsson HH (2018) Motor imagery involves predicting the sensory consequences of the imagined movement. Nat Commun 9:1617. https:// doi.org/10.1038/s41467-018-03989-0 11. Walsh LD, Taylor JL, Gandevia SC (2011) Overestimation of force during matching of externally generated forces. J Physiol 589: 547–557. https://doi.org/10.1113/jphysiol. 2010.198689 12. Palmer CE, Davare M, Kilner JM (2016) Physiological and perceptual sensory attenuation have different underlying neurophysiological correlates. J Neurosci 36:10803–10812. https://doi.org/10.1523/JNEUROSCI. 1694-16.2016 13. Kilteni K, Ehrsson HH (2017) Body ownership determines the attenuation of selfgenerated tactile sensations. Proc Natl Acad Sci 114:8426–8431. https://doi.org/10. 1073/pnas.1703347114 14. Kilteni K, Ehrsson HH (2020) Functional connectivity between the cerebellum and somatosensory areas implements the attenuation of self-generated touch. J Neurosci 40:894–906. https://doi.org/10.1523/JNEUROSCI. 1732-19.2019 15. Wolpe N, Zhang J, Nombela C et al (2018) Sensory attenuation in Parkinson’s disease is related to disease severity and dopamine dose. Sci Rep 8:15643. https://doi.org/10.1038/ s41598-018-33678-3 16. Wolpe N, Ingram JN, Tsvetanov KA et al (2016) Ageing increases reliance on sensorimotor prediction through structural and functional differences in frontostriatal circuits. Nat Commun 7:13034. https://doi.org/10. 1038/ncomms13034 17. Voss M, Bays PM, Rothwell JC, Wolpert DM (2007) An improvement in perception of selfgenerated tactile stimuli following theta-burst stimulation of primary motor cortex. Neuropsychologia 45:2712–2717. https://doi.org/ 10.1016/j.neuropsychologia.2007.04.008 18. Teufel C, Kingdon A, Ingram JN et al (2010) Deficits in sensory prediction are related to delusional ideation in healthy individuals. Neuropsychologia 48:4169–4172. https://doi. org/10.1016/j.neuropsychologia.2010. 10.024 19. Kilteni K, Engeler P, Ehrsson HH (2020) Efference copy is necessary for the attenuation of self-generated touch. iScience 23:100843. https://doi.org/10.1016/j.isci.2020.100843

51

20. Kilteni K, Houborg C, Ehrsson HH (2019) Rapid learning and unlearning of predicted sensory delays in self-generated touch. Elife 8: 1–17. https://doi.org/10.7554/eLife.42888 21. Bays PM, Flanagan JR, Wolpert DM (2006) Attenuation of self-generated tactile sensations is predictive, not postdictive. PLoS Biol 4:281– 284. https://doi.org/10.1371/journal.pbio. 0040028 22. Hesse MD, Nishitani N, Fink GR et al (2010) Attenuation of somatosensory responses to self-produced tactile stimulation. Cereb Cortex 20:425–432. https://doi.org/10.1093/ cercor/bhp110 23. Blakemore S-J, Wolpert DM, Frith CD (1998) Central cancellation of self-produced tickle sensation. Nat Neurosci 1:635–640. https:// doi.org/10.1038/2870 24. Shergill SS, White TP, Joyce DW et al (2013) Modulation of somatosensory processing by action. Neuroimage 70:356–362. https://doi. org/10.1016/j.neuroimage.2012.12.043 25. Ackerley R, Hassan E, Curran A et al (2012) An fMRI study on cortical responses during active self-touch and passive touch from others. Front Behav Neurosci 6:51. https://doi.org/ 10.3389/fnbeh.2012.00051 26. Arikan BE, van Kemenade BMV, Podranski K et al (2019) Perceiving your hand moving: BOLD suppression in sensory cortices and the role of the cerebellum in the detection of feedback delays. J Vis 19:4. https://doi.org/10. 1167/19.14.4 27. Pazen M, Uhlmann L, van Kemenade BM et al (2020) Predictive perception of self-generated movements: commonalities and differences in the neural processing of tool and hand actions. Neuroimage 206:116309. https://doi.org/ 10.1016/j.neuroimage.2019.116309 28. Boehme R, Hauser S, Gerling GJ et al (2019) Distinction of self-produced touch and social touch at cortical and spinal cord levels. Proc Natl Acad Sci 116:2290–2299. https://doi. org/10.1073/pnas.1816278116 29. Blakemore SJ, Frith CD, Wolpert DM (2001) The cerebellum is involved in predicting the sensory consequences of action. Neuroreport 12:1879–1884. https://doi.org/10.1097/ 00001756-200107030-00023 30. Diedrichsen J, Verstynen T, Hon A et al (2003) Anticipatory adjustments in the unloading task: is an efference copy necessary for learning? Exp Brain Res 148:272–276. https://doi.org/10.1007/s00221-0021318-z 31. Klaffehn AL, Baess P, Kunde W, Pfister R (2019) Sensory attenuation prevails when

52

Konstantina Kilteni

controlling for temporal predictability of selfand externally generated tones. Neuropsychologia 132:107145. https://doi.org/10.1016/ j.neuropsychologia.2019.107145 32. McNamee D, Wolpert DM (2019) Internal models in biological control. Annu Rev Control Robot Auton Syst 2:339–364. https://doi. org/10.1146/annurev-control060117-105206 33. Wolpert DM, Flanagan JR (2001) Motor prediction. Curr Biol 11:R729–R732. https:// doi.org/10.1016/S0960-9822(01)00432-8 34. Franklin DW, Wolpert DM (2011) Computational mechanisms of sensorimotor control. Neuron 72:425–442. https://doi.org/10. 1016/j.neuron.2011.10.006 35. Shadmehr R, Krakauer JW (2008) A computational neuroanatomy for motor control. Exp Brain Res 185:359–381. https://doi.org/10. 1007/s00221-008-1280-5 36. Kawato M (1999) Internal models for motor control and trajectory planning. Curr Opin Neurobiol 9:718–727. https://doi.org/10. 1016/S0959-4388(99)00028-8 37. Faisal AA, Selen LPJ, Wolpert DM (2008) Noise in the nervous system. Nat Rev Neurosci 9:292–303. https://doi.org/10.1038/ nrn2258 38. Bays PM, Wolpert DM (2007) Computational principles of sensorimotor control that minimize uncertainty and variability. J Physiol 578:387–396. https://doi.org/10.1113/ jphysiol.2006.120121 39. Brooks JX, Cullen KE (2019) Predictive sensing: the role of motor signals in sensory processing. Biol Psychiatry Cogn Neurosci Neuroimaging 4:842–850. https://doi.org/ 10.1016/j.bpsc.2019.06.003 40. Synofzik M, Vosgerau G, Newen A (2008) Beyond the comparator model: a multifactorial two-step account of agency. Conscious Cogn 17:219–239. https://doi.org/10.1016/j.con cog.2007.03.010 41. Synofzik M, Vosgerau G, Voss M (2013) The experience of agency: an interplay between prediction and postdiction. Front Psychol 4:1–8. https://doi.org/10.3389/fpsyg.2013.00127 42. Haggard P (2017) Sense of agency in the human brain. Nat Rev Neurosci 18:196–207. https://doi.org/10.1038/nrn.2017.14 43. Parees I, Brown H, Nuruki A et al (2014) Loss of sensory attenuation in patients with functional (psychogenic) movement disorders. Brain 137:2916–2921. https://doi.org/10. 1093/brain/awu237 44. Shergill SS, Samson G, Bays PM et al (2005) Evidence for sensory prediction deficits in

schizophrenia. Am J Psychiatry 162:2384– 2386. https://doi.org/10.1176/appi.ajp. 162.12.2384 45. Shergill SS, White TP, Joyce DW et al (2014) Functional magnetic resonance imaging of impaired sensory prediction in schizophrenia. JAMA Psychiat 71:28–35. https://doi.org/ 10.1001/jamapsychiatry.2013.2974 46. Blakemore S-J, Smith J, Steel R et al (2000) The perception of self-produced sensory stimuli in patients with auditory hallucinations and passivity experiences: evidence for a breakdown in self-monitoring. Psychol Med 30: 1131–1139. https://doi.org/10.1017/ S0033291799002676 47. Lemaitre A-L, Luyat M, Lafargue G (2016) Individuals with pronounced schizotypal traits are particularly successful in tickling themselves. Conscious Cogn 41:64–71. https:// doi.org/10.1016/j.concog.2016.02.005 48. Frith C, Blakemore SJ, Wolpert DM (2000) Explaining the symptoms of schizophrenia: abnormalities in the awareness of action. Brain Res Rev 31:357–363. https://doi.org/10. 1016/S0165-0173(99)00052-1 49. Pynn LK, DeSouza JFX (2013) The function of efference copy signals: implications for symptoms of schizophrenia. Vis Res 76:124– 133. https://doi.org/10.1016/j.visres.2012. 10.019 50. Fletcher PC, Frith CD (2009) Perceiving is believing: a Bayesian approach to explaining the positive symptoms of schizophrenia. Nat Rev Neurosci 10:48–58. https://doi.org/10. 1038/nrn2536 51. Moberget T, Ivry RB (2019) Prediction, psychosis, and the cerebellum. Biol Psychiatry Cogn Neurosci Neuroimaging 4:820–831. https://doi.org/10.1016/j.bpsc.2019. 06.001 52. Crapse TB, Sommer MA (2008) Corollary discharge across the animal kingdom. Nat Rev Neurosci 9:587–600. https://doi.org/10. 1038/nrn2457 53. Cullen KE (2004) Sensory signals during active versus passive movement. Curr Opin Neurobiol 14:698–706. https://doi.org/10.1016/ j.conb.2004.10.002 54. Straka H, Simmers J, Chagnaud BP (2018) A new perspective on predictive motor signaling. Curr Biol 28:R232–R243. https://doi.org/ 10.1016/j.cub.2018.01.033 55. Schneider DM, Mooney R (2018) How movement modulates hearing. Annu Rev Neurosci 41:553–572. https://doi.org/10.1146/ annurev-neuro-072116-031215

Somatosensory Attenuation 56. Poulet JFA (2006) The cellular basis of a corollary discharge. Science (80- ) 311:518–522. https://doi.org/10.1126/science.1120847 57. Poulet JFA, Hedwig B (2003) Corollary discharge inhibition of ascending auditory neurons in the stridulating cricket. J Neurosci 23: 4717–4725. https://doi.org/10.1523/ JNEUROSCI.23-11-04717.2003 58. Schneider DM, Sundararajan J, Mooney R (2018) A cortical filter that learns to suppress the acoustic consequences of movement. Nature 561:391–395. https://doi.org/10. 1038/s41586-018-0520-5 59. Fukutomi M, Carlson BA (2020) A history of corollary discharge: contributions of mormyrid weakly electric fish. Front Integr Neurosci 14: 42. https://doi.org/10.3389/fnint.2020. 00042 60. Sawtell NB (2017) Neural mechanisms for predicting the sensory consequences of behavior:

53

insights from electrosensory systems. Annu Rev Physiol 79:381–399. https://doi.org/10. 1146/annurev-physiol-021115-105003 61. Cullen KE (2012) The vestibular system: multimodal integration and encoding of selfmotion for motor control. Trends Neurosci 35:185–196. https://doi.org/10.1016/j.tins. 2011.12.001 62. Brooks JX, Carriot J, Cullen KE (2015) Learning to expect the unexpected: rapid updating in primate cerebellum during voluntary self-motion. Nat Neurosci 18:1310–1317. https://doi.org/10.1038/nn.4077 63. Roy JE (2004) Dissociating self-generated from passively applied head motion: neural mechanisms in the vestibular nuclei. J Neurosci 24:2102–2111. https://doi.org/10.1523/ JNEUROSCI.3988-03.2004

Chapter 3 Muscle Tendon Vibration: A Method for Estimating Kinesthetic Perception Anne Kavounoudias, Caroline Blanchard, Caroline Landelle, and Marie Chancel Abstract Perception of self-body movements (i.e., kinesthesia) relies on muscle proprioception. Moving our body segments activates proprioceptors, especially those located within muscles that are sensitive to muscle stretching, and skin mechanoreceptors due to the concomitant stretching or brushing of the skin. To date, impairment of kinesthetic function has been assessed in clinical studies by comparing detection thresholds or position matches of passively moved segments. These traditional methods simultaneously activate muscle and skin. For research purposes, as well as for diagnosis, it is critical to disentangle these two sensory sources, as they may not contribute equally to kinesthesia across the lifespan, and may be affected unequally in pathological circumstances. This chapter presents a well-established method for specifically activating muscle afferents, the so-called muscle tendon vibration. Applied on a muscle tendon of a relaxed participant, vibrations can give rise to an illusory movement in the direction of the stretching of the vibrated muscle, with a velocity proportional to the vibration frequency. The required context and appropriate vibration parameters for eliciting such an illusion are provided in this chapter. We describe psychophysiological approaches that can be used to collect relevant indices quantifying proprioceptive sensitivity. In functional magnetic resonance imaging (fMRI) studies, non-magnetic vibrators can also be used to examine the brain activity of the sensorimotor network during tendon vibration. Highlighting the advantages of this approach, recent studies have shown that muscle proprioception is more impaired than touch in the elderly, a result that could not have been observed using traditional approaches. Key words Muscle proprioception, Kinesthetic illusion, Self-movement, Psychophysics, Physiotherapeutic rehabilitation

1

Introduction The static sense of limb position (i.e., statesthesia) and the dynamic sense of limb movement (i.e., kinesthesia) are two distinct abilities that are encompassed in the more general notion of proprioception [1–3]. This classification of proprioception into two senses is supported by observations that the perception of position can change

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_3, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

55

56

Anne Kavounoudias et al.

without changing the perception of movement [4, 5]. This differential perception can be explained by the fact that different mechanoreceptors in the skin, muscles, joints, and tendons have different mechanical properties that make them more sensitive to either sustained stimulation, or to sudden changes in stimulation. In particular, primary endings of muscle spindles respond more to the magnitude and speed of change in muscle length, which are parameters related to motion; whereas secondary endings of muscle spindles are sensitive more to the length change itself, and therefore only contribute to the sense of position [2]. In clinical practice, the sense of static position is commonly assessed by passively moving joints and by asking participants to match this imposed joint angle deviation using their other limb [6]. It is generally well accepted that the magnitude of matching errors can be a useful indicator of proprioceptive acuity. Similarly, kinesthetic acuity is often measured as our ability to detect passively imposed movement, which usually refers to the latency and/or minimal angular speed required to detect the imposed movement [7–9]. Less frequently, direction of movement (flexion or extension) is also considered in assessing kinesthesia using a discrimination task. However, our ability to finely encode the velocity of movement has not been considered as a systematic index of kinesthesia in clinical assessments [10]. In addition, moving body parts increases the activity of both proprioceptors within stretched muscles, and skin mechanoreceptors, due to the concomitant stretching or brushing of the skin. The activity of these two sensory receptors encodes information about the speed and direction of limb movement [11]. As a result, conventional methods based on passively imposed movement are not able to disentangle these two sensory sources, although they may be differentially affected by various clinical disorders. In this chapter, we present an efficient method that preferentially activates muscle afferents for self-body movement perception, the so-called muscle tendon vibration. We combine this with a psychophysical approach to assess the discrimination threshold of movement velocity, and with an electromyographic (EMG) method to perform an objective quantification of the associated motor responses. During the 1970s, several microneurographic studies performed in humans (see Chapter 15, this volume) demonstrated that a mechanical vibration applied on a muscle tendon increases the activity of muscle spindle endings. In particular, primary (Ia) afferents are specifically activated by small amplitude muscle tendon vibration (0.5 mm peak to peak), with a firing rate linearly proportional to vibration frequency up to 80 Hz [12, 13] (Fig. 1). This artificial message is sent to the central nervous system and interpreted as an actual lengthening of the stimulated muscle. It is therefore possible to use these physiological properties to induce

Proprioceptive Assessment Tool

57

Fig. 1 Muscle spindle afferents. (a) Schematic representation of muscle spindles. (b) Typical responses of Ia afferents to a mechanical vibration of various frequencies (note that this is an illustration, not actual recordings)

illusory joint movements, the stimulated limb thus being perceived as moving by its motionless participant. In a pioneering study, Goodwin et al. [14] applied a vibratory stimulus to the tendon of the biceps (flexor muscle of the arm) on blindfolded participants, who reported a clear illusion of arm extension movement, as if the biceps muscle had been stretched. Since this report, a large number of studies have described illusions of movement evoked by muscle tendon vibration of different body segments such as the hand [15– 17], head [18, 19], eyes [20], ankles [13, 21, 22], or even the whole body standing upright [23–26]. In addition to the stimulated muscle, during vibration-induced illusions of movement a small involuntary contraction was also recorded in the antagonist muscle to the vibrated muscle, that is, the muscle that would have been contracted during real movement [16, 27, 28]. Interestingly, this muscle response increased with the vibration frequency as well as with the illusion velocity. It is thought to have a central origin, since similar motor responses are observed during illusory movement induced without vibration, but by a pure visual stimulation [16]. It can also be perturbed by transcranial magnetic stimulation over the primary sensorimotor cortex [29]. In fact, several neuroimaging studies showed that not only sensory but also motor brain regions are activated during an illusion of movement evoked by muscle tendon vibration [30–34]. Muscle tendon vibration is a relevant methodological candidate to experimentally manipulate kinesthesia and elucidate its neural

58

Anne Kavounoudias et al.

basis. It also allows us to assess the integrity of the muscle proprioception modality through its perceptual and motor effects. The present chapter describes the technical, methodological, and practical aspects of this method. Recent results illustrate the benefit of coupling a psychophysical approach to mechanical tendon vibration. This has enabled us to precisely evaluate any potential alteration of muscle proprioception, for example as observed in people aged over 65 years, and to disentangle the respective contribution of muscles and skin to such kinesthetic alterations [28, 35]. Application of this method to clinical evaluations, sensorimotor rehabilitation, brain plasticity, as well as brain imaging, will also be discussed.

2

Materials

2.1 Mechanical Vibrators

Mechanical vibrators used for stimulating muscle tendons differ from those used for stimulating skin receptors (see Chapters 1 and 4 of this volume). They usually consist of biaxial DC motors equipped with small eccentric masses forming a cylinder of various sizes (Fig. 2a: Top: 5 cm long vibrator, diameter 2 cm; Bottom: 7 cm long vibrator, diameter 3 cm). They can be attached to the participant’s wrist (Fig. 2b) or ankle (Fig. 2c) by elastic bands. A majority of the studies studying muscle tendon vibration effects used custom devices (for a detailed example of the technical features of such devices see ref. 36). Vibrators can be controlled using a specific user interface built using National Instruments’ LabVIEW tools, or any other software system capable of driving analog cards. The customized program must drive the analog voltage outputs of the card, to control the rotations of the eccentric masses. A calibration is thus needed to measure the tension-frequency relationship to determine the required DC input giving rise to the chosen vibration frequencies. Based on recordings from an accelerometer mounted on the vibrator, Fig. 3 shows an example of a calibration curve with a vibration frequency increasing linearly between 30 and 80 Hz, and a voltage varying from 1 to 3 V. For several vibrators, one curve of calibration per vibrator is recommended.

2.2

To quantify illusions of movement (e.g., direction, velocity, and latency), participants can reproduce any movement they perceived during the vibration, either in real time using any free body segment or after the stimulation has ended using the vibrated body segment. A potentiometer or a motion capture system can be used to record the movement reproduced by the participant (Fig. 4). Concomitant involuntary motor responses can also be recorded using surface EMG electrodes placed over the antagonist muscle to the vibrated muscle (more details in Subheading 3.4). It is

Evaluation Tools

Proprioceptive Assessment Tool

59

Fig. 2 Tendon vibration. Vibrating devices of two different sizes (a), applied to the tendon muscle at the wrist (b) or ankle (c) level. (d) The control system for operating the vibrators and recording signals from a potentiometer and electromyographic (EMG) electrodes simultaneously, using only one analog I/O card driven by a customized software implemented, for example, in LabVIEW

Fig. 3 Example of calibration curve of a vibrator. The frequency of vibration (Hz) is plotted as a function of the input voltage (V)

possible to verbally assess the salience of illusions thanks to a subjective scale (e.g., from “1”—very weak illusion to “4”—very salient illusion, “as if my hand actually moved”), and/or to ask participants to choose which of two perceived illusory movements is the fastest during a forced-choice discrimination task (see Subheading 3.3).

60

Anne Kavounoudias et al.

Fig. 4 Illustration of an experimental set-up. The seated participant is exposed to vibratory stimulation applied on their right pollicis longus muscle tendon. Any illusory sensation perceived in the right hand during the stimulation is copied with a potentiometer held in the left hand. Electromyographic (EMG) signals from the right pollicis longus and extensor carpi muscles are recorded

3

Methods

3.1 Experimental Design and Procedure 3.1.1 Vibration Parameters

When vibration of 0.2–0.5 mm amplitude is applied perpendicularly to the distal tendon of a muscle, most of the primary terminations of the neuromuscular spindles respond, cycle by cycle, to the vibratory stimulus, up to approximately 80 Hz. Increasing the vibration frequency from 80 to 200 Hz generally shows a dropout of the spindles, which then no longer respond cycle by cycle, but only every two or three cycles (i.e., at subharmonic frequencies) [13]. Therefore, the range of vibration frequency should be 20–80 Hz. The amplitude of vibrations should also remain low to specifically activate the primary afferents. Indeed, increasing the amplitude of the vibration results in a greater recruitment of secondary terminations of the neuromuscular spindles, which in turn produces illusions of a displaced body segment instead of a moving body segment [1] (see Note 4.1). Within a given frequency range (optimally between 20 and 80 Hz), the vibratory stimulation can be constant or can be

Proprioceptive Assessment Tool

61

modulated to follow a ramp, or any complex pattern, to induce more complex illusions, for instance, those of written symbols [21]. Indeed, using five vibrators applied on different muscles around the ankle, Albert et al. (2006) delivered complex spatiotemporal patterns of vibration to mimic the proprioceptive feedback evoked during an actual movement of the foot, for example when the foot draws a letter or a symbol. For best responses, the vibration should last at least 6 s, but the optimal duration may vary from person to person. In addition, it has been reported that after a sustained vibration lasting 30 s, muscle spindles continue to discharge after stimulation has ended [37], and that motor after-effects occur which may last for several minutes [38]. Thus, to prevent such potential after-effects, the vibration should not exceed 12 s duration (see Note 4.2). For the same reason, inter-stimulus breaks are necessary, with a minimum duration of 6 s. It is also preferable to deliver vibration to several muscles in pseudo-randomized order, so that the stimulated muscle is different each time and the same muscle is never vibrated twice consecutively. 3.1.2

Sensory Context

Multisensory information contributes concurrently to the perception of body movements [15, 16]. Thus, to maximize the perceived illusion, it is important to prevent any inconsistent messages from other sensory sources, by keeping participants’ eyes closed and limiting other skin contacts that would indicate that the body segment is not actually moving.

3.1.3

Instructions

As an experimenter, you must ask participants to relax and you should check that the targeted muscle is not contracted when you apply the vibration. Indeed, muscular contraction alters the illusion of movement, to the point of even preventing its occurrence [1].

3.1.4

Recording

Potentiometer, motion capture and/or EMG signals can be recorded depending on the responses required: verbal, manual, pedal, on-line, or off-line (Fig. 4).

3.1.5 Practice and Familiarization Trials

Familiarization to the vibration is required as it is not obvious for participants to experience illusions of movement, especially those arising from a sense other than vision. The site of stimulation for the familiarization phase can be any muscle tendon of different muscles so that the participant can experience different directions and speeds of movement illusions. The vibrated muscle should be moderately elongated or stretched for better recruitment of neuromuscular spindles, and therefore a better illusion. For example, the elbow should be open at 110–120° for optimal vibration of the biceps and closed at 60–70° for vibration applied to the triceps. Note that some participants may be considered as non-responders if the vibration stimulation does not elicit any movement illusion.

62

Anne Kavounoudias et al.

This is typically fewer than 10% of participants (see Note 4.3). It is up to the experimenter to decide if their research question requires the inclusion of these non-responders or not. A preliminary practice of the online reproduction of perceived movement is also recommended, as it takes time for participants to feel the illusion of movement and to reproduce it at the same time using another body segment. Sessions of ~30–40 min with breaks are recommended to prevent fatigue or adaptation (see Note 4.4). 3.2 Quantification of Movement Illusions

Angular deviations recorded using a potentiometer during the online reproduction of illusory movements can be used to quantify illusions, compare responses to different vibration frequencies or compare performance between groups. Figure 5 shows two individual signals, one from a younger adult (dark gray line) and one from an older adult (light gray line), centered on the baseline before stimulation onset. Vibration at 60 Hz was applied to the right wrist abductor muscle. Several indices can be extracted from these recordings: latency of the illusion (i.e., how long it takes the illusion to occur from the start of the stimulation), direction of movement (upward direction in this example means a clockwise rotation of the hand), and mean velocity of movement derived from the regression of angular deviation on time between the onset of the illusion and the end of the stimulation. Note in this example that the older participant reported a later and slower illusion of clockwise hand rotation than the younger participant. This example has been extended to 13 younger and 19 older participants tested in the

Fig. 5 Examples of curves and extracted indices obtained from a potentiometer recording. Potentiometric recordings showing clockwise angular deviations (°) copied online with the left hand by the two participants. In this example, although the intensity of stimulation was the same for the two participants, the younger participant has a stronger proprioceptive illusion compared to the older participant. (Data replotted from Chancel et al. [35])

Proprioceptive Assessment Tool

63

study by Chancel and colleagues [35]. Using the muscle tendon vibration method, this study shows that kinesthesia is altered with aging, and in particular the ability to encode movement velocity based on muscle proprioceptive afferents. 3.3 Assessment of Discrimination Thresholds

Between-subjects differences found in vibration-induced illusions may be related to anatomical differences between individuals, or to their differing ability to relax, or to many other parameters not directly linked to the proprioceptive system. Therefore, one way to more accurately assess the proprioceptive acuity of a given population is to estimate its ability to discriminate the velocity of two proprioceptive illusions. To this end, a classical two-interval forced choice (2IFC) discrimination task can be used (see also Chapter 1, this volume). It consists of a 2IFC discrimination task with constant stimuli: in each trial participants receive a pair of proprioceptive stimuli (muscle tendon vibration), always including a reference stimulation randomly presented in either the first or second interval. Several frequencies of vibration around the reference frequency are systematically tested (test stimulation) against the reference stimulation. For example, a reference illusory movement elicited by a vibration of the pollicis longus muscle at 50 Hz can be compared to illusions elicited by a vibration of the same muscle ranging from 35 to 65 Hz [28]. Participants are instructed to focus on their hand to estimate as accurately as possible if the first or the second stimulation led to a faster perceived movement by answering verbally after each pair “first” or “second”. To evaluate the discrimination threshold for each participant, the difficulty of the task can be adjusted. In this case, the range of the tested stimuli around the reference can be widened or shrunk to make the discrimination task easier or harder, respectively. Each stimulation intensity should be tested at least 12 times to ensure a robust measure of the participant’s perception. The duration of the stimuli should be long enough for the illusion to arise but not too long that the perceived movement would not be anatomically possible to execute (e.g., 10 s in Landelle et al. [28]). The presentation order of the stimulation conditions should be counterbalanced for each participant. To avoid any fatigue, the experiment should be divided into several sessions of about 10 min. As mentioned before, proprioceptive inputs can be integrated with many additional sensory inputs (e.g., visual or auditory), thus we recommend performing the experiment in the dark, and for participants to wear earplugs. To evaluate and compare participants’ perceptual performance, the psychometric data (i.e., the proportion of “faster than the reference” answers at different stimulation intensities) are fitted by a cumulative Gaussian function in agreement with classical psychophysical procedures [39, 40]. Two indices characterize the participant’s performance: the point of subjective equality (PSE)

64

Anne Kavounoudias et al.

Fig. 6 Example of data collected during a psychophysical task using muscle vibration. (a) Examples of psychometric curves obtained from one young and one old participant. The just noticeable difference (JND) corresponds to half the intensity difference between the 75% and 25% points of the psychometric function. (b) Individual and mean JND for the 16 younger (dot symbols) and 17 older (triangle symbols) participants in response to proprioceptive stimulation. Full symbols are the group means, empty dark symbols are individual values, and the bars are the median of the groups **p < 0.01. (Data replotted from Landelle et al. [28])

corresponds to the stimulation intensity leading the participant to perceive an illusory movement on average as fast as the reference movement; the just noticeable difference (JND) corresponds to half the intensity difference between 25% and 75% points of the psychometric function, which is inversely related to the participant’s discrimination sensitivity. In other words, a smaller JND value corresponds with a higher sensitivity in the discrimination task. Using these parameters, the experimenter can compare different groups of participants, as has been done, for example, by Landelle and collaborators [28]—comparing between 16 younger and 25 older individuals. In this study, the authors observed a significant increase in the JND in the older group compared to the younger group (Fig. 6), highlighting the lower discriminative abilities in the elderly in responses to proprioceptive stimulation. 3.4 Measurements of Motor Responses 3.4.1

Recording

EMG activity of the agonist and antagonist muscles at the vibrating tendon can be recorded to assess motor responses. An involuntary motor response is expected in the antagonist muscle of the vibrated muscle, while no muscle response is expected in the agonist muscle. A pair of surface EMG electrodes should be placed several centimeters apart on clean skin, over the belly of the muscle, and in

Proprioceptive Assessment Tool

65

parallel to its fibers; one reference electrode should be located close to a bony region. The experimenter should first inspect the EMG signal at baseline, ensuring it is below 10–15 μV in peak-to-peak amplitude (e.g., 50 or 60 Hz background noise from the power supply could increase this baseline, and should be avoided). Then ask the participant to contract the corresponding muscle to evaluate the recording during an actual muscular contraction and check the electrode placement. It is strongly recommended to record a maximum voluntary contraction as a value of a reference contraction for each muscle and then express the EMG responses recorded during vibratory stimulation as a percentage of this maximal contraction. This makes it possible to correct for inter-individual variability due to electrode placement, but also to physiological changes related to the study group, for example when comparing different age groups [28]. 3.4.2

Processing

3.5 Neural Basis of the Proprioceptive System

Most EMG devices pre-amplify the signal (e.g., ×1000) and should be set to filter the signal with band-pass from 10 Hz to at least 500 Hz, otherwise it is recommended to use analog filtering with this pass-band. The signal must have a sampling frequency at least twice the EMG band (1000 Hz). The raw signal should first be centered on the mean motor activity calculated before stimulation onset (e.g., 750 ms in Fig. 7) and rectified (negative amplitudes converted to positive amplitudes). To quantify individual EMG responses, the root mean squared (RMS) values are calculated over the duration of the stimulation. The mean individual EMG responses calculated as the mean RMS value can then be expressed as a percentage of the maximum voluntary contraction (% MVC). Muscle tendon vibration can also be used to investigate the sensorimotor network and its integrity. Pneumatic vibrators driven by air pressure have been developed to be compatible with the magnetic environment to perform fMRI studies. Several investigations converge and show that vibration-induced movement illusions activate both somatosensory- and motor-related brain areas [30–34]. Interestingly, using this approach, it has been possible to identify the brain sensorimotor network involved in kinesthesia [31–34], to compare brain activations associated with a proprioceptive- versus a tactile-induced illusion of movement [30], and to demonstrate that these cerebral networks are not equally affected by aging [32]. In particular, the latter authors showed that an alteration in the inter-hemispheric balance between the two primary sensorimotor cortices may be responsible for this age-related impairment, which affects proprioception more than touch. This method is thus a potential tool for clinical perspectives for both diagnosis and follow-up of longitudinal treatment carried out on patients suffering from sensorimotor disorders.

66

Anne Kavounoudias et al.

Fig. 7 Effect of vibration frequency on electromyography (EMG) amplitude. (a) Individual raw EMG activity (mV) recorded respectively in the extensor carpi ulnaris (ECU) and the pollicis longus (PL) muscles of the stimulated right hand, during the vibration of low (light gray) and high frequency (dark gray) of the pollicis longus. (b) Mean amplitudes (RMS: root mean square) of EMG responses in the ECU and PL muscles with respect to the vibration frequency applied

3.6 A Tool for Rehabilitation Perspectives

Muscle tendon vibration is also a promising tool for rehabilitation purposes. It has been shown that during transient deafferentation caused by immobilization of a limb, daily treatment of muscle vibrations can prevent cortical disruption in the sensorimotor network [41]. This method also seems efficient to reduce spasticity in patients with spinal cord injuries [42], to improve motor abilities in patients suffering from muscular dystrophy [43] or stroke [44], and to facilitate the processing of muscle afferents in patients with Parkinson’s disease [45]. By activating peripheral muscle spindle afferents, muscle tendon vibration is thus an efficient and non-invasive tool to promote the recovery of kinesthetic and motor functions in patients with sensorimotor disabilities.

Proprioceptive Assessment Tool

67

This method can also be useful for diagnostic purposes. For instance, using an fMRI approach, we investigated spinal cord activations during single vibrations applied at different levels of the lower limbs. This approach has recently allowed us to identify preserved sensorimotor circuits in the spinal cord of a patient suffering from a complete section of the spine [46].

4

Notes

4.1 Establishing Illusion Presence

The first problem to manage related to this method is to make sure that the participant really has a movement illusion (Subheading 3.1). To this end, you should: first, never tell the participant the direction of movement you are expecting; second, confirm the correct answer on another muscle site; and finally, when an illusion is present, it is often associated with a small involuntary contraction of the antagonistic muscle of the vibrating muscle—you can feel it if you hold the body segment of the participant (Subheading 3.4).

4.2 Avoiding AfterEffects

Another concern is the fact that after-effects can occur after only 30 s of sustained exposure to vibration. As mentioned in the Subheading 3, the duration of vibration should be limited to a few seconds with breaks between stimuli. Consecutive repetitions of a same stimulation should be avoided (Subheading 3.1).

4.3 Individual Differences

For most participants, vibration-induced illusion is not immediate, and often requires several tests before it occurs. Some participants (usually fewer than 10%) will never even feel any kinesthetic illusion (Subheading 3.1). One possible explanation may be that without a motor command there is a top-down process preventing illusion occurring in these participants.

4.4 Participant’s Attentional State

When applying vibration, the participant must be relaxed and not see (i.e., by closing the eyes) the vibrated limb, otherwise an illusion will not be induced, and an involuntary contraction may occur in the vibrated muscle instead. As mentioned in the Subheading 3.1, certain maneuvers such as the addition of extremely slow passive movements in the direction of the movement illusion are a very effective way to increase the perceptive effects of vibration, at least in the training phase.

References 1. McCloskey DI (1973) Differences between the senses of movement and position shown by the effects of loading and vibration of muscles in man. Brain Res 61:119–131 2. Proske U, Gandevia SC (2009) The kinaesthetic senses. J Physiol 587:4139–4146

3. Sittig AC, van der Gon JJD, Gielen CCAM (1985) Separate control of arm position and velocity demonstrated by vibration of muscle tendon in man. Exp Brain Res 60:445–453. https://doi.org/10.1007/BF00236930

68

Anne Kavounoudias et al.

4. Allen TJ, Proske U (2006) Effect of muscle fatigue on the sense of limb position and movement. Exp Brain Res 170:30–38. https://doi. org/10.1007/s00221-005-0174-z 5. Brown LE, Rosenbaum DA, Sainburg RL (2003) Limb position drift: implications for control of posture and movement. J Neurophysiol 90:3105–3118. https://doi.org/10. 1152/jn.00013.2003 6. Goble DJ (2010) Proprioceptive acuity assessment via joint position matching: from basic science to general practice. Phys Ther 90: 1176–1184. https://doi.org/10.2522/ptj. 20090399 7. Barrack RL, Skinner HB, Cook SD, Haddad RJ (1983) Effect of articular disease and total knee arthroplasty on knee joint-position sense. J Neurophysiol 50:684–687. https://doi.org/ 10.1152/jn.1983.50.3.684 8. Westlake KP, Wu YS, Culham EG (2007) Sensory-specific balance training in older adults: effect on position, movement, and velocity sense at the ankle. Phys Ther 87:560– 568 9. Wright ML, Adamo DE, Brown SH (2011) Age-related declines in the detection of passive wrist movement. Neurosci Lett 500:108–112. https://doi.org/10.1016/j.neulet.2011. 06.015 10. Hillier S, Immink M, Thewlis D (2015) Assessing proprioception: a systematic review of possibilities. Neurorehabil Neural Repair 29:933– 9 4 9 . h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 1545968315573055 11. Aimonetti J-M, Roll J-P, Hospod V, RibotCiscar E (2012) Ankle joint movements are encoded by both cutaneous and muscle afferents in humans. Exp Brain Res 221:167–176. https://doi.org/10.1007/s00221-0123160-2 12. Burke D, Hagbarth KE, Lofstedt L, Wallin BG (1976) Responses of human muscle-spindle endings to vibration of non-contracting muscles. J Physiol 261:673–693 13. Roll JP, Vedel JP (1982) Kinaesthetic role of muscle afferents in man, studied by tendon vibration and microneurography. Exp Brain Res 47:177–190 14. Goodwin GM, McCloske D, Matthews PB (1972) Proprioceptive illusions induced by muscle vibration – contribution by musclespindles to perception. Science 175:1382– 1384 15. Blanchard C, Roll R, Roll JP, Kavounoudias A (2011) Combined contribution of tactile and proprioceptive feedback to hand movement perception. Brain Res 1382:219–229

16. Blanchard C, Roll R, Roll JP, Kavounoudias A (2013) Differential contributions of vision, touch and muscle proprioception to the coding of hand movements. PLoS One 8:e62475 17. Brun C, Metral M, Chancel M, Kavounoudias A, Luyat M, Guerraz M (2015) Passive or simulated displacement of one arm (but not its mirror reflection) modulates the involuntary motor behavior of the other arm. Neuroscience 285:343–355. https://doi.org/ 10.1016/j.neuroscience.2014.11.036 18. Lackner JR, Levine MS (1979) Changes in apparent body orientation and sensory localization induced by vibration of postural muscles: vibratory myesthetic illusions. Aviat Space Environ Med 50:346–354 19. Guinet M, Michel C (2013) Prism adaptation and neck muscle vibration in healthy individuals: are two methods better than one? Neuroscience 254:443–451. https://doi.org/10. 1016/j.neuroscience.2013.08.067 20. Roll JP, Vedel JP, Roll R (1989) Eye, head and skeletal muscle spindle feedback in the elaboration of body references. Prog Brain Res 80: 113–123; discussion 57–60 21. Albert F, Bergenheim M, Ribot-Ciscar E, Roll JP (2006) The Ia afferent feedback of a given movement evokes the illusion of the same movement when returned to the subject via muscle tendon vibration. Exp Brain Res 172: 163–174 22. Regueme SC, Barthe`lemy J, Gauthier GM, Nicol C (2007) Changes in illusory ankle movements induced by tendon vibrations during the delayed recovery phase of stretchshortening cycle fatigue: an indirect study of muscle spindle sensitivity modifications. Brain Res 1185:129–135. https://doi.org/10. 1016/j.brainres.2007.09.046 23. Kavounoudias A, Gilhodes JC, Roll R, Roll JP (1999) From balance regulation to body orientation: two goals for muscle proprioceptive information processing? Exp Brain Res 124: 80–88 24. Hlavacka F, Mergner T, Krizkova M (1996) Control of the body vertical by vestibular and proprioceptive inputs. Brain Res Bull 40:431– 434 25. Ceyte H, Cian C, Zory R, Barraud P-A, Roux A, Guerraz M (2007) Effect of Achilles tendon vibration on postural orientation. Neurosci Lett 416:71–75. https://doi.org/10. 1016/j.neulet.2007.01.044 26. Kasai T, Yahagi S, Shimura K (2002) Effect of vibration-induced postural illusion on anticipatory postural adjustment of voluntary arm movement in standing humans. Gait Posture 15:94–100. https://doi.org/10.1016/ S0966-6362(01)00177-1

Proprioceptive Assessment Tool 27. Calvin-Figuiere S, Romaiguere P, Gilhodes JC, Roll JP (1999) Antagonist motor responses correlate with kinesthetic illusions induced by tendon vibration. Exp Brain Res 124:342–350 28. Landelle C, El Ahmadi A, Kavounoudias A (2018) Age-related impairment of hand movement perception based on muscle proprioception and touch. Neuroscience 381:91–104. https://doi.org/10.1016/j.neuroscience. 2018.04.015 29. Romaiguere P, Calvin S, Roll JP (2005) Transcranial magnetic stimulation of the sensorimotor cortex alters kinaesthesia. Neuroreport 16: 693–697 30. Kavounoudias A, Roll JP, Anton JL, Nazarian B, Roth M, Roll R (2008) Propriotactile integration for kinesthetic perception: an fMRI study. Neuropsychologia 46:567–575 31. Naito E, Morita T, Saito DN, Ban M, Shimada K, Okamoto Y, Kosaka H, Okazawa H, Asada M (2017) Development of right-hemispheric dominance of inferior parietal lobule in proprioceptive illusion task. Cereb Cortex 27:5385–5397. https://doi.org/10. 1093/cercor/bhx223 32. Landelle C, Anton J-L, Nazarian B, Sein J, Gharbi A, Felician O, Kavounoudias A (2020) Functional brain changes in the elderly for the perception of hand movements: a greater impairment occurs in proprioception than touch. Neuroimage 220:117056. https://doi. org/10.1016/j.neuroimage.2020.117056 33. Cignetti F, Vaugoyeau M, Nazarian B, Roth M, Anton J-L, Assaiante C (2014) Boosted activation of right inferior frontoparietal network: a basis for illusory movement awareness. Hum Brain Mapp 35:5166–5178. https://doi.org/ 10.1002/hbm.22541 34. Goble DJ, Coxon JP, Van Impe A, Geurts M, Van Hecke W, Sunaert S, Wenderoth N, Swinnen SP (2012) The neural basis of central proprioceptive processing in older versus younger adults: an important sensory role for right putamen. Hum Brain Mapp 33:895–908. https://doi.org/10.1002/hbm.21257 35. Chancel M, Landelle C, Blanchard C, Felician O, Guerraz M, Kavounoudias A (2018) Hand movement illusions show changes in sensory reliance and preservation of multisensory integration with age for kinaesthesia. Neuropsychologia 119:45–58. h t t p s : // d o i . o r g / 1 0 . 1 0 1 6 / j . neuropsychologia.2018.07.027 36. Pereira MP, Pelicioni PHS, Gobbi LTB (2017) RCVibro System: full description of a custommade vibratory system and its reliability. Braz J Phys Ther 21:440–448. https://doi.org/10. 1016/j.bjpt.2017.09.001

69

37. Ribot-Ciscar E, Rossi-Durand C, Roll JP (1998) Muscle spindle activity following muscle tendon vibration in man. Neurosci Lett 258:147–150 38. Duclos C, Roll R, Kavounoudias A, Roll JP, Forget R (2007) Vibration-induced posteffects: a means to improve postural asymmetry in lower leg amputees? Gait Posture 26:595– 602 39. Wichmann FA, Hill NJ (2001) The psychometric function: I. Fitting, sampling, and goodness of fit. Percept Psychophys 63:1293– 1313 40. Kingdom FAA, Prins N (2010) Psychophysics: a practical introduction. Elsevier Science & Technology Books, Jordan Hill 41. Roll R, Kavounoudias A, Albert F, Gay A, Fabre B, Roll JP (2012) Illusory movements prevent cortical disruption caused by immobilization. Neuroimage 62:510–519 42. Murillo N, Kumru H, Vidal-Samso J, Benito J, Medina J, Navarro X, Valls-Sole J (2011) Decrease of spasticity with muscle vibration in patients with spinal cord injury. Clin Neurophysiol 122:1183–1189. https://doi.org/10. 1016/j.clinph.2010.11.012 43. Ribot-Ciscar E, Milhe-De Bovis V, Aimonetti J-M, Lapeyssonnie B, Campana-Salort E, Pouget J, Attarian S (2015) Functional impact of vibratory proprioceptive assistance in patients with facioscapulohumeral muscular dystrophy. Muscle Nerve 52:780–787. https://doi.org/10.1002/mus.24605 44. Mortaza N, Abou-Setta AM, Zarychanski R, Loewen H, Rabbani R, Glazebrook CM (2019) Upper limb tendon/muscle vibration in persons with subacute and chronic stroke: a systematic review and meta-analysis. Eur J Phys Rehabil Med 55:558–569. https://doi.org/ 10.23736/S1973-9087.19.05605-3 45. Ribot-Ciscar E, Aimonetti J-M, Azulay J-P (2017) Sensory training with vibrationinduced kinesthetic illusions improves proprioceptive integration in patients with Parkinson’s disease. J Neurol Sci 383:161–165. https:// doi.org/10.1016/j.jns.2017.11.016 46. Rowald A, Komi S, Demesmaeker R, Baaklini E, Hernandez-Charpak SD, Paoles E, Montanaro H, Cassara A, Becce F, Lloyd B, Newton T, Ravier J, Kinany N, D’Ercole M, Paley A, Hankov N, Varescon C, McCracken L, Vat M, Caban M, Watrin A, Jacquet C, BoleFeysot L, Harte C, Lorach H, Galvez A, Tschopp M, Herrmann N, Wacker M, Geernaert L, Fodor I, Radevich V, Van Den Keybus K, Eberle G, Pralong E, Roulet M, Ledoux JB, Fornari E, Mandija S, Mattera L, Martuzzi R, Nazarian B, Benkler S, Callegari S,

70

Anne Kavounoudias et al. Greiner N, Fuhrer B, Froeling M, Buse N, Denison T, Buschman R, Wende C, Ganty D, Bakker J, Delattre V, Lambert H, Minassian K, van den Berg CAT, Kavounoudias A, Micera S, Van De Ville D, Barraud Q, Kurt E, Kuster N, Neufeld E, Capogrosso M, Asboth L, Wagner

FB, Bloch J, Courtine G (2022) Activitydependent spinal cord neuromodulation rapidly restores trunk and leg motor functions after complete paralysis. Nat Med 28(2):260–271. https://doi.org/10.1038/ s41591-021-01663-5. Epub 2022 Feb 7

Chapter 4 Creating Tactile Motion Tatjana Seizova-Cajic´, Xaver Fuchs, and Jack Brooks Abstract Interactions with the world and between our own body parts often result in motion across the skin, and sensory processing of this motion is of interest to both basic and applied researchers. We describe motion cues and common approaches to creating tactile motion to help researchers make choices for their study, and we give two illustrative examples. Major cues that inform about direction and speed of tactile motion are displacement over the skin, skin stretch, and skin vibration. The most common approach for creating tactile motion relies on apparent motion, the motion sensation arising from discrete—often vibrating—stimuli presented in quick succession to simulate object displacement. Other methods rely on real displacement of objects such as plates and brushes. They create friction and skin stretch, allowing the study of that cue and resulting in a richer motion experience. Friction can also be implemented in devices that rely on apparent motion, but this is still very rare. Some recent solutions, such as movement-contingent air turbulation, are driven by the desire to create immersive tactile experience in virtual reality. Illusions abound in perception of motion, as in any other aspect of perception. They include mislocalization of discrete motion stimuli known as sensory saltation, and filling-in of non-stimulated skin area with the sensation of continuous motion—the two effects we focus on in the Methods section. We use them to illustrate two manners of motion delivery (discrete stimulation vs real object displacement) and a variety of ways to measure perception of the moving stimulus. Key words Touch, Tactile motion, Motion cues, Motion perception, Apparent motion, Numb spots, Filling-in, Sensory saltation, Cutaneous rabbit, Somatosensory experiments

1

Introduction Most forms of touch involve motion across the skin, whether we do the touching, are touched ourselves, or both. Systematic investigations of perception of tactile motion date back to the nineteenth century and were concerned with direction discrimination [1]. A string of studies in the early twentieth century investigated apparent motion [2–5] and motion aftereffect [6], inspired by the analogous vision research. Most subsequent research on tactile motion used psychophysics, and a smaller number of studies, neurophysiological methods. A number of papers review this literature: see [7] for

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_4, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

71

72

Tatjana Seizova-Cajic´ et al.

twentieth-century studies on motion cues, [8, 9] for neural basis of tactile motion, [10] for the social and affective role of stroking the skin, a burgeoning field motivated partly by implementation in haptic interfaces [11], and [12–14] for substantial engineering work on haptic technology that either delivers tactile stimuli or registers them using artificial sensors (see Note 3.1 for different meanings of “haptics”). Specific topics studied over the years are represented in our sample of 50 papers selected at random from more than 623 papers retrieved through a systematic search (see Note 3.2 for search terms and links to the full list of references). 1.1

Chapter Aim

Our first problem in undertaking the following lines of research, (. . .) which so far as we know are mainly new, was to devise a suitable apparatus. (Hall & Donaldson, 1885, [1]).

Devising or choosing a suitable apparatus remains the important problem in research on tactile motion and is likely one of the reasons basic research on touch lags well behind vision [15]. The aim of this chapter is to help researchers with the choice of devices, and to illustrate their application using two specific devices as examples. Researchers who use fMRI or other imaging methods need to also consider the compatibility of the tactile motion device with the machine (for fMRI, see [16], and Chapter 18, this volume) [17]. 1.2 Sources of Information (Cues) About Tactile Motion

Motion across the skin stimulates the somatosensory system in a specific manner, providing the potential motion cues listed in Table 1. Two major cues are successive positions (displacement) and skin stretch, illustrated in Fig. 1. In ecological perception, moving texture elements, ridges, and edges often successively stimulate neighboring skin locations. In artificial displays, we experience motion between successively stimulated locations separated by non-stimulated areas (known as apparent motion [4]). The quality and vividness of the motion percept in the latter case (apparent motion) varies with many stimulus parameters (see [17–22]), including: (1) interstimulus onset interval (ISOI), also known as the stimulus onset asynchrony (SOA), the time from the onset of one stimulus to the onset of another; (2) inter-stimulus interval (ISI), the time from offset of one stimulus until the onset of the next; (3) stimulus duration; (4) other properties of the stimuli comprising the array (frequency and amplitude of vibration; indentation depth); (5) size and number of stimulators; (6) their spatial layout (spacing, 2D arrangement, and distance between the body parts if motion “jumps” across empty space). Interaction between the above variables is complex and parameters for optimal motion experience need to be tested for any given setup. In one study on the effect of different cues on the quality of perceived motion [20], a linear array of 4 stimulators that were 5.1 mm apart was applied to the finger of participants who judged

Creating Tactile Motion

73

Table 1 Cues (sources of information) for tactile motion; cues listed in the table inform about direction, speed, or both, as indicated Cue

Description

Spatiotemporal separation between successive Sequential stimulation of neighboring or relatable positions, or displacement (direction and speed positions; see text for details and [7] for review of cue) twentieth-century studies focusing on this cue. Skin stretch/friction (direction and speed cue)

Friction between the skin and a moving object creates a tangential force that pulls the skin in the direction of motion. It reduces direction threshold several-fold relative to the otherwise comparable frictionless stimuli [23]. If the skin is pulled in the direction opposite to stimulus displacement, detection of displacement is impaired [24]. See [7] for review of the twentieth-century studies.

Frequency of texture-elicited skin vibration (speed cue)

Moving textures and edges result in skin vibration, and the faster they move, the higher the vibration frequency. Texture density is a confound (denser textures also result in higher vibration frequencies), but our tactile system is able to extract speed, although imperfectly [25–27].

Proprioceptive and other information about posture and movement (direction and speed cues)

When body parts move relative to each other and through the environment, proprioception, vision and the vestibular system inform about those movements and link skin-centric and other frames of reference (ego-centric or external/allocentric frames; see [28–33].

motion as “definitely absent”, “hardly perceivable or ambiguous”, “unimpressive or discontinuous”, or “impressive and continuous”. At a stimulus duration of 100 ms, perceived motion quality gradually increased from 30 to 90 ms SOA, and then decreased at 100 and 110 ms SOA. However, other stimulus durations resulted in differently shaped functions and in poorer motion experience overall. The analysis of cues is an active area of research and the above list is without doubt incomplete and will expand. For example, recent studies suggest that wave fields that propagate through the skin during motion are a possible motion cue [34, 35] and new potential speed cues are identified in vibration patterns [36]. How we process tactile motion cues also depends on the spatial and temporal context, as demonstrated by grouping, aftereffects, and other phenomena. Gestalt grouping principles of proximity and similarity operate in tactile motion [37]. Adaptation to motion results in bias to perceive motion in the opposite direction, i.e., the negative aftereffect (although, unlike vision, touch requires a dynamic stimulus to make the negative aftereffect observable)

74

Tatjana Seizova-Cajic´ et al.

Fig. 1 Two major motion cues in touch. See text for details

[6, 38–40]. Perceived speed also adapts and may decrease by 30–40% following short periods of exposure [41, 42]. Masking and interference may occur between simultaneously and sequentially presented tactile patterns [43], and between tactile and visual motion patterns [44]. Days or weeks long exposure of the same skin area to intensive stimulation reduces its ability to process stimuli and may result in anomalous sensations such as ghosttouches (double or triple sensations from a single stimulus) [45]. On the other hand, training with specific patterns may also stabilize and refine the percept. Bekesy [17] described a “peculiar” outcome of training with alternating vibrations on contralateral thighs: “In time, (. . .) sensation can be localized in the free space between the knees, and [the observer] will be able to experience a displacement of the sensation in this free space (. . .).” (pp. 223). 1.3 Tactile Motion Illusions and Distortions

When point stimuli are applied to the skin in quick succession, some distance apart, the sensation of a single object moving across the skin may arise, known as apparent motion. This motion percept can be described as illusory because it does not require the presence of a moving object, although a real object, such as the rotating wheel with spikes, could in principle create the pattern described. The minimal number of points required to sense motion is two, and two-point stimulation alone can give rise to illusory effects. Bekesy [17] described the illusory motion between two vibrating tactors when the vibration intensity of one decreased while

Creating Tactile Motion

75

the other increased. He also described funneling, where the simultaneous activity of two tactors at equal intensity results in single touch perceived mid-way between them (for use of funneling in applications, see [46] and Fig. 2e). Adjusting the relative intensity of the two tactors results in an apparent shift toward the stronger stimulus. One much-studied illusion was first described by Geldard and Sherrick [47], who delivered brief taps at one location in quick succession, followed by a quick tap at another location. This resulted in the feeling of hops spread out between the two locations, known as sensory saltation or cutaneous rabbit, which does not occur if the temporal separation between the successive stimuli is longer (also see [48, 49]). The above illusion is only one instance of distortion at high velocities of motion. Others effects also show velocity-dependence: the perceived path of fast-moving stimuli (above 25 cm/s) is compressed and their endpoints mislocalized in the direction opposite to the motion direction. This is true of continuous motion ([50], motion crossing a gap [51], and apparent motion [52]. By contrast, the endpoint of a relatively slow motion is mislocalized in the direction of motion [52]. For example, the perceived path is several times shorter for velocities approaching 250 cm/s (which far exceeds the range of typical tactile motions) than for velocities within the 5–20 cm/s range [50]. Large gaps in motion trajectory may go unnoticed—10 cm gaps on the forearm [53] and 2 mm gaps on the fingertips [54]. This is known as filling-in and is analogous to the filling-in of blind spots in vision. Shapes of motion trajectories can also be misperceived. In one of the oldest reported illusions—the Weber illusion—parallel motion trajectories appear to converge or diverge due to different receptor density of skin patches they traverse [55, 56]. Perception of tactile motion can also be distorted due to medical conditions and is used as a diagnostic tool [57]. 1.4 Devices Used to Create Tactile Motion

There are many different ways to create tactile motion in the laboratory and in applications. Example devices used for this purpose are shown in Fig. 2, split into two groups according to the type of motion stimulus they create: continuous or shifting motion, or discrete stimulation leading to apparent motion. The top panel of Fig. 2 shows devices that create continuous (shifting) motion. How they contact the skin varies: some allow object tracking because the whole contact area translates across the skin (e.g., translating brush, Fig. 2a). Comparable ecological stimuli are crawling insects, self-touch during grooming, or affectionate strokes. In others, textured surfaces run over the same skin area, for example, the fingertips (Fig. 2b). This is similar to surface exploration using fingertips in everyday life, except that ecological exploration is normally active.

76

Tatjana Seizova-Cajic´ et al.

Fig. 2 Example research devices used to create tactile motion. (a) Brush driven across the skin by a servomotor. Stimulated area is limited by an aperture (see Fig. 1 in [58]). (b) A rotating ball with ridges moves over a limited skin region (see Fig. 1A in [59]). (c). Strong air currents impact the skin enabling users to feel virtual 3D objects. AIREAL air jet technology (see Fig. 1 in [60]). (d)*. Linear array of piezoelectric stimulators. The Princeton Linear Array of piezo tactors (see Fig. 1 in [22]) (e) A 2-D array of vibrators named “Tactile brush” (see Fig. 6 in [46]). (f) A dense array of laterally moving pins, “Latero” with 60 tactors (see Fig. 9 in [61])

Continuous motion stimuli have high spatial density, and density matters: direction thresholds for high-density stimuli are lower than for low-density stimuli when displacement is kept constant [62, 63]. Many of these devices create a level of friction, but some don’t: continuous frictionless motion is created using water, air jets, air puffs (see [7]), and more recently, ultrasound ([60] Fig. 2c). Continuous motion stimuli have been extensively used in basic research but have one disadvantage: they rely on physical motion, in which motion cues (see Sect. 1.2) are necessarily consistent with each other. It is sometimes preferable to study cues in isolation or in conflict with each other to find out how they interact. Moving objects also make for inflexible experimental setups and are usually not convenient for applications. Artificial displays using discrete elements to create the illusion of motion solve some of these problems and we turn to them next. Devices that rely on discrete stimuli and create apparent motion in touch (bottom half of Fig. 2) are analogous to artificial visual displays such as monitors and stereo goggles. Vibrators are a popular choice, used in approximately 53% of studies we surveyed (see Note 3.2). Even sparse devices, such as those illustrated in Fig. 2d, e, can deliver varied and smooth motion percepts, because perception is modulated by temporal parameters and stimulus intensity, as described earlier. The cost of hardware can be low: vibrators/

Creating Tactile Motion

77

eccentric motors (motors with off-center load) are mass-produced at low cost for use in mobile phones. Attached to the skin using medical tape, elastic bands, or with other means, they create a constant level of pressure and allow reliable stimulus delivery. There are also sophisticated instruments whose densely packed tactors can present complex motion patterns indistinguishable from real motion (except that they typically lack skin stretch cues, see below). One such device, Optacon, originally designed in the 1960s as a sensory substitution device [64], has been adjusted for use in research (see, e.g., [21]). Optacon has 144 pins arranged in a 6 × 24 matrix within approximately 3 cm2. Another instrument fits 400 vibrating pins within 1 cm2 [65], and the third—Latero, Fig. 2f—fits 60 laterally moving piezoelectric benders within the same area [61]. These devices allow the use of featureless, random dot patterns akin to random-dot kinematograms used in vision to study motion mechanisms (see Note 3.3). One limitation of discrete devices is that they typically rely on the successive positions cue and not on skin stretch. Given that we are very sensitive to skin stretch [23] and that it influences our perception of motion direction [24], it is desirable to include this cue by enabling lateral or rotational motion of discrete elements. The pins in Latero [61] move back-and-forth and stretch the skin, but the bidirectional stretch does not inform about motion direction (moving objects stretch the skin in the direction of motion). One solenoid-based device creates skin stretch by lateral motion of silicon-covered pin heads, which retract after moving in one direction to reposition for the next sweep; however, they stretch the skin too slowly compared to the speed from the successive positions cue of the same device [24]. A recent promising design solution uses a linear array of mini-wheels to mimic stroking of the skin (Lateral Skin-Slip Haptic Device [66]). 1.5 Approaches to the Study of Tactile Motion

We conducted a survey of empirical studies on tactile (across-theskin) motion and chose 50 original papers at random from a set of 623 papers (see Note 3.4). Approximately 40% of papers come from engineering sources whose emphasis tends to be on applications such as sensory substitution, virtual reality, robotics, and biomedical engineering, and 60% come from perception, neuroscience, and other science publications. Table 2 summarizes typical devices and measures used in those studies. A majority of studies use discrete motion stimuli. Discrete devices identified in our survey include small vibrators (53%), piezoelectric benders (22%), electrocutaneous and electrostatic devices (10%), solenoids (3%), and many others (12%), such as pneumatic stimuli. They are more commonly used than devices that rely on real, continuous motion, especially in publications from engineering sources. A typical study comprises simple linear translation applied to a body part at rest. Absence of bodily movements allows

78

Tatjana Seizova-Cajic´ et al.

Table 2 Survey results for a random sample of 50 studies on tactile motion (column and row totals exceed 50 because some studies used multiple stimuli and/or measures)

Type of motion stimulus Tasks/measures (one study used both (a number of studies used stimulus types) multiple measures)

Publication Engineering Perception, (19) Neuroscience, Other (31)

Continuous/shifting motion stimulus (used in 20 studies)

Psychophysical (19) Motor action (4) Neural recording (4)

5 1 0

14 3 4

Discrete motion stimulus (used in 31 studies)

Psychophysical (31) Motor action (3) Neural recording (4)

16 3 1

15 0 3

See text and Note 3.2 for more details

for better stimulus control but is a non-ecological situation reminiscent of vision studies that use a fixation point rather than free eye movements. There is interest in active touch in virtual reality research, where simulated touching is contingent on active interaction with virtual objects (e.g., [60]). Most studies use psychophysics and investigate perceived direction, speed, and position of the moving stimulus. Motor (action) responses and neurophysiological measures (including imaging) are used far less often (Table 2). For review of technical characteristics, advantages and disadvantages of a variety of devices, see [67] (esp. their Table 1, Appendix) and [13, 68].

2

Materials and Methods for One Discrete and One Continuous Motion Study In this section, we describe methods for creating and measuring two motion illusions described earlier: the cutaneous rabbit illusion or sensory saltation [47], and filling-in of a non-stimulated area (“numb-spot”) with illusory motion [51, 53, 54]. Sensory saltation is created using discrete stimulators, and filling-in, using continuous motion.

2.1 Sensory Saltation (the Cutaneous Rabbit Illusion)

The basic aspect of saltation is perceptual compression of distance between—or mislocalization of—successive taps separated by a short time interval. The extent of perceptual compression or displacement depends on the number and intensity of the taps, their spatiotemporal separation, and spatial sensitivity of the skin [69–71]. In the study that accidentally revealed a saltatory effect (described in [69]), Geldard used multiple taps in one location, followed by one or more taps in the second location. The nickname

Creating Tactile Motion

79

for the effect, “cutaneous rabbit”, vividly illustrates how some of the taps delivered to the first location perceptually “hop” away from it and toward the second location. When measuring saltation, we may wish to capture the perceived relative distances between the taps as well as their perceived locations, but measuring both at the same time is difficult if not impossible. Geldard used different methods to capture one or the other (see pp. 50–54 in [69]). In the matching method, participants adjusted the time interval between taps in the saltation pattern until perceived distances between them felt identical to distances of hops across equidistant physical locations presented on the contralateral limb. In what Geldard called verbal sectioning, participants expressed the degree of displacement of the second tap in the three-tap sequence as a proportion of the perceived distance between the first and the third tap. Geldard also used localization with a stylus, in which participants indicated the perceived position of the second tap (in some recent studies, the localization takes the form of a forced-choice task: participants report whether they perceive a tap in the given location or not [71–73]). This simple response does not indicate what the whole pattern feels like, because other taps (first and third in the Geldard’s three-tap sequence) are not necessarily perceived in their correct locations either. One reason for large shifts of the overall saltation pattern are shifts in attention [48], illustrated in Fig. 3.

Fig. 3 This figure illustrates two effects: the illusory displacement of saltation stimuli toward each other (open circles indicate their perceived locations, and filled circles, their physical locations) and how perceived position of the pair shifts depending on the instructions to participants about which part of the forearm they should attend to (no instructions, ‘Attend distal’, ‘Attend proximal’). The SOA for the saltation pair was 60 ms. Black points also indicate the physical positions of additional taps presented 500 ms before and after the saltation pair (Adapted from Fig. 1 in [48])

80

Tatjana Seizova-Cajic´ et al.

Fig. 4 Experimental setup and computation of displacement index in the study of sensory saltation. (a) Equipment and setup (see Note 3.4, which gives details regarding the function of digital acquisition device (DAQ), digital-to-analog converter (DAC), amplifier, and switch box). Black circles on participant’s forearm: active vibrators; grey circles: sham vibrators. The participant performs a localization task indicating perceived positions of stimuli on the forearm shown on the computer screen. (b) Position of the six active stimulators and two sham stimulators on the participant’s forearm. (c) Computation of a displacement index: Distance between first and second response (responses are indicated by ‘x’) is expressed as a proportion, i.e., percentage of distance between first and third response. Note that physical distance between 1st and 2nd tap in this example study is always zero but their perceived distance is non-zero

Several recent studies asked the participants to localize all presented taps in their perceived order (using a stylus [49], a finger slider [74], or graphical representation of the limb on a screen) [70, 75]. This method is not free of problems either: participants often report taps in the wrong spatial order (the second tap is not placed between the first and the third). Such errors are ambiguous because they may reflect misperceived temporal order, a genuine swap of spatial order, or random error. Given that proportion of trials with this error is large (24% [76]), the effect may be of interest on its own. With this caveat, we describe one such method for capturing the saltation effect. In this method, two taps are delivered at one location, and one tap 10 cm further away using solenoids (Fig. 4). Participants report perceived locations of taps by clicking on a picture of the forearm in the order in which they perceived them delivered. A typical percept includes taps in three distinct locations. 2.1.1

Materials

1. Six vibrotactile solenoids for the basic set-up and two more to use as sham stimulators (Dancer Design tactor, http://www. dancerdesign.co.uk/); adhesive rings to attach them to the skin.

Creating Tactile Motion

81

2. One amplifier for the solenoids (Dancer Design Tactamp, http://www.dancerdesign.co.uk/); one switchbox that routes the signal to the six stimulators (http://www.dancerdesign.co. uk/); a Labjack T4 digital data acquisition device (www. labjack.com) with digital-analogue-converter and digital outs (TTL); PC running Windows 10 operating system with PsychoPy (v.2020.2.8; [77]) installed (free Python-based software, see Note 3.4). 3. A monitor with mouse or trackpad for participants to use in the localization task (see Fig. 4a). 4. A hairdresser’s gown or another occluder (e.g., cardboard box). 5. Earplugs and/or headphones to mask the sound of solenoids. 2.1.2

Methods

Setup The participant is seated, with solenoids and arm occluded from view to prevent them seeing the exact locations or number of stimuli. Six solenoids are attached to the dorsal side of the forearm, in a linear array with 3.3 cm center-to-center distance (Fig. 4b). The middle of the array is centered at the mid-point between the wrist and elbow creases; these are the active stimulators. Two sham stimulators can be added, one at each end of the array, to further reduce the chance for the participants to anticipate stimulation points. Devices (stimulator, switch box, amplifier, DAQ-device, and PC) are connected as shown in Fig. 4a. The participant is seated in front of the monitor wearing earplugs (white or pink noise at comfortable volume can be played over the stereo headphones). The participant performs a localization task by mouse-clicking on the perceived positions of stimuli on a photographic representation of the forearm presented on a computer screen. Importantly, they are instructed to click on locations in the order in which they were perceived. Design and procedure The independent variable is test pattern. The saltation and control patterns are described below. They both involve three taps, two directions of motion (proximal and distal), and two stimulus locations 10 cm apart. All individual stimulus durations are 30 ms, and vibration frequency is 200 Hz (alternatively, non-vibratory, tap-like stimuli can be used; see [70] for an example). Each test pattern is presented a number of times in random or pseudorandom order. The optimal number of repetitions depends on response variability and the effect size of interest and needs to be determined in a pilot study (examples: 5 to 10 repetitions were used for direct localization in [49]; 25 repetitions were used in forced choice task in [71]).

82

Tatjana Seizova-Cajic´ et al.

1. Sensory saltation pattern: two stimulators are placed 10 cm apart; the first location receives two taps, and the second, one tap; the critical parameter is time interval between the second and third tap: SOA (tap 1 to tap 2) = 700 ms; SOA (tap 2 to tap 3) = 100 ms; different location pairs are used throughout the experiment; example patterns (numbers indicate locations) are: 1-1-4, 6-6-3; 2-2-5, 3-3-6, 4-4-1; different pairs and motion directions are presented in random order in order to prevent formation of expectations; variations also minimize adaptation effects. 2. Control pattern: same as the saltation pattern, except that SOAs of 700 ms are used for both intervals. Following the presentation of the whole pattern, the participant localizes the three stimuli in their perceived order by clicking on a schematic of the forearm that does not show the positions of the stimulators (see Fig. 4a). Instructions “You will feel three brief taps on your forearm. Please pay attention to where you feel the taps. After the taps, a picture of a forearm will appear on the screen. Please indicate the positions where you felt the taps by clicking on the corresponding part of the forearm. Please always report the taps in the order in which you have felt them”. Practice trials These will ensure that participants understand the instructions (criterion performance can be set) and reduce variability due to initial learning, but practice should be minimized if learning itself is of interest. Data analysis A displacement index can be computed from localization responses: the distance between responses to the first and second tap is expressed as the proportion of the distance between responses to the first and third tap (see Fig. 4c). The expected effect is a larger displacement index for the saltation than for the control condition. Limitations of the method The task may be demanding, especially if stimuli are faint. As explained earlier, participants sometimes report the taps in a spatial order different from the one expected, resulting in an ambiguous response. It may help to make taps distinguishable by a stimulus feature such as duration, vibration frequency, or intensity, keeping in mind that such variations may also influence the saltation effect itself—that is, perceived motion or position (differential stimulus intensity, for example, results in spatial shifts ([17, 69], p. 59). Another solution is to present the stimulus several times before each response (e.g., [48] used 10 repeats).

Creating Tactile Motion

2.2 Filling-in of a Numb Spot with Continuous Motion

83

In both vision and touch, discontinuity in the impressions stimuli make on the sensory surface (proximal stimulus) is a norm: visual objects get occluded, and tactile objects extend between contact points with our fingers. Nevertheless, we perceive external objects and motion events as continuous. In some circumstances, we actively sense stimulus presence in the non-stimulated area, known as (modal) perceptual completion or filling-in (see Note 3.5). Perceptual filling-in has been studied intensely in vision [78], but far less so in touch. Here we describe a method for investigation of motion filling-in in touch. Filling-in with motion was first reported anecdotally when an aperture with two openings separated by an occluder (a split aperture) was placed over the perioral skin and brushed over longitudinally [79]. Recent studies describe filling-in of the 10 cm gap on the forearm [51, 53] and 2 mm gap on the fingertip [54] with motion that quickly skips across the gap, that is, finishes on the one side and almost immediately continues on the other side. Other manners of motion, including slower gap-crossing speeds, also result in fillingin, but take many repeated motion sweeps (unpublished results). Many stimulus parameters can potentially influence whether filling-in occurs and at what rate, including gap size, motion direction, speed of motion across the gap as described above, and more (see Table 3), and most of these variables have not been investigated. The method described below focuses on gap size. It also

Table 3 Stimulus parameters (potential independent variables) and example experimental conditions for the study of filling-in

Parameter A. Extent of the gap (numb spot)

Example experimental conditions narrow wide

Parameter D. Direction of surround motion (and its relatability i.e., potential for grouping by motion)

Example experimental conditions across the gap (strong) along the gap (strong) mixed (weak)

B. Extent of surround stimulation

narrow E. Nature of motion/ attribution of motion wide

trackable object

C. Nature of surround stimulation

motion F. Acceleration across the gap of trackable stimuli taps

high

motion field

zero

84

Tatjana Seizova-Cajic´ et al.

Fig. 5 Operational definitions of filling-in and spilling-over. (a) Filling-in is the sensation mislocalized within the numb spot, when the numb spot is flanked by stimulation; motion sweeps cross the gap. (b) Spilling-over is also a mislocalized sensation, but it occurs in the presence of a single stimulated area. Motion sweeps occur on one side of the gap only, either left or right, within a given experimental run

investigates the importance of surround stimulation as opposed to one-sided stimulation in mislocalization of motion to the non-stimulated area. Different measures—dependent variables—can be used to capture filling-in and related effects. A simple binary judgment regarding the presence of a gap lends itself to a forced-choice task and was used in [54]. If the exposure is repeated or lasts longer, time-tocompletion, i.e., filling-in can also be measured. However, even if the gap is perceived, it may have contracted. Perceived extent is thus a more sensitive measure than a forced-choice task. The dependent variable in the method we describe below is perceived location of motion. It is reported during prolonged, repetitive stimulation and all the above measures can be derived from it. It also captures any location shifts over time. In addition to filling-in—and to help interpret it—we also measure “spilling-over”, that is, mislocalization of the stimulus when only one side of the gap—one skin area—is brushed (Fig. 5). Two half-apertures (which together make up the split aperture) are separately tested to establish the degree of spillingover. Responses from those conditions are superimposed and compared to responses from the split-aperture condition to determine whether their additive effects account for filling-in, or whether filling-in occurs because stimuli on two opposite sides of the gap interact when presented one after the other or simultaneously. 2.2.1

Materials

To create a non-stimulated area surrounded by stimulation, the artificial numb spot, or simply the numb spot, an occluder is placed over the middle of the finger, with flanking areas on both sides exposed to longitudinal brushing (Fig. 6a shows an example occluder that is 3.4 cm wide, with flanking areas of 0.8 cm each). Beveling apertures’ edges minimize the fluctuations in force as the brush moves on and off the skin. Two half-apertures are also used to expose only distal or only proximal skin patch to brushing

Creating Tactile Motion

85

Fig. 6 Photographic and schematic depiction of materials and method for studying of tactile motion filling-in. Top row: 3-D printed apertures used in different experimental conditions to control which skin area is exposed to motion. (a) The split aperture creates a “numb spot” in the middle of the finger, flanked by two brushed areas, accessed via distal and proximal openings; small, circular magnets visible in the photo are used to attach the aperture to the large, fixed frame shown in panel (c); (b) Proximal and distal aperture; Bottom row: The brushing apparatus and reporting method. (c) The brush moves across the aperture attached to a frame, with participant’s hand placed on the opposite side (we sketched bristles in white for clarity). (d) A photo of the participant’s finger (taken before the run) is presented on the computer monitor, with the response grid superimposed on it; the black rectangles are example responses that indicate where the participant felt the motion during a given period of brushing; (e) A photo is taken after the test run to document where the brushing occurred. See procedure for more details

(Fig. 6b). Each aperture attaches to a bigger frame using magnets (Fig. 6c). The aperture and frame together completely occlude the rest of the hand from brushing. Due to individual differences in finger size, apertures will cover different anatomical landmarks in different people. Brushes can be acquired from a hardware or arts store and stiff brushes should be avoided to minimize irritation. They can be further shaped to purpose and need to be replaced with use, as fibers may split and bend. The brush moves back-and-forth along a linear path at a constant speed of approximately 15 cm/s. It is attached to a carrier driven by a stepper motor (e.g., X57-40M, Excitron) mechanically coupled to belt-driven sliding mechanism (Excitron WS70). A custom-made computer program (e.g., in

86

Tatjana Seizova-Cajic´ et al.

LabView) sends commands to a microcontroller via a USB-to-serial adapter to specify direction, steps, steps per second, and ramp up/down slope. The carrier moves along the adjustable track constructed using T-slotted aluminum bars, angle brackets, sliding inserts, and fasteners (e.g., https://8020.com.au/ https:// aluminiumprofile.com.au/). Participants’ localization responses to brushing are measured using a clickable display consisting of n fields (see Note 3.6). In our example, there are 12 fields, each 6 mm long, for a total grid length of 7.2 cm. This provides more space to respond than the maximal 5-cm span of stimulated skin (for an occluder extent of 3.4 cm flanked by two 0.8 cm apertures, as detailed below). A single click on a transparent cell changes its colour to light gray to report a light motion stimulus, and double click makes it black to report that the stimulus feels more intense (Fig. 6d). Bristles can be dipped in the skin-friendly (cosmetic) pigment to mark the contact area on the skin prior to application to the skin; photographs taken after brushing provide clear verification about the dimensions of the area that received stimulation (Fig. 6e). 2.2.2

Methods

Setup The seated participant equipped with headphones playing pink noise faces the monitor, their outstretched left forearm parallel to the mid-sagittal plane, hand pressing against the aperture shown in Fig. 6c. The arm and brushing apparatus are occluded from view using a curtain or divider. During brushing, participants hold their hand actively against the aperture holder, monitored visually by the experimenter. They respond using their right hand, by clicking on the grid presented on the monitor. Design and procedure The two independent variables are extent of the numb spot, created with different occluder sizes (1.4, 2.4, and 3.4 cm) and aperture type, that is, skin area exposed to brushing (three conditions: the split aperture, Fig. 6a; proximal-only and distal-only aperture, Fig. 6b left and right, respectively). The nine experimental conditions (3 extents × 3 apertures) can be completed in three sessions in a fully crossed, repeated-measures design. The order of apertures in each session is counterbalanced across participants, with only one numb spot size done in each session, their order also counterbalanced across participants. Brush either trackable or large brush can be used (Table 3E). Brush sweeps along the finger once per second, at a constant speed of 15 cm/s, and in both directions. Duration of the repetitive brushing should be determined during piloting. It should be long enough to allow filling-in of some gap sizes in most participants (if the aim is to compare filling-in with spilling-over). For example, a 2.4-cm gap fills-in within 4 min if large brush is used (according to our unpublished results), but this will vary with other stimulus parameters (Table 3).

Creating Tactile Motion

87

Participants report perceived location of motion throughout the brushing, using clickable display as described above. Initially, they may clearly feel a gap in stimulation, but it typically shrinks and may completely disappear as brushing progresses. Perceived stimulus intensity can also be reported: a single click on a transparent cell turns it light gray to report a low-intensity touch, and double click, into dark gray for a more intense stimulus. Instructions Participants unaccustomed to perception experiments need to be encouraged to report whatever they experience. Example instructions: “Our interest is in perception and not in your understanding or guessing of the experimental situation; please simply report what you experience even if you cannot explain it”; a color aftereffect, projected on the wall, may be used to illustrate that perception does not always correspond to reality. Example of specific instructions: “Please report the location and intensity of felt motion on your index finger. An image of your hand is shown on the screen; every 30 seconds or so, a 12-field grid will come up, superimposed on the photo of your finger; please indicate where you feel the motion by clicking on the corresponding cell; to indicate stimulus intensity, make the cell lighter or darker by clicking once or twice. At the beginning of each trial, you will feel brushing for 10 s before the grid appears; brushing will continue for 20 s while you click on the grid, after which both the grid and the stimulus will disappear. The brushing will continue, and this cycle will be repeated few more times”. Short practice trials (1–2 min) using the same aperture as in the experimental condition will help familiarize participants with the stimuli and task; participants are encouraged to make small adjustments to their hand position until they are comfortable. The pigment applied to the brush shows area(s) of contact and is checked after practice. Data analysis Raw data are locations of cells clicked on at different points in time. The main derived measure is perceived gap size, which is expected to decrease over time in the split aperture (critical) condition. Its decrease is the operational definition of filling-in with the present method (see Note 3.7). The proximal- and distalaperture controls will be superimposed and compared to the response in the split-aperture condition, as explained earlier. The difference, if any, would be due to the interaction between the two areas when successively stimulated during a single sweep. Other measures can be extracted, including, for example, shift over time in the perceived stimulus location, its overall compression or expansion, or the overall form (topography) of the response.

88

Tatjana Seizova-Cajic´ et al.

Limitations of the method (a) Motion of the brush in the above setup is linear, resulting in variations in pressure due to curved contours of body parts; (b) additional trial-to-trial variations may also occur due to small adjustments in finger position within the aperture; (c) the brushing apparatus does not allow quick changes in stimulation patterns, which limits experimental options (see Note 3.8); (c) direct reporting using click-on-the-grid method is open to expectation bias and attention effects as non-naı¨ve participants may expect filling-in to occur; using naı¨ve observers and complementing studies that rely on direct reporting with those using forced-choice tasks (see Note 3.9) may alleviate this concern.

3

Notes

3.1 Haptic Terminology

The terms haptic or haptics are today often used interchangeably with tactile or touch, especially in the engineering literature represented by the reviews cited in the main text. This differs significantly from classic perception literature, where “haptics” means active or exploratory touch arising from active movements of the hand, as distinct from passive touch received by a body part at rest. Active touch relies not only on cutaneous input but also on proprioception, which includes motor commands [80–83]. The distinction was historically used to stress the importance of active, ecological obtaining of information.

3.2 Systematic Review of Tactile Motion

We conducted a systematic search in Scopus, IEEE, Medline, PsychInfo, Web of Science, and ProQuest Central, with the aim to survey the literature on tactile motion (date of the search: 29th June 2021). We used the following terms: ((“tactile motion” OR “tactile speed” OR “tactile velocity” OR “tactile apparent movement” OR “tactile apparent motion” OR “sensory saltation” OR “cutaneous rabbit” OR “haptic pattern” OR “cutaneous directional sensitivity” OR “vibrotactile patterns”) AND (perc* OR sens*)) OR (“spatiotemporal integration” AND “tactile“) OR ((“affective touch“OR “pleasant touch“OR “gentle touch“) AND (“speed”)). With 102 papers from TSC’s personal library, and following removal of duplicates, the total was 623 papers. Fifty papers were randomly chosen from this database for a quick survey presented in the chapter (the full list of 623 papers, and the chosen 50 papers are given here: https://osf.io/fv9a3/?view_only=2a412cffeb644822 9591f7543008cb57). The 50 papers had to satisfy the below inclusion and exclusion criteria; for each paper that didn’t, another paper was randomly chosen from the full list:

Creating Tactile Motion

89

Inclusion criteria Empirical study in humans or non-human primates; stimuli moved across the skin (stimuli applied to the skin or explored with the skin) or were glued to the skin and moved; study investigated perception of motion stimulus (e.g., speed, direction, location, extent, pleasantness) in basic or applied context (e.g., VR, sensory substitution) OR perception of self-motion (vection) induced by tactile motion, in basic or applied context, OR neural mechanisms for perception of motion. Exclusion criteria Study focused on other aspects of perception (e.g., pain, limb ownership); not a first-order empirical study but review, opinion, or other. 3.3 Random-dot Patterns

Kuroki and Nishida [83] described similarities between visual random-dot kinematograms (RDKs) and their tactile analogs: “Visual RDKs have been widely used to investigate the visual motion processing in the brain. One of the advantages of using RDKs is density controllability. (When translated into touch, density is the number of possible points of contact on the skin.) With dense stimuli, trackable features are masked; thus, high-level feature tracking becomes difficult. In other words, increasing dot density creates a bias towards lowlevel motion sensing (Braddick, 1974*). Another advantage is lifetime controllability. Lifetime is defined as the number of successive image frames in which each dot appears and moves before it extinguishes. (When translated into touch, lifetime is the number of individual actuators successively representing the same moving dot at different times.)” * [84].

3.4 Controlling Tactile Stimuli

The intensity, duration, and frequency (in case of vibratory stimuli) of tactile stimuli need to be controlled with hardware that offers precise timing. Ideally, time-critical processes can be performed by specialized hardware that is built for the given application and uses a microcontroller. In many situations, tactile stimuli need to be coordinated with other events, for example other stimuli, or an experimental task running on a computer. When using a computer to control stimuli, it is important to use software that allows precise timing and low-latency communication with the hardware. Typical applications are programming languages such as Matlab (www. mathworks.com), Python (www.python.org), LabView (www.ni. com), or specialized stimulus presentation software such as Presentation (www.neurobs.com). If the computer is configured correctly and background processes are limited to a minimum, precision of timing is usually at the sub-millisecond level. However, interfaces to control external stimulators, such as soundcards or USB-to-serial converters, can have longer and variable latencies that compromise timing. Therefore, the use of a low-latency device, or special data acquisition devices

90

Tatjana Seizova-Cajic´ et al.

(DAQs) is recommended. Depending on the device, both digital (on/off) or analog signals can be written and read, and many devices include a digital-to-analog converter (DAC) that allows creation of waveforms, for example, to drive vibratory stimuli. Some manufacturers of such devices are Labjack (www.labjack. com) or National Instruments (www.ni.com). National Instruments also provides cards for a computer’s PCI bus with parallel ports (digital in and out ports) that can be used to trigger stimuli, other devices (e.g., an EEG system), or read events (e.g., button presses) with precise timing. These devices can also be used to control switchboxes to route the signal to several stimulators or electrodes. The software configuration and hardware need to be optimized for the given purpose and computer load kept at a low level to ensure best timing. Stimulus definitions in particular must not overwhelm the update rate of the communication via the DAQ. While digital ports are usually fast, the DACs update rate is limited when controlled by a computer. This can lead to problems when, for example, attempting to define a sophisticated waveform such as a sine wave using computer commands. Many devices provide “stream out” configurations where the values defining the waveform are stored in a buffer on the hardware device itself and do not require real-time update of the DAC which improves timing. A collection of functions to control up to 10 stimulators (as used in the setup described for sensory saltation) may be useful for a range of applications and can be downloaded from this software repository: https://github.com/xaverfuchs/pytact. 3.5 Amodal and Modal Completion

We can distinguish between two types of completion: amodal and modal. In amodal completion, the object or motion event are perceived as continuous without being seen or felt in the nonstimulated area [78]. An example is partial occlusion of a car by a tree: the car is perceived as a whole, but its color and shape are not seen in the place of occlusion. In touch, we feel uninterrupted, continuous surface when we touch it with outstretched fingertips (see Sections 11 and 34 in [85]), but we don’t sense surface properties in-between the fingers. By contrast, in modal completion, at least some features of the stimulus surrounding the non-stimulated area are perceived within the area (described in [78], p. 228 as “the illusory perception of a feature in a region where it is physically absent”). The method we describe results in modal completion.

3.6 Number of Response Options

The optimal number of clickable response cells depends on the time allowed for the response, degree of precision required, and amount of practice the participant has had. Too many cells may result in incomplete response and tire the participant, and too few will not give a sensitive measure. A practice run makes the task easier, but may result in a perceptual change, which the experiment is

Creating Tactile Motion

91

supposed to capture. The practice should therefore be brief, with a break between the practice and test. 3.7 Definitions of “Filling-in”

Other operational definitions can be had with different control conditions. For example, perceived gap size when brushing laterally across the split aperture can be used as the baseline.

3.8 Discrete Versus Continuous Stimuli

Some of these limitations illustrate why apparent motion using discrete stimuli is a more popular method for studying tactile motion (Table 2). Unlike a brush moving along a straight path, an array of small vibrators or other tactors attached to the skin would follow the limb curvature and stimulus intensity across identical tactors would be constant. The motion pattern is also easier to alter across experimental conditions using discrete motion devices, and some stimulus patterns, such as taps (Table 3C), can hardly be created using brushes. However, brushing also has important advantages: it is a gentle motion stimulus that results in good motion sensation and can be applied repeatedly without excessive adaptation. Vibrating tactors that make up commonly used discrete devices are much stronger stimuli that create traveling waves and would thus encroach onto the intended numb spot. How much vibration spreads depends on the stimulator size, spacing, and intensity of vibration. Therefore, more subtle tactors such as piezoelectric stimulators, densely packed to achieve smooth motion, might be a viable alternative to brushing. The finest discrete devices [61, 64, 65] can be used on the fingertip (and have been used - see [54]) but they are not widely available.

3.9 Direct Report Versus Forced-choice Responses

Filling-in on the fingertip was studied using Latero (Fig. 2f) in a forced-choice task [54]. The 10 mm motion contained a 2 mm gap, crossed using a range of gap-crossing times. Participants reported whether motion sweeps, presented one at a time, contained a gap or not. However, the forced-choice task does not provide information about where motion is perceived, which the method described above of clicking on the grid does. There is, in other words, a trade-off between simplifying the task to better control for potential bias and a wealth of information we can obtain. As often is the case, it seems best to combine multiple approaches to the study of motion filling-in.

References 1. Hall GS, Donaldson HH (1885) Motor sensations on the skin. Mind 10(40):557–572 2. Benussi V (1913) Kinematohaptische Erscheinungen (Kinematohaptic illusion). Arch f d ges Psych 29:385–388

3. Benussi V (1914) Kinematohaptische Scheinbewegung und Auffassungsumformung (Kinematohaptic apparent motion and perceptual change), Ber. ii. d. VI Kong. f. exp. Psych. Gottingen: 31–35

92

Tatjana Seizova-Cajic´ et al.

4. Burtt H (1917) Tactual illusions of movement. J Exp Psychol 2(5):371–385 5. Whitchurch A (1921) The illusory perception of movement on the skin. Am J Psychol 32(4): 472–489 6. Thalman W (1922) The after-effect of movement in the sense of touch. Am J Psychol 33: 268–276 7. Essick GK (1998) Factors affecting direction discrimination of moving tactile stimuli. In: Morley JW (ed) Advances in psychology 127. Neural Aspects of Tactile Sensation, NorthHolland, p 1–54 8. Pack CC, Bensmaia SJ (2015) Seeing and feeling motion: canonical computations in vision and touch. PLoS Biol 13(9):e1002271 9. Pei YC, Bensmaia SJ (2014) The neural basis of tactile motion perception. J Neurophysiol 112: 3023–3032 10. Morrison I, Lo¨ken LS, Olausson H (2010) The skin as a social organ. Exp Brain Res 204:305– 314 11. Moscatelli A, Nimbi FM, Ciotti S, Jannini EA (2020) Haptic and somesthetic communication in sexual medicine. Sex Med Rev. https://doi.org/10.1016/j.sxmr.2020. 02.003 12. Shull PB, Damian DD (2015) Haptic wearables as sensory replacement, sensory augmentation and trainer - a review. J Neuroengi Rehab 12:59 13. Culbertson H, Schor S, Okamura (2018) Haptics: the present and future of artificial touch sensation. Annu Rev Contr Robot Auton Syst 1(1):385–410 14. Al-Handarish Y, Omisore MO, Igbe T, Han S, Li H, Du W, Zhang J, Wang L (2020) A survey of tactile-sensing systems and their applications in biomedical engineering. Adv Mater Sci Eng 2020:Article ID 4047937 15. Hutmacher F (2019) Why is there so much more research on vision than on any other sensory modality? Front Psychol 10:2246 16. Hartwig V, Carbonaro N, Tognetti A, Vanello N (2017) Systematic review of fMRI compatible devices: design and testing criteria. Ann Biomed Eng 45:1819–1835 17. Puckett AM, Sanchez Panchuelo RM (2023) Imaging somatosensory cortex: human functional magnetic resonance imaging (fMRI), present volume 17. Bekesy G (1967) Sensory inhibition. Princeton UniversityPress, Princeton 18. Sherrick CE, Rogers R (1966) Apparent haptic movement. Percept Psychophys 1:175–180 19. Kirman JH (1974) Tactile apparent movement: the effects of interstimulus onset interval and stimulus duration. Percept Psychophys 15:1–6

20. Kirman J (1975) The effect of number of stimulators on the optimal interstimulus onset interval in tactile apparent movement. Percept Psychophys 17:263–267 21. Gardner EP, Sklar BF (1994) Discrimination of the direction of motion on the human hand: a psychophysical study of stimulation parameters. J Neurophysiol 71(6):2414–2429 22. Cholewiak RW, Collins AA (2000) The generation of vibrotactile patterns on a linear array: influences of body site, time, and presentation mode. Percept Psychophys 62:1220–1235 23. Olausson H, Norrsell U (1993) Observations on human tactile directional sensibility. J Physiol 464:545–559 24. Seizova-Cajic T, Karlsson K, Bergstrom S, McIntyre S, Birznieks I (2014) Lateral skin stretch influences direction judgments of motion across the skin. In: Haptics: neuroscience, devices, modeling, and applications, pp 425–431 25. De´peault A, Meftah E-M, Chapman CE (2008) Tactile speed scaling: contributions of time and space. J Neurophysiol 99:1422–1434 26. Dallmann CJ, Ernst MO, Moscatelli A (2015) The role of vibration in tactile speed perception. J Neurophysiol 114:3131–3139 27. Delhaye BP, O’Donnell MK, Lieber JD, McLellan KR, Bensmaia SJ (2019) Feeling fooled: texture contaminates the neural code for tactile speed. PLoS Biol 17(8):e3000431 28. Lederman SJ, Klatzky RL (2009) Haptic perception: a tutorial. Atten Percep Psychophys 71:1439–1459 29. Kuroki S, Watanabe J, Mabuchi K, Tachi S, Nishida S (2012) Directional remapping in tactile inter-finger apparent motion: a motion aftereffect study. Exp Brain Res 216:311–320 30. Badde S, Heed T (2016) Towards explaining spatial touch perception: weighted integration of multiple location codes. Cogn Neuropsychol 33(1–2):26–47 31. Liaci E, Bach M, Tebartz van Elst L, Heinrich SP, Kornmeier J (2016) Ambiguity in tactile apparent motion perception. PLoS One 11: e0152736 ˜ o´n E, Longo MR (2019) A con32. Tame` L, Azan ceptual model of tactile processing across body features of size, shape, side, and spatial location. Front Psychol 10:291 33. Haladjian HH, Anstis S, Wexler M, Cavanagh P (2020) The tactile quartet: comparing ambiguous apparent motion in tactile and visual stimuli. Perception 49:61–80 34. Shao Y, Hayward V, Visell Y (2016) Spatial patterns of cutaneous vibration during wholehand haptic interactions. Proc Nat Acad Sci 113(15):4188–4193

Creating Tactile Motion 35. Reardon G, Shao Y, Dandu B, Frier W, Long B, Georgiu O, Visell Y (2019) Cutaneous wave propagation shapes tactile motion: evidence from air-coupled ultrasound. IEEE Haptics Conf (WHC), pp 628–633 36. Greenspon CM, McLellan KR, Lieber JD, Bensmaia SJ (2020) Effect of scanning speed on texture-elicited vibrations. J R Soc Interface 17:20190892 37. Gallace A, Spence C (2011) To what extent do Gestalt grouping principles influence tactile perception? Psychol Bull 137:538–561 38. Lerner EA, Craig JC (2002) The prevalence of tactile motion aftereffects. Somatosens Mot Res 19:24–29 39. Watanabe J, Hayashi S, Kajimoto H, Tachi S, Nishida S (2007) Tactile motion aftereffects produced by appropriate presentation for mechanoreceptors. Exp Brain Res 180:577– 582 40. McIntyre S, Birznieks I, Vickery RM, Holcombe AO, Seizova-Cajic T (2016) The tactile motion aftereffect suggests an intensive code for speed in neurons sensitive to both speed and direction of motion. J Neurophysiol 115(3):1703–1712 € 41. von Rausch E (1960) Uber kinaesthetische nachwirkungen. Psychol Beitr€age 5:232–320 42. McIntyre S, Holcombe AO, Birznieks I, Seizova-Cajic T (2012) Tactile motion adaptation reduces perceived speed but shows no evidence of direction sensitivity. PLoS One 7: e45438 43. Craig JC (1996) Interference in identifying tactile patterns: response competition and temporal integration. Somatosens Mot Res 13: 199–213 44. Craig JC (2006) Visual motion interferes with tactile motion perception. Perception 35:351– 367 45. Craig JC (1993) Anomalous sensations following tactile stimulation. Neuropsychologia 31: 277–291 46. Israr A, Poupyrev I (2011) Tactile brush: drawing on skin with a tactile grid display. Proceed SIGCHI Conf Human Fact Comput Syst. pp 2019–2028 47. Geldard F, Sherrick CE (1972) The cutaneous “rabbit”: a perceptual illusion. Science 178: 178–179 48. Kilgard MP, Merzenich M (1995) Anticipated stimuli across skin. Nature 373:663 49. Trojan J, Stolle AM, Carl AM, Kleinbo¨hl D, Tan HZ, Ho¨lzl R (2010) Spatiotemporal integration in somatosensory perception: effects of sensory saltation on pointing at perceived positions on the body surface. Front Psychol 1:206

93

50. Whitsel BL, Franzen O, Dreyer DA, Hollins M, Young M, Essick GK, Wong C (1986) Dependence of subjective traverse length on velocity of moving tactile stimuli. Somatosens Res 3:185–196 51. Seizova-Cajic T, Taylor JL (2014) Somatosensory space abridged: rapid change in tactile localization using a motion stimulus. PLoS One 9(3):e90892 52. Merz S, Deller J, Meyerhoff HS, Spence C, Frings C (2019) The contradictory influence of velocity: representational momentum in the tactile modality. J Neurophysiol 121:2358– 2363 53. Nguyen EHL, Taylor JL, Brooks J, SeizovaCajic T (2016) Velocity of motion across the skin influences perception of tactile location. J Neurophysiol 115:674–684 54. Kaneko S, Kajimoto H, Hayward V (2018) A case of perceptual completion in Spatiotemporal tactile space. In: Prattichizzo D, Shinoda H, Tan HZ, Ruffaldi E, Frisoli A (eds) BT – haptics: science, technology, and applications. Springer International Publishing, pp 49–57 55. Weber EH (1834/1856/1978) The sense of touch. Translated by HE Ross & DJ Murray. Academic Press 56. Longo MR, Haggard P (2011) Weber’s illusion and body shape: anisotropy of tactile size perception on the hand. JEP: HPP 37:720– 726 57. Bender MB, Stacy C, Cohen J (1982) Agraphesthesia. A disorder of directional cutaneous kinesthesia or a disorientation in cutaneous space. J Neurol Sci 53:531–555 58. Edin BB, Essick GK, Trulsson M, Olsson KA (1995) Receptor encoding of moving tactile stimuli in humans. I. Temporal pattern of discharge of individual low-threshold mechanoreceptors. J Neurosci 15:830–847 59. Hsu YC, Yeh CI, Huang JJ, Hung CH, Hung CP, Pei YC (2019) Illusory motion reversal in touch. Front Neurosci 13:9 60. Arafsha F, Zhang L, Dong H, Saddik AE (2015) Contactless haptic feedback: State of the art. IEEE international symposium on haptic, audio and visual environments and games, HAVE 2015 61. Wang Q, Hayward V (2009) Biomechanically optimized distributed tactile transducer based on lateral skin deformation. Int J Robot Res 29:323–335 62. Shimizu Y, Wake T (1982) Tactile sensitivity to two types of stimulation: continuous and discrete shifting of a point stimulus. Percept Mot Skills 54(3):1111–1118

94

Tatjana Seizova-Cajic´ et al.

63. Gardner EP, Sklar BF (1994) Discrimination of the direction of motion on the human hand: a psychophysical study of stimulation parameters. J Neurophysiol 71:2414–2429 64. Bliss JC, Katcher MH, Rogers CH, Shepard RP (1970) Optical-to-tactile image conversion for the blind. IEEE Trans Man-Machine Sys MMS-11:58–65 65. Killebrew J, Bensmaia S, Dammann J, Denchev P, Hsiao S, Craig JC, Johnson KO (2007) A dense array stimulator to generate arbitrary spatio-temporal tactile stimuli. J Neurosci Methods 161:62–74 66. Nunez CM, Williams SR, Okamura A, Culbertson H (2019) Understanding continuous and pleasant linear sensations on the forearm from a sequential discrete lateral skin-slip haptic device. IEEE Trans Haptics 12:414–427 67. Pasquero J (2006) Survey on communication through touch. Technical report: TR-CIM 06.04. Center for Intelligent Machines. McGill University 68. MacLean KE, Hayward V (2008) Do it yourself haptics: part II. IEEE Robot Autom Mag 15:104–119 69. Geldard FA (1975) Sensory saltation: Metastability in the perceptual world. Lawrence Erlbaum 70. Flach R, Haggard P (2006) The cutaneous rabbit revisited. J Exp Psychol: Human Perc Perform 32(3):717 71. Eimer M, Forster B, Vibell J (2005) Cutaneous saltation within and across arms: a new measure of the saltation illusion in somatosensation. Atten Perc Psychophys 67(3):458–468 72. Blankenburg F, Ruff CC, Deichmann R, Rees G, Driver J (2006) The cutaneous rabbit illusion affects human primary sensory cortex somatotopically. PLoS Biol 4(3):e69 73. Warren JP, Santello M, Helms Tillery SI (2010) Electrotactile stimuli delivered across fingertips inducing the Cutaneous Rabbit Effect. Exp Brain Res 206(4):419–426

74. Miyazaki M, Hirashima M, Nozaki D (2010) The “Cutaneous Rabbit” hopping out of the body. J Neurosci 30(5):1856–1860 75. Martel M, Fuchs X, Trojan J, Gockel V, Habets B, Heed T (2021) Illusory tactile movement crosses arms and legs is coded in external space. BioRxiv, 20210520445020 76. Trojan J, Heil M, Maiho¨fner C, Ho¨lzl R, Kleinbo¨hl D, Flor H, Benrath J (2014) Spatiotemporal integration of tactile patterns along and across fingers. Neuropsychologia 53:12– 24 77. Peirce J, Gray JR, Simpson S, MacAskill M, Ho¨chenberger R, Sogo H, Kastman E, Lindeløv JK (2019) PsychoPy2: experiments in behavior made easy. Behav Res Methods 51(1):195–203 78. De Weerd P (2006) Perceptual filling-in: more than the eye can see. Prog Brain Res 154:227– 245 79. Essick GK, Franzen O, Mcmillian A, Whitsel B (1991) Utilization of temporal and spatial cues to judge the velocity and traverse length of a moving tactile stimulus. In: Franzen et al (eds) Information processing in the somatosensory system, pp 341–352 80. Gibson J (1966) The senses considered as perceptual systems. Houghton Mifflin, Boston 81. Lederman S, Klatzky R (2009) Haptic perception: a tutorial. Atten Percept Psychophys 71: 1439–1459 82. Carello C, Turvey MT (2017) Useful dimensions of haptic perception: 50 years after the senses considered as perceptual systems. Ecol Psychol 29(2):95–121 83. Kuroki S, Nishida Y (2021) Motion direction discrimination with tactile Random-Dot Kinematograms. Iperception 12:1–19 84. Braddick O (1974) A short-range process in apparent motion. Vis Res 14(7):519–527 85. Katz D (1925/1989) The world of touch. Lawrence Erlbaum Associates

Chapter 5 Measuring Tactile Distance Perception Matthew R. Longo Abstract Illusions of tactile distance have been studied since the start of scientific research on the sense of touch in the nineteenth century. In the past 15 years, these illusions have become increasingly popular among researchers due to their connections with basic aspects of somatosensory neurophysiology, higher-level aspects of mental body representation, and relation to clinical disorders. This chapter will discuss methods for measuring tactile distance perception, focusing on two broad classes of methods. One type of method involves making estimates of the distance between a single pair of touches applied to the skin. The other method involves making a judgment about the relative distance between two such pairs. These methods can be applied to a range of experimental designs, body parts, and experimental conditions. Key words Touch, Distance perception, Size perception, Anisotropy

1

Introduction Ernst Weber’s studies in the nineteenth century were among the first systematic investigations of the sense of touch [1]. Among many seminal observations, Weber observed an intriguing tactile illusion which now bears his name (Weber’s illusion). As he moved the two points of a compass across his skin, it felt to him like the distance between them increased as he moved them from a region of relatively low tactile sensitivity to a region of higher sensitivity. Several subsequent studies have confirmed this general pattern that touches feel farther apart on skin surfaces with high than with low tactile spatial acuity [2–4]. Analogous illusions have also been reported for stimuli in different orientations on single skin surfaces, with distances generally being perceived as farther apart when oriented across the width of the body [5–7]. Such anisotropies have been reported extensively on the hand [6], but also on other body parts including the forearm [5], thigh [5], shin [8], foot [9], and face [10]. The only place where this effect does not seem to

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_5, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

95

96

Matthew R. Longo

appear is on the torso, with no apparent anisotropy on the belly [11], and two recent studies finding a reversed effect on the lower back [12, 13]. Longo and Haggard [6] related both Weber’s illusion and tactile distance anisotropies to the geometry of the receptive fields (RFs) of neurons in the somatosensory cortex. RFs are smaller on highly sensitive skin surfaces compared to less sensitive skin surfaces [14], and are also generally oval-shaped with the long axis of the RF running parallel to the long axis of the limbs. This is found in the somatosensory cortex [15], the spinal cord [16], and even in individual peripheral afferent fibers [17]. Correspondingly, several studies have reported that tactile spatial acuity [1, 18] and the precision of tactile localization [19, 20] are higher across the width of the limbs than along their length. Consistently, adaptation aftereffects for perceived tactile distance have been reported [21], which show selectivity for a range of characteristics (e.g., orientation, location, skin surface) suggesting that they arise from relatively early stages of somatosensory processing. At the same time, tactile distance perception also appears linked to higher-level aspects of body perception, being modulated by body size illusions [3, 22, 23] (see also Chapter 13, this volume), tool-use [24–26], and categorical segmentation of the body into discrete parts [27–29]. Similarly, the baseline distortions in tactile distance perception in which distances across body-part width are overestimated are similar to perceptual distortions of body size measured using a variety of other tasks [30–32]. Finally, several recent studies have found abnormalities of tactile distance perception in clinical disorders, such as anorexia nervosa [33] and obesity [34], low-back pain [35], as well as after surgical elongation of the arm [36].

2

Materials A range of stimuli have been used to measure perceived tactile distance, including wooden, metal, or plastic sticks [3, 5, 6, 28, 29, 37–40], calipers [8, 27, 33, 34, 41, 42], von Frey hairs [7, 43], solenoid tappers [22], electric shocks [44], vibrotactile stimuli [2, 12], air puffs [45], and a laser which selectively stimulates nociceptive afferents [43]. While no research to my knowledge has directly compared these stimuli, it is worth noting that broadly comparable results (e.g., Weber’s illusion) are apparent across a range of stimulus types. For example, anisotropies of similar magnitude have been found on the hand dorsum measured with sticks [6], von Frey hairs [7], and air puffs [45]. Similarly, comparable anisotropy in the opposite direction has recently been found on the lower back using both vibrotactile stimuli [12] and sticks [13]. Intriguingly, however, one study that compared tactile (von

Tactile Distance Perception

97

Frey hairs) with nociceptive (infrared laser) distance perception found that participants were unable to make meaningful distance judgments of nociceptive stimuli at all [43]. Many studies have used simple verbal responses, either of judged size [2, 5, 7, 37, 43, 46] or of which of two stimuli is perceived as bigger [3, 6, 21, 22, 47], or manual entry of numbers [44], which do not require any specialized measurement equipment. Other approaches, however, do require other equipment. For example, Tame` and colleagues [45] used a visual comparison procedure in which participants manually adjusted the length of a line presented on a monitor to match the perceived distance between two touches. The script was controlled by a custom MATLAB script using the Psychophysics toolbox [48]. Some other studies [8, 33, 34, 41, 49] have used kinesthetic judgments in which participants use two fingertips to match the perceived distance between two touches. While this can be done using paper-and-pencil and a ruler [34, 49], most studies have used a more automated procedure in which distances are measured using a touchscreen tablet computer [8, 33, 41].

3

Methods Methods for measuring tactile distance perception fall into two broad families, those involving estimating the distance between a single pair of touches (size estimation methods) and those involving comparing the relative distance between two different pairs (two-interval forced-choice, 2IFC methods, see Chapter 1, this volume for discussion of different experimental designs).

3.1 Size Estimation Methods

The first set of methods involves size estimation of a single tactile distance, which can involve four procedures:

3.1.1 In Magnitude Estimation

In which participants give a verbal estimate of distance using an arbitrary magnitude scale [2, 5]. Green [5], for example, asked participants to respond with “a number that reflected the apparent distance between the two stimuli” (pg. 316), while explicitly avoiding mapping these numbers onto known metric units such as inches or centimeters.

3.1.2 In Absolute Estimation

In which participants give verbal [7, 11, 27, 43, 46, 50] or written [44] estimates of the distance between two touches using an absolute metric scale (e.g., cm or inches). It is important to note that absolute over- or under-estimation of tactile distance using this method could be due to misrepresentation of the measuring unit, rather than an actual tactile distortion. Thus, in my view, inferences using this method should be restricted to comparisons of different

98

Matthew R. Longo

experimental conditions, not to veridical size. One approach to addressing this issue would be to have a visual ruler present showing the actual size of a cm. Even in this case, however, interpreting absolute values is potentially problematic. 3.1.3 In Visual Comparison

In which participants compare a tactile distance with a visual comparison stimulus [45]. In the recent study by Tame` and colleagues [45], the stimuli were presented on a computer monitor, but they could also conceivably be printed on sheets of paper or even be physical 3-D objects. Compared to absolute estimation, this provides a more valid measure of over- or under-estimation of a single stimulus, though it is important to keep in mind that deviations from veridical judgments could just as well reflect misperception of visual as of tactile stimuli. For example, different estimates will likely be obtained if the visual comparison stimulus is oriented vertically versus Horizontally, due to the well-known visual horizontal-vertical illusion [51].

3.1.4 In Kinesthetic Estimation

In which participants use the distance between their thumb and index fingertips to match the perceived distance between two touches [8, 33, 34, 41, 49]. In these studies, the kinesthetic judgments have been made with a hand that was not being stimulated. As with absolute estimation and visual comparison, deviations from veridical judgments could reflect biases in kinesthetic perception, just as much as tactile distance perception.

3.2 Size Estimation Analysis

Whichever of these estimation methods is used, different analysis approaches can be employed:

3.2.1

Linear Regression

Some studies have used linear regression to assess how perceived tactile distance relates to actual tactile distance [5, 43]. Different skin surfaces, or different orientations on a single surface, can be compared either in terms of slope or y-intercept.

3.2.2

ANOVA

Other studies using similar designs have used analysis of variance (ANOVA) approaches to analyze data [2, 11, 46]. For example, the top row of Fig. 1 shows data on the hand and belly [11]. Because these skin surfaces have very different two-point discrimination thresholds [52, 53], different actual tactile distances needed to be used (Fig. 1, top left panel), leading to deviation from a purely factorial design. Re-expressing perceived distance as overestimation as a proportion of actual distance (Fig. 1, top right panel) can facilitate analysis and allow comparison of skin surfaces which require different absolute distances due to differences in 2-PDT (see Note 4.1). The results show a clear anisotropy on the hand, but not on the belly, as well as relative overestimation on the (relatively sensitive) hand compared to the (relatively insensitive) belly (i.e., Weber’s illusion).

Tactile Distance Perception

99

Fig. 1 Methods for analyzing results from size estimation procedures. Top row: Results (N = 37) showing tactile distance anisotropy on the hand and belly [11]. Results were analyzed using ANOVA assessing judged size as a function of actual size and orientation (top left), and assessing overestimation as a percentage of actual size (top right; error bars show SEM). Middle row: Use of multidimensional scaling (MDS) to reconstruct

100

Matthew R. Longo

3.2.3 Multidimensional Scaling

Two recent studies [7, 45] have used multidimensional scaling (MDS) to reconstruct perceptual maps of tactile space. MDS is a statistical procedure, akin to principal components analysis, positioning items in a multi-dimensional space based on a matrix of the pairwise distances or dissimilarities between items, such that the pairwise distances between points match the distance matrix as closely as possible [54, 55]. Applied to tactile distance perception, a fixed set of locations on the skin is stimulated and across trials tactile distance estimates are obtained from each pair of locations, producing a full perceptual distance matrix. MDS applied to this distance matrix produces coordinates in 2-D (or other dimensionality, if so desired) space. The middle panel of Fig. 1 shows results from the study of Longo and Golubova [7] which used MDS to reconstruct the tactile space of the hand dorsum (Fig. 1, middle left panel). Overall distortion in these maps was quantified by finding the deformation applied to an idealized square grid that minimized the dissimilarity (quantified as the Procrustes distance) with each perceptual map (Fig. 1, middle right panel). Tame` and colleagues [45] also applied the same logic to representational dissimilarity matrices measured using fMRI to compare perceptual and neural maps of tactile space.

3.2.4 Computational Models

Finally, a recent study by Fiori and Longo [37] presented stimuli at a range of orientations and used a simple computational model to quantify the magnitude and orientation of “stretch” of tactile space. The basic idea is that if anisotropy on a skin surface reflects a geometrically simple stretch of tactile space, perceived distance as a function orientation should show a sinusoidal function across a range of orientations (Fig. 1, bottom row). This approach uses least-squares regression to fit a three-parameter model to individual participant data, allowing the magnitude and orientation of stretch to be quantified. For example, the bottom right panel of Fig. 1 shows the orientation of maximal stretch for each of the 25 participants on the hand dorsum.

3.3 Forced-Choice Methods

Whereas the size estimation methods described so far involve making judgments about single stimuli, another set of methods ask participants to compare the relative size of two pairs of touches, presented either on two different skin surfaces or in two

ä Fig. 1 (continued) perceptual maps of the tactile space of the hand dorsum [7] (middle left; N = 12). Overall distortion in these maps was quantified by identifying the deformation of a square grid that minimized the dissimilarity (i.e., Procrustes distance; middle right; shaded area shows SEM). Bottom row: The model used by Fiori and Longo [37] to assess whether tactile distance illusions reflect a geometrically simple stretch of tactile space (bottom left; N = 25; error bars show SEM). This approach allows the orientation of maximal anisotropy to be estimated in a data-driven way for each participant (bottom right)

Tactile Distance Perception

101

Fig. 2 An example of a psychometric function fit to 2IFC data (N = 18; error bars show SEM; dotted vertical line shows the point of subjective equality; from [50])

orientations on a single skin surface. In all cases, this has involved sequential, rather than simultaneous, presentation of the pairs of stimuli. Some studies have simply quantified the percentage of trials on which one type of stimuli is judged as larger [3, 22]. Most studies, however, have adopted some form of the method of constant stimuli, using psychometric functions to quantify biases in the perceived tactile distance [6, 10, 21, 25, 28, 29, 38, 39, 47]. Figure 2 shows a typical example, taken from the “together” condition in Experiment 1 of [50]. On each trial, two tactile distances were applied sequentially, one oriented with the mediolateral (“across”) hand axis the other with the proximo-distal (“along”) axis, and the participant judges which one feels larger (i.e., by saying “first” or “second”). Across trials, five different pairs of distances were used, varying the relative size of the stimuli in the two orientations, according to the method of constant stimuli. The proportion of trials on which the across stimuli were judged as larger was calculated as a function of the ratio of the across and the along stimuli. The psychometric function was modeled using a cumulative Gaussian curve fit using maximum likelihood estimation using the Palamedes MATLAB toolbox [56]. The mean of this curve indicates the point of subjective equality (PSE), that is the ratio between the across and the along stimuli where the participant is equally likely to say that each orientation is bigger. The results shown in Fig. 2 indicate a typical anisotropy on the hand dorsum, as the PSE corresponds to a ratio between the across and the along stimuli less than 1, meaning that the along stimulus needs to be bigger than the across stimulus for them to be judged as being the same size.

102

4

Matthew R. Longo

Notes

4.1 The Two-point Discrimination Task

As noted above (see Subheading 3.2.2), one issue that frequently comes up in designing studies of tactile distance perception is the relation between the actual distances applied and the two-point discrimination threshold (2PDT) on that skin surface. While the 2PDT has been criticized as a measure of tactile acuity [57] (see Chapter 1, this volume), in this context what is relevant is whether or not the participant experiences one point or two. If the participant only feels a single point, it is not sensible to ask them to judge how far apart the stimuli felt. In many studies, participants are asked to assume that if they feel just one point to assume that it was a small distance [6], and in others to consider the spatial extent of that single point [5]. In some studies using verbal responses, they are explicitly told that they can give a response of “0 cm” [37]. This, however, does not necessarily solve the problem, and such trials (whether or not they are included in the analysis) can potentially distort the pattern of results. To address this issue, researchers can consult studies that have measured 2PDT across the body [52, 53]. Ideally the smallest tactile distance applied should be larger than the average 2PDT on that skin surface, although since the numbers reported in those papers are averages, even this does not guarantee that participants will feel two touches on all trials. Another point is that while largescale studies of 2PDT [52, 53] have assessed the perception of stimuli in a single orientation, it has been known since Weber’s work that 2PDT varies with the orientation of stimuli [1]. Thus, stimuli which are felt as two distinct touches in one orientation will not necessarily be felt the same way in another orientation on that same surface. On some body parts, this can leave a relatively narrow range of usable stimuli that are both large enough to be felt as two points and small enough to actually fit on the skin for all participants. On the palm and dorsum of the hand and on the arms, for example, we have found that a range of 2–4 cm is appropriate, with stimuli less than 2 cm commonly felt as a single point and stimuli more than 4 cm not fitting on some participants’ bodies.

4.2 Testing Different Parts of the Body

Another consequence of varying 2PDT across the body is that the stimuli appropriate for testing one body part may not be appropriate for other body parts. For example, as shown in the top left panel of Fig. 1, a recent study comparing anisotropy of tactile distance on the hand and belly [11] used different sets of stimuli on each body part for exactly this reason. While it would obviously be preferable to use identical stimuli on each skin surface, the combination of different sensitivity and different size of skin regions makes this impossible. Analogous problems arise in many types of psychophysical studies, for example comparing different parts of the retina in

Tactile Distance Perception

103

vision, or different frequencies in audition. One approach to dealing with this violation of a fully factorial experimental design is to re-express each judgment in terms of overestimation of actual distance as a percentage of actual distance, as shown in the top right panel of Fig. 1.

References 1. Weber EH (1834) De subtilitate tactus. In: Ross HE, Murray DJ (eds) E. H. Weber on the tactile senses. Academic Press, London, pp 21–128 2. Cholewiak RW (1999) The perception of tactile distance: influences of body site, space, and time. Perception 28:851–876. https://doi. org/10.1068/p2873 3. Taylor-Clarke M, Jacobsen P, Haggard P (2004) Keeping the world a constant size: object constancy in human touch. Nat Neurosci 7:219–220. https://doi.org/10.1038/ nn1199 4. Miller LE, Longo MR, Saygin AP (2016) Mental body representations retain homuncular shape distortions: evidence from Weber’s illusion. Conscious Cogn 40:17–25. https://doi. org/10.1016/j.concog.2015.12.008 5. Green BG (1982) The perception of distance and location for dual tactile pressures. Percept Psychophys 31:315–323. https://doi.org/10. 3758/BF03202654 6. Longo MR, Haggard P (2011) Weber’s illusion and body shape: anisotropy of tactile size perception on the hand. J Exp Psychol Hum Percept Perform 37:720–726. https://doi. org/10.1037/a0021921 7. Longo MR, Golubova O (2017) Mapping the internal geometry of tactile space. J Exp Psychol Hum Percept Perform 43:1815– 1827. https://doi.org/10.1037/xhp0000434 8. Stone KD, Keizer A, Dijkerman HC (2018) The influence of vision, touch, and proprioception on body representation of the lower limbs. Acta Psychol 185:22–32. https://doi.org/10. 1016/j.actpsy.2018.01.007 9. Manser-Smith K, Tame` L, Longo MR (2021) Tactile distance anisotropy on the feet. Atten Percept Psychophys 83(8):3227–3239 10. Longo MR, Amoruso E, Calzolari E et al (2020) Anisotropies of tactile distance perception on the face. Atten Percept Psychophys 82: 3636–3647. https://doi.org/10.3758/ s13414-020-02079-y 11. Longo MR, Lulciuc A, Sotakova L (2019) No evidence of tactile distance anisotropy on the

belly. R Soc Open Sci 6:180866. https://doi. org/10.1098/rsos.180866 12. Plaisier MA, Sap LIN, Kappers AML (2020) Perception of vibrotactile distance on the back. Sci Rep 10:17876. https://doi.org/10.1038/ s41598-020-74835-x 13. Nicula A, Longo MR (2021) Tactile distance perception on the back. Perception 50:677 14. Sur M, Merzenich MM, Kaas JH (1980) Magnification, receptive-field area, and size in areas 3b and 1 of somatosensory cortex in owl monkeys. J Neurophysiol 44:295–311. https://doi.org/10.1152/jn.1980.44.2.295 15. Brooks VB, Rudomin P, Slayman CL (1961) Peripheral receptive fields of neurons in the cat’s cerebral cortex. J Neurophysiol 24:302– 325. https://doi.org/10.1152/jn.1961.24. 3.302 16. Wall PD (1960) Cord cells responding to touch, damage, and temperature of skin. J Neurophysiol 23:197–210. https://doi.org/ 10.1152/jn.1960.23.2.197 17. Burgess PR, Petit D, Warren RM (1968) Receptor types in cat hairy skin supplied by myelinated fibers. J Neurophysiol 31:833– 848. https://doi.org/10.1152/jn.1968.31. 6.833 18. Cody FWJ, Garside RAD, Lloyd D, Poliakoff E (2008) Tactile spatial acuity varies with site and axis in the human upper limb. Neurosci Lett 433:103–108. https://doi.org/10.1016/j. neulet.2007.12.054 19. Margolis AN, Longo MR (2015) Visual detail about the body modulates tactile localisation biases. Exp Brain Res 233:351–358. https:// doi.org/10.1007/s00221-014-4118-3 20. Medina S, Tame` L, Longo MR (2018) Tactile localization biases are modulated by gaze direction. Exp Brain Res 236:31–42. https://doi. org/10.1007/s00221-017-5105-2 ˜ o´n E, Danvers M et al (2017) 21. Calzolari E, Azan Adaptation aftereffects reveal that tactile distance is a basic somatosensory feature. Proc Natl Acad Sci 114:4555–4560. https://doi. org/10.1073/pnas.1614979114

104

Matthew R. Longo

22. de Vignemont F, Ehrsson HH, Haggard P (2005) Bodily illusions modulate tactile perception. Curr Biol 15:1286–1290. https:// doi.org/10.1016/j.cub.2005.06.067 23. Tajadura-Jime´nez A, V€aljam€ae A, Toshima I et al (2012) Action sounds recalibrate perceived tactile distance. Curr Biol 22:R516– R517. https://doi.org/10.1016/j.cub.2012. 04.028 24. Canzoneri E, Ubaldi S, Rastelli V et al (2013) Tool-use reshapes the boundaries of body and peripersonal space representations. Exp Brain Res 228:25–42. https://doi.org/10.1007/ s00221-013-3532-2 25. Miller LE, Longo MR, Saygin AP (2014) Tool morphology constrains the effects of tool use on body representations. J Exp Psychol Hum Percept Perform 40:2143–2153. https://doi. org/10.1037/a0037777 26. Miller LE, Longo MR, Saygin AP (2017) Visual illusion of tool use recalibrates tactile perception. Cognition 162:32–40. https:// doi.org/10.1016/j.cognition.2017.01.022 27. de Vignemont F, Majid A, Jola C, Haggard P (2008) Segmenting the body into parts: evidence from biases in tactile perception. Q J Exp Psychol 62:500–512. https://doi.org/10. 1080/17470210802000802 28. Le Cornu Knight F, Longo MR, Bremner AJ (2014) Categorical perception of tactile distance. Cognition 131:254–262. https://doi. org/10.1016/j.cognition.2014.01.005 29. Le Cornu KF, Bremner AJ, Cowie D (2020) Does the language we use to segment the body, shape the way we perceive it? A study of tactile perceptual distortions. Cognition 197: 104127. https://doi.org/10.1016/j.cogni tion.2019.104127 30. Dolan BM, Birtchnell SA, Lacey JH (1987) Body image distortion in non-eating disordered women and men. J Psychosom Res 31: 385–391. https://doi.org/10.1016/00223999(87)90009-2 31. Longo MR, Haggard P (2010) An implicit body representation underlying human position sense. Proc Natl Acad Sci 107:11727– 11732. https://doi.org/10.1073/pnas. 1003483107 32. Fuentes CT, Longo MR, Haggard P (2013) Body image distortions in healthy adults. Acta Psychol 144:344–351. https://doi.org/10. 1016/j.actpsy.2013.06.012 33. Keizer A, Smeets MAM, Dijkerman HC et al (2012) Aberrant somatosensory perception in anorexia nervosa. Psychiatry Res 200:530– 537. https://doi.org/10.1016/j.psychres. 2012.05.001

34. Scarpina F, Castelnuovo G, Molinari E (2014) Tactile mental body parts representation in obesity. Psychiatry Res 220:960–969. https:// doi.org/10.1016/j.psychres.2014.08.020 35. Adamczyk WM, Sługocka A, Mehlich K et al (2018) Preliminary validation of a two-point estimation task for the measurement of sensory dissociation in patients with chronic low back pain. Pain Med 20:2472–2478. https://doi. org/10.1093/pm/pny220 36. Cimmino RL, Spitoni G, Serino A et al (2013) Plasticity of body representations after surgical arm elongation in an achondroplasic patient. Restor Neurol Neurosci 31:287–298. https:// doi.org/10.3233/RNN-120286 37. Fiori F, Longo MR (2018) Tactile distance illusions reflect a coherent stretch of tactile space. Proc Natl Acad Sci 115:1238–1243. https://doi.org/10.1073/pnas.1715123115 38. Tame` L, Bumpus N, Linkenauger SA, Longo MR (2017) Distorted body representations are robust to differences in experimental instructions. Atten Percept Psychophys 79:1204– 1216. https://doi.org/10.3758/s13414017-1301-1 39. Longo MR, Morcom R (2016) No correlation between distorted body representations underlying tactile distance perception and position sense. Front Hum Neurosci 10:593. https:// doi.org/10.3389/fnhum.2016.00593 40. Spitoni GF, Pireddu G, Cimmino RL et al (2013) Right but not left angular gyrus modulates the metric component of the mental body representation: a tDCS study. Exp Brain Res 228:63–72. https://doi.org/10.1007/ s00221-013-3538-9 41. Keizer A, Aldegonda M, Smeets M et al (2011) Tactile body image disturbance in anorexia nervosa. Psychiatry Res 190:115–120. https:// doi.org/10.1016/j.psychres.2011.04.031 42. Bassolino M, Finisguerra A, Canzoneri E et al (2015) Dissociating effect of upper limb non-use and overuse on space and body representations. Neuropsychologia 70:385–392. h t t p s : // d o i . o r g / 1 0 . 1 0 1 6 / j . neuropsychologia.2014.11.028 43. Mancini F, Steinitz H, Steckelmacher J et al (2015) Poor judgment of distance between nociceptive stimuli. Cognition 143:41–47. https://doi.org/10.1016/j.cognition.2015. 06.004 44. Marks LE, Girvin JP, Quest DO et al (1982) Electrocutaneous stimulation II. The estimation of distance between two points. Percept Psychophys 32:529–536. https://doi.org/10. 3758/BF03204206

Tactile Distance Perception 45. Tame` L, Tucciarelli R, Sadibolova R et al (2021) Reconstructing neural representations of tactile space. NeuroImage 229:117730. https://doi.org/10.1016/j.neuroimage. 2021.117730 46. Longo MR, Sadibolova R (2013) Seeing the body distorts tactile size perception. Cognition 126:475–481. https://doi.org/10.1016/j. cognition.2012.11.013 47. Longo MR, Ghosh A, Yahya T (2015) Bilateral symmetry of distortions of tactile size perception. Perception 44:1251–1262. https://doi. org/10.1177/0301006615594949 48. Brainard DH (1997) The psychophysics toolbox. Spat Vis 10:433–436. https://doi.org/ 10.1163/156856897X00357 49. Le Cornu Knight F, Cowie D, Bremner AJ (2017) Part-based representations of the body in early childhood: evidence from perceived distortions of tactile space across limb boundaries. Dev Sci 20:e12439. https://doi.org/10. 1111/desc.12439 50. Longo MR (2017) Hand posture modulates perceived tactile distance. Sci Rep 7:9665. https://doi.org/10.1038/s41598-01708797-y

105

51. Avery GC, Day RH (1969) Basis of the horizontal-vertical illusion. J Exp Psychol 81: 3 7 6 – 3 8 0 . h t t p s : // d o i . o r g / 1 0 . 1 0 3 7 / h0027737 52. Weinstein S (1968) Intensive and extensive aspects of tactile sensitivity as a function of body part, sex, and laterality. In: Kenshalo DR (ed) The skin senses. Thomas, Springfield, pp 195–222 53. Mancini F, Bauleo A, Cole J et al (2014) Whole-body mapping of spatial acuity for pain and touch. Ann Neurol 75:917–924. https:// doi.org/10.1002/ana.24179 54. Shepard RN (1980) Multidimensional scaling, tree-fitting, and clustering. Science 210:390– 398. https://doi.org/10.1126/science.210. 4468.390 55. Everitt BS, Rabe-Hesketh S (1997) The analysis of proximity data. Arnold, London 56. Prins N, Kingdom FAA (2009) Palamedes: Matlab routines for analyzing psychophysical data. http://www.palamedestoolbox.org 57. Craig JC, Johnson KO (2000) The two-point threshold: not a measure of tactile spatial resolution. Curr Dir Psychol Sci 9:29–32. https:// doi.org/10.1111/1467-8721.00054

Part II Affective Touch, Pain, Wetness, Itch, and Interoception

Chapter 6 Affective Touch: Psychophysics, Physiology and Vicarious Touch Perception Connor J. Haggarty, Adarsh Makdani, and Francis McGlone Abstract It is hypothesized that the affective quality of touch is conveyed through the stimulation of a group of unmyelinated low threshold mechanoreceptors (LTM) called C-Tactile afferents (CTs). Research has shown that CTs selectively respond to slow gentle stroking touch, delivered at skin temperature (~32 ° C). Over the past three decades many studies have benefited from the development of bespoke psychophysical tools and methods to optimally stimulate these afferents, while controlling for activity in discriminative A-β afferents. More recently, the affective quality of CT-optimal touch has been measured using facial electromyography (EMG), allowing us to capture the more subtle affective responses to CT stimulation. One of the key arguments for this afferent being important for sociability is that the touch is vicariously experienced when we observe others receiving an optimal tactile stimulus. A series of studies have shown that cortical activity matches the activity elicited during first-hand touch, so we also discuss these videobased studies here. Key words Affect, Social Touch, Psychophysics, Physiology, Vicarious, C-Tactile Afferent

1 1.1

Introduction The Periphery

The skin is innervated by multiple classes of afferents responsible for the sensations of touch, temperature, pain, and itch. These afferents can be divided into morphologically and functionally distinct groups, such as C-fibers, which are thin, unmyelinated, slowly conducting nerve fibers, known for encoding nociceptive (see Chapter 7, this volume) and pruritic (see Chapter 8, this volume) sensations. More recently, a subgroup of C-fibers has been discovered in the hairy (non-glabrous) skin of humans; dubbed C-Tactile afferents (CTs) and are analogous to the C-low threshold mechanoreceptors (C-LTM) found in other animals. Microneurography studies (see Chapter 15, this volume) show that CTs respond optimally to innocuous tactile stimulation [1], and that they are selectively tuned, firing most frequently to dynamic touch delivered slowly (1–10 cm/s) and gently (~0.3 N) [2].

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_6, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

109

110

Connor J. Haggarty et al.

Fig. 1 The properties of afferent nerve fibers in the skin. (a) Shows the firing frequency for velocity preferences of the CT-afferent with optimal firing for velocities between 1 and 10 cm/s. Stimuli below and above this range result in a lower firing frequency. Similarly, participant’s ratings of pleasantness (b) follow this same pattern of activity whereby stimuli delivered at a CT-optimal range result in the greatest levels of pleasantness relative to slower and faster velocity stroking. Figure (c) shows a strong positive relationship between firing frequency and subjective pleasantness ratings. (Reproduced from [2] with permission from XXXX)

The slow conduction velocity of C fibers means that CTs are thought to have little discriminative value, but are instead hypothesized to encode for affective and social touch, by conveying the emotional quality of the stimulus, as C-nociceptors do for pain [3, 4]. The fact that there is an emotional or affective, in addition to a discriminative, dimension to touch has only been recognized anatomically relatively recently, but it is present for both nociceptive and tactile input [5]. This relationship between the mechanoreceptive and affective responses is supported by the close correlation between CT firing and the perception of pleasantness at naturalistic stroking velocities and forces (Fig. 1) [2]. More recently it has been shown that the optimal temperature of touch delivery is ~32 °C, close to the temperature of human skin [6]. Taken together, this evidence suggests that CT fibers are optimally activated by touch reminiscent of a gentle caress from another individual, and that the skin is responsive to a social touch [7]. Research into affective touch relies on the combination of psychophysics with an understanding of the underlying neural mechanisms. As a result of the complexity of the somatosensory system, where any cutaneous stimulus will activate multiple afferents in parallel, much of the foundation of knowledge in this area comes from studies employing single unit microneurography (see Chapter 15, this volume), or in-patient populations with a wellcharacterised neuronopathy. For example, patients with a congenital reduction in C-fiber density show reduced cortical activation and reduced hedonic ratings to felt and seen dynamic touch [8]. Conversely, studies involving patients with large fiber neuronopathies (e.g., IW and GL, who lack large myelinated afferent

Affective Touch

111

function below the neck) have provided an opportunity to understand the role of C-fibers (and finely myelinated afferents) in isolation. With this insight into the peripheral system, it has been possible to design psychophysical studies using finely controlled affective touch stimuli, to investigate sensations and perceptions in healthy participants. 1.2

The Cortex

Unlike myelinated afferents, CTs project to the dorsal posterior insula cortex, a region associated with homeostatic functioning and nociceptive input (Fig. 2). The affective quality of these afferents

Fig. 2 The functional properties of Aβ and CT afferents and their central projections. At the lowest level, Aβ and CT afferents respond differently to stimuli. Where Aβ afferents’ discharge frequency increases linearly with stimulus velocity, CTs respond optimally to stimuli of between 1 and 10 cm/s, with a lower discharge frequency to slower and faster stimuli. These afferents project to different lamina in the dorsal horn: while Aβ fibers project via the dorsal column to primary and secondary somatosensory cortices, CTs are believed to project via the spinothalamic tract to the posterior insula cortex. (Reproduced with permission from [5] XXXX)

112

Connor J. Haggarty et al.

comes from further projections to the anterior cingulate and the orbitofrontal cortex, regions associated with reward and affective valuation of stimuli. When compared with pain processes (which follow a similar pattern—see Chapter 7, this volume) it is possible to see the juxtaposed role of the pain and affective touch systems, with one having negative affective value and motivational function (pain), and the other having a positive affective value and motivational reward (affective touch). Recent studies have shown that optimal activation of the latter afferents results in electrophysiological indicators of positive affect [9–11]. These studies used facial electromyography (EMG) which measures affective responses directly from muscles associated with changes in affective state. For example, the zygomaticus major muscle in the cheek is responsible for drawing the mouth into a smile, and thus represents positive affect, while the corrugator supercilii muscles are responsible for drawing the brow into a negative affective state of frowning. The cortical mechanisms and subsequent behavioral consequences of somatosensory perception are observed both during physical stimulation and vicarious observation of touch [12– 15]. This effect is likely the result of the human ability to empathize with another’s cognitive and emotional states [16, 17]. This is supported by the fact that vicarious experience of both pleasant and unpleasant somatosensory stimuli has been shown to activate regions of the cortex associated with imitation and socio-emotional behavior, such as the anterior insula, anterior cingulate cortex, and temporoparietal junction [18–20]. In two separate experiments, Morrison and colleagues [15] reported that both receiving and observing CT-optimal touch resulted in selective activation of the posterior insula cortex and not the somatosensory cortex (see Chapters 17 and 18, this volume for neuroimaging of somatosensory cortex). Participants rated the touch-giver as more pleasant and likeable after observing reciprocal interactions between the touchgiver and touch-receiver [21]. It was also reported that the touchgiver was attended to more often during subsequent viewing periods. Furthermore, the observation of images depicting social tactile interactions (bonding stimuli) has been reported to increase participants’ subjective feelings of sociability, to lower feelings of isolation, and to increase subsequent electroencephalographic (EEG) activity [22]. In this chapter, we will describe some of the methods used to measure behavioral and EMG responses to affective touch; we also included some of the hints and tips necessary for measuring responses to this type of touch stimulus.

Affective Touch

2

113

Materials

2.1 Stimulus Delivery

In psychophysics it is important to control for variables that may affect the stimulus you are trying to deliver. In studies using tactile stimuli there are numerous methods that have been used to deliver replicable stimuli in a controlled manner (for overview of methods for delivering tactile motion, see Chapter 4, this volume). Here we describe both automatic and manual methods for direct stimulus delivery, and video methods that test vicarious affective touch.

2.1.1 Rotary Tactile Stimulator and Force Transducer

The rotary tactile stimulator (RTS, Fig. 3) is a custom-built stroking robot. Designed in partnership with the Biomedical Engineering Department at Chapel Hill (NC) and engineer Chris Dancer (Dancer Design, UK), the RTS is an optimal way to deliver consistently reliable and replicable tactile stimulation to various body sites, free from the social context, for example, of being stroked by an experimenter (see Subheading 2.1.2). The RTS comprises 1 to 4 probes that stroke across the skin in a precise arc. With an inbuilt force transducer, the RTS is calibrated to apply each stroke at a programmable force and velocity at the selected body site (See Note 4.1). The RTS can use multiple probes, to vary, for example, the texture of the stimuli. The material of the probe is an important factor to consider depending on the question of interest. Some earlier studies looked at the difference in pleasantness between,

Fig. 3 A typical rotatory tactile stimulator (RTS). The participant’s arm is resting comfortably on a cushion or pillow and the brush stroking probe runs back and forth over their arm. Participants rate the pleasantness and/or intensity of the perceived touch using the VAS scale in their opposite hand. (Reproduced with permission from [24] XXXX)

114

Connor J. Haggarty et al.

for example, soft materials and burlap [23]. However, most CT experiments use a smooth low-friction material to emulate skin, or a soft brush such as an artists’ goat-hair brush or cosmetic brush. It is not typical to see differences in pleasantness ratings for these different materials, unless the texture is coarse and unpleasant. There is scope for innovation to consider new questions; for example, Ackerley and colleagues [6] used a mechanical-thermal probe with the RTS to simulate tactile stimuli whilst controlling for temperature. Dancer Design has created a Graphical User Interface (GUI) written in Python using LabVIEW, that allows a researcher to easily program experiments, delivering stimuli at controlled velocities and forces in a programmable order. The RTS can be controlled via the GUI or can send and receive TTL triggers to interact with other hardware, or experimental software such as E-Prime or PsychoPy (e.g., via the parallel port). Although the questions of interest may be different across experiments, most have focused on understanding differences in pleasantness between CT-optimal and non-optimal velocities, through implicit or explicit measures. The questions can be presented in many forms, though most often these are programmed via the GUI or one of the aforementioned programs, so that, for example, participants are presented with an electronic visual-analog scale (VAS) or Likert scale after each stroke (or block). Due to the fine velocity and force control, the RTS is the goldstandard means of delivering controlled dynamic touch (see also Chapter 4, this volume). However, this comes at a significant monetary cost, and at the expense of some flexibility and maneuverability. Other devices have been developed, for example, to deliver (static) punctate or vibrating tactile stimuli. One such alternative currently in development is the low-cost MultiTAC device (https://somaffect.org), which is hoped will be able to deliver both linear and circular movements, at a controlled velocity and constant force. 2.1.2 Manual Stroking Techniques

Some study designs are not well suited to robotic or computercontrolled stimulus delivery, whether due to cost of equipment, or practicalities such as a desire to stimulate multiple different body locations in a short period of time. In these instances, with some training, it is entirely possible for researchers to deliver controlled, manual strokes (see Note 4.2). It is important to initially mark the region of interest on the skin to ensure that each of the stimuli is delivered to the same location to the arm. This can be done simply by using a marker pen; however, some researchers may choose to put something on either side of this area such as plastic film [25] to limit the sensations outside of this region. Typically, this region is ~9 cm long and as wide as the probe or brush, to easily allow for the typical stimulus velocities of 0.3 cm/s, 3 cm/s, and 30 cm/s.

Affective Touch

115

Fig. 4 A screenshot showing the metronome during one of its runs. The metronome was programmed in E-Prime 2.0 (Psychology Software Tools, Pittsburgh, PA). For studies where participants are stroked for a set period the metronome runs back and forth for the allotted time meaning that stimuli were matched for contact time on the skin. The red metronome line represented a proximal to distal stroke whereas a white metronome (not shown) represented distal to proximal strokes. Some studies may have a single color metronome to represent a single stroke per trial irrespective of the direction of the touch being delivered

Variations in these velocities are not uncommon, for example, 18 cm/s [26, 27], which was chosen as a more natural speed, as 30 cm/s may not reflect the typical speed of interactions between individuals. To control for stimulus force, touch can be delivered using a weighted probe designed to apply a set force to the skin, or using a soft brush where the force applied by the buckled bristles is relatively constant. The accuracy of this constant-force delivery can be trained by practicing stimulus delivery on a sensitive balance or scale. To control for velocity, control measures such as visual (Fig. 4) or auditory metronomes can help to train and guide researchers. These are essential tools to minimize human error. The use of metronomes typically follows the same procedure in all studies. At the beginning of each trial, the velocity, and other parameters such as the location of the touch are signaled to the experimenter only, for example, on a computer screen located behind the participant. It is important that participants are not unduly cued to the metronome stimulus, and should focus on the physical sensation, not the metronome (see Note 4.3). A short countdown primes the experimenter, before the metronome begins to guide the experimenter in delivering the correct velocity of touch. In a visual metronome, an empty rectangle represents the region of interest or aperture on the skin and is filled at the same rate that the experimenter should stroke for that trial. Some researchers choose to have a set period for stroking, for example, matched contact time (i.e., where strokes at any velocity are

116

Connor J. Haggarty et al.

matched for the amount of time the skin is stroked), whereas others opt for single strokes on the skin. Pawling and colleagues [9] found that controlling for contact time on the skin between fast and slow stroking velocities (i.e., a single 3 cm/s stroke has the same contact time as ten 30 cm/s strokes) did not have any effect on the pleasantness of the touch or physiological arousal in participants, so it is suggested that this is not an important control measure to consider. In studies where timing is important, the start and end of strokes may need to be recorded, for example, using triggers from the metronome to the data recording. Manually administered brush strokes have been used effectively across a number of previous CT-focused studies [9, 28–31], and typically elicit pleasant sensations in comparison to other materials [32]. Haggarty and colleagues [33] used a soft make-up brush, whereas Yu and collaborators [25] used a goat-hair brush, both achieving similar patterns of ratings. Although stroking participants with a brush has little external social validity (outside of behaviors such as applying make-up), it allows for a social component to be included with the stimulus where a human is delivering the touch. This social component is not present when touch is administered using an RTS. Although efforts have been made to simulate a more naturalistic interaction with manual stroking of the skin using the hand [34], this comes with its own challenges. Using a brush, as opposed to a hand, ensures stimulus consistency without any individual differences in skin texture or temperature affecting the velocity or force of the stroking [35]. This is important when the goal of the study is to activate CTs optimally, but not when the goal involves social context-dependent responses. Importantly, Triscoli and collaborators [36] reported that touch delivered by an RTS was comparable in terms of perceived pleasantness and intensity to manually administered brush stroking (see Notes 4.4, 4.5 and 4.6). 2.1.3 Videos for Vicarious Touch

Walker and colleagues [37] created a series of videos depicting touch delivered at CT-optimal and non-CT-optimal velocities (see also Chapter 8, this volume on itch-inducing videos). These videos were created to measure the vicarious affective responses to dynamic social touch. Researchers can compare ratings across skin sites hypothesized to have differing innervation densities of CTs, for example, hairy skin sites from the back (densely innervated) to the lower arm (low innervation density). The videos comprise 5 s clips of touches being delivered to different body sites, with either static placement of the hand on the skin or at different stroking velocities. Participants watch videos depicting one actor touching the upper body of another actor. The actors (one male and one female) stand in front of a white screen. To minimize the effect of social context and top-down

Affective Touch

117

Fig. 5 Screenshots from Walker and colleagues’ [37] videos. The images show touch delivered to (from top left, clockwise) the back, upper arm, ventral forearm, palm, and dorsal forearm. Just the hand and the stroked body part are visible to remove social influences and the actors stand in front of a plain background

representation of the touch, only the touched location and the toucher’s arm are visible in each shot (Fig. 5). This set of videos shows touch delivered to five locations (palm, ventral forearm, dorsal forearm, upper arm, and back) at three different velocities (static touch, 3 cm/s, and 30 cm/s). Recently, an updated series of videos has been created where both the male and the female play the role of the stroker/stroked and a further sub-optimal stroking velocity of 0.5 cm/s is included. More recently still, these videos have been used in a series of studies with typically developing children [38] and children with autism [39], whereby participants rated perceived pleasantness of the touch using “smiley face” scales. After viewing each video, participants are asked to rate “How pleasant was that action for the person being touched?” and “How much would you like to be touched like that?” These were answered on a seven-point Likert scale running from 1, “very unpleasant” or “not at all” to 7, “very pleasant” or “very much so”. As these participants are not experiencing the sensation themselves, these questions measure their ability to vicariously experience the touch, and how they feel about receiving that touch themselves [37–39]. The preference for CT-optimal stroking touch has been shown in a relatively small number of trials. In fact, Walker and colleagues [37] used a sample of videos to show that both velocity and location

118

Connor J. Haggarty et al.

affected preference for CT touch in as little as 15 trials (5 locations × 3 velocities). Similarly, most first-hand touch studies show reliable results from just three stimulations, one at each of the different velocities. This ensures that an average preference is taken across three separate stimulations rather than the touch being directly influenced by the preceding touch stimulus. 2.2 Implicit Measures of Affective Touch 2.2.1 Facial Electromyography (EMG)

2.2.2

Microneurography

Facial EMG is a physiological technique used to measure the activity of muscles in the face. This has been of particular interest to social neuroscientists studying emotion and social interactions. In the field of affective touch, it is used for a more implicit observation. It is hypothesized that CTs inherently encode pleasant sensation. Furthermore, these fibers project to regions of the cortex associated with processing the relative value and rewarding aspect of stimuli. As participants typically rate CT-optimal touch as the most pleasant type of touch, it stands to reason that these are inherently pleasant stimuli. Using facial EMG, researchers can measure the physiological affective arousal that is the automatic response to a pleasant stimulus such as a smile. Despite this being a relatively new avenue for CT research, some interesting results have already been obtained, with two groups reporting that stimulation targeting CTs does indeed have a pleasant hedonic value [9, 10]. The facial EMG in these studies showed different results. In Mayo and colleagues’ study [10] the positive valence was shown by the decrease in the activity of the muscle responsible for frowning, the corrugator supercilii (CS, Fig. 6), whereas in Pawling and colleagues’ study [9], the positive valence was shown by an increase of activity in a smiling muscle, the zygomaticus major (ZM, Fig. 6). Despite there being activity in different facial muscles here, it is important to note that these results show the same thing—an increase in positive affect and a decrease of negative affect. Thus CT-optimal touch carries a positive affective value. As the anatomy and physiology of these muscles is well known, it is a simple enough task to choose regions of interest for a given experiment. Most facial EMG studies in the affective touch domain focus on the CS and ZM as these reflect opposing affective states, where an increase in one of these muscles is likely to cause no effect or even a decrease in the other. Many companies who provide equipment for physiological measurements such as heart rate or galvanic skin response also provide the electrodes necessary to complete facial EMG recordings. Although microneurography is covered in greater detail elsewhere (see Chapter 15, this volume), its specific value to affective touch research should not be overlooked. Single-unit microneurography allowed the identification and characterization of CTs, providing a neural mechanism which is the foundation for the psychophysical

Affective Touch

119

Fig. 6 A diagram showing the approximate locations of the zygomaticus major (ZM) and corrugator supercilii (CS) muscles on the human face. Placed near the hairline is a reference electrode which provides baseline activity measures free from spontaneous and background muscle activity present even at rest. Electrodes are attached using a small adhesive disk, and electrode caps are filled with conducting gel

studies in this field. Microneurography is an electrophysiological technique which allows for the recording of single afferent activity—such as from a mechanoreceptor in the skin, from the nerves of conscious human participants using a tungsten microelectrode inserted percutaneously. It can be used in combination with any of the psychophysical methods described above to investigate afferent responses to controlled stimuli. This activity is captured and amplified through a head stage and captured by a data acquisition system (e.g., ADInstruments NeuroAmp and PowerLab). Spike sorting and analysis software (e.g., ADInstruments Labchart 8) can then be used to quantify this response.

3

Methods

3.1 Control Conditions and Stimuli

It is important to consider control conditions required for any psychophysical study. There are several considerations in relation to CT psychophysical studies. For example, these afferents have been shown to fatigue (adapt), and this is particularly prominent with continued stroking. In fact, this fatigue appears to reduce the pleasantness of the stroking touch [41]. It is therefore important that this is considered when these studies are being designed and conducted. One stimulus control is to ensure that there is a suitable

120

Connor J. Haggarty et al.

interstimulus interval (ISI). Here, 10 s between stimuli is chosen to ensure that the afferents can return to their resting state without becoming fatigued. Furthermore, given the preference of CTs for slow gentle stroking [2], it is important to have an appropriate nonCT-optimal velocity of touch for comparison (continued in Subheading 3.2). Also, the location of the touch is an important factor in stimulus delivery. Not only will this be an important consideration for the method being used but also for the innervation densities on different body parts. For example, it is common in these studies to control for CT activation using a manipulation of velocity, but it is not often considered that a location control can also be used. CTs have not been found in the glabrous skin of the palm—so far, although this has more recently been contested in microneurography experiments [42]. 3.2

Stimulus Velocity

3.3 Stimulus Evaluation

The velocities of the stroking touch can vary across experiments; however, it is important to consider that CT-optimal velocity touch is between 1 and 10 cm/s [2]. This is not to say that CTs do not fire outside this range, but that firing frequency is optimal within this range of velocities. Comparatively, it is important to include a control or “non-CT-optimal” velocity. However, exactly what this velocity should be has long been discussed. Most frequently 30 cm/s is chosen, as this is at the upper end of an inverted-U or quadratic function (where the firing frequency is at its lowest), where stimuli are both considered less pleasant and also result in the reduced firing frequency of CTs. Often, researchers will opt to include a velocity less than 1 cm/s, so that it is sub-CT-optimal, and one faster supra-CT-optimal (e.g., 30 cm/s), to compare against the CT-optimal velocity of between 1 and 10 cm/s. Simpler studies that want to examine the underlying mechanisms that drive maximal CT firing (slow stroking) and maximal Aβ firing (fast stroking) will compare 3 to 30 cm/s velocities directly (see Note 4.7). When planning a study, you need to consider the measurements that you are taking. For example, if you are taking a physiological measure, then it is important to consider that any significant changes in affective arousal state may not happen until after the initial stroking period, and an “evaluation” period needs to be considered. This appears to be particularly pertinent in facial EMG and EEG research, where the affective evaluation of the stroking stimulus may in fact be most prominent after stimulus offset. In the post-evaluation period, you should then consider a measure to determine the subjective experience of the stimulus. Examples of this are to ask how pleasant a stimulus is, how intense it is, or for video-based studies participants could be asked how much they empathize with the person being touched. Rating scales have been used with various electronic and paper-based methods, ranging from 100-point VAS scales to five-point Likert scales. From the

Affective Touch

121

experience in our lab, none of these seem to be more effective than others in teasing out the differences in individual preferences of the touch stimulus. 3.4 Manual and Robotic Stroking Procedures

One important consideration, when developing a touch paradigm is not just the location of interest but also that the participants will remain comfortable for the duration of the study. Therefore, a great deal of effort is taken to ensure that the participant is comfortable, but also that they do not move during the study. In RTS-based studies, the participant’s arm can be secured in place using a vacuum cushion; however, a similar procedure is used in manual stroking studies. For example, Haggarty and colleagues [33] used a foam block to make the participant comfortable, whereas Pawling’s group [9] positioned the participant’s arm on a cushion. Once the participant’s arm is in a comfortable posture then a target region of interest can be drawn onto the arm to ensure that stroking occurs in the same location consistently. Interestingly Hauser and colleagues [34] used an infrared tracking system to objectively measure the stroking activity of the touch-giver (as in this case it was a participant performing the touch). Prior to completing an RTS component of a study, it should be made clear to participants that they must keep their arm still throughout the experiment otherwise the RTS will need to be recalibrated whenever participants move. After instructing participants about the procedure and whether they will have to close their eyes or if their vision is occluded in some way, you should run practice trials. This is not a study or type of stimulation that participants will necessarily be used to. For example, running a make-up brush across the skin may be “normal” to someone who applies make-up regularly, but this will not be a common sensation to others. Furthermore, it is not often that you would encounter a robot that provides soft tactile sensations to you, so getting used to the sounds and appearance of the RTS is an important precursor to completing the study. It is important to include both the different velocities and different locations in this practice period to ensure participants are aware of all aspects of the study. Furthermore, there are some populations, such as of autistic individuals, who may have sensitivities to tactile stimuli, so it is important to prepare them for the types of stimuli they will experience in the experiment.

3.5 Video-Based Induction

The aim of video-based studies of affective touch is to determine how empathic vicarious experiences of touch impact a participant’s ratings of observed touch (see also Chapter 8, this volume on videobased itch induction). This is reliant on a participant’s empathic abilities and underlying preferences for touch. It is important to consider that the aim of the questions asked in these studies is to measure empathic ability—“How pleasant was that action for the

122

Connor J. Haggarty et al.

person being touched?”—that is, to understand how someone else might experience that touch and the participant’s own desire for touch like that: “How much would you like to be touched like that?” In these studies, participants watch each of the videos, then immediately after are asked the aforementioned questions. It is beneficial to guide the participants to the start of each trial, typically by using a fixation cross that is then replaced by the video. Participants should have the questions presented to them at the beginning of the study so that they are thinking about what the touch means for the person receiving it and what it would mean to them if they were to receive it themselves. Videos are an interesting tool in this research in that they do not necessarily need to be presented in the laboratory, and can be run with large numbers of participants relatively easily. Some researchers use these videos within brain scanners, while others are interested in the vicarious experiences reflected through behavioral measures, and the mediating effects of personality traits. In these studies, it is possible to run these studies online using a survey methodology which greatly increases participant numbers. By running a study online, however, you run the risk of participants completing the study incorrectly or without sufficient attention to the stimuli. That said, online data presented in Haggarty and colleagues [38, 39] were comparable to those from lab-based studies [37], where CT-optimal velocity touch at CT-innervated locations was consistently deemed to be the most pleasant in typically developing participants. This suggests that online studies are possible in this modality, particularly as the participants are not just answering questionnaires but also have an engaging videobased task. One of the important questions to consider for video-based studies is at what age this mode of presentation is applicable. Firsthand studies of touch in infants show that CT-optimal velocity touch is the most pleasant [40]; however, recent evidence [38, 39] suggests that the vicarious experience of this touch is different in children, and that they may not detect a difference in the touch when it is delivered to another individual. This raises the question of when empathic responses for a subtle affective experience created by affective touch are present in development (see Chapter 11, this volume, for methods of studying atypical development). 3.6

Facial EMG

The electrodes for facial EMG should be set up before the participant arrives in the lab. This means attaching the adhesive collars and filling the electrode cups with conductive gel. This allows the conductive gel to settle into the electrode cup so that the wires are attached to the appropriate areas of the collection hardware. There are different types of electrodes that can be used for EMG,

Affective Touch

123

ADI Computer Researcher

Metronome

ADI System

Participant Computer

Participant

Fig. 7 A typical set up for a manual stroking experiment, with facEMG set up designed by Connor Haggarty. Participants are sat facing the computer screen where they are instructed to close their eyes during the stroking trial. A post-stimulus period for “evaluation” of the sensation follows stimulation, then a tone sounds to signify that participants are to open their eyes and answer questions relating to pleasantness and intensity of the touch. Participants cannot see the visual metronome located behind them and observation of the stroking is prevented by them closing their eyes during each trial

including flat electrodes typically used to measure eye movement in EEG studies, and needle electrodes which are more invasive and therefore not ideal for measuring natural facial muscle activity. The hardware should be set up close enough to the participant’s chair so that the cables are not stretched and so that the participant can move or adjust slightly without pulling the electrodes off their face (Fig. 7) (see Note 4.8). It is wise to avoid using the word “electrode” when discussing the procedure with participants, as this may have negative connotations. The word “sensors” acts as a nice buffer for participants to understand what the procedure is and to understand that something is being measured from them without any negativity (see Notes 4.9 and 4.10).

124

Connor J. Haggarty et al.

An important factor to consider in facial EMG research is that many studies have found that an evaluative post-stimulus period is important for considering participants’ affective state. It is therefore necessary to include an ISI period (~5 s) between the stimulus and any questions to account for any changes in affective arousal during this post-stimulus period.

4

Notes

4.1 Participant Movement

It is important that the participant does not move the target body site once the RTS calibration is complete, as any movement will alter the force delivered. Ensure that participants are as comfortable (and still) as possible by lying or sitting participants down in an adjustable chair or bed; using adjustable straps and a vacuum cushion to form a temporary cast which supports and immobilizes the target site or limb.

4.2

For any new researcher, it is important to consider the training required to provide replicable tactile stimulation. For the RTS this requires some minimal training for how the RTS is programmed for each study and how the experiment and RTS computer communicated with each other. By contrast, manual stroking is more complicated and tiring. Fundamentally, it is expected that this technique will be more open to differences between researchers. As described in Subheading 2.1.2 on manual stroking, there are different ways to support researchers during the experiment. This includes a visual metronome showing the velocity of the stroking and using the appropriate stroking tool to ensure a smooth stroking trial. Training is also necessary. It is important that researchers run pilot participants (usually from within the lab group) so that they can get used to the technique, using the stroking tool and using the metronome to guide them.

Training

4.3 Participant Blinding and Attention

As with most studies, it is important for participants to be blinded to the purpose of the study. This is just as important in these studies, whether participants are naı¨ve to the field of CTs or not. It is important for them to be focused on the sensation they are feeling and not the action of being touched. For example, consider an experiment where the participant is being stroked by the RTS. This is not a typical situation for them and if their focus is on the RTS itself there are likely top-down influences on the perception of the touch based on the novel situation. This is equally an issue in manual stroking studies where participants are being stroked by a stranger by some form of tool. Two ways to control for these top-down influences are first to have participants close their eyes during the stroking procedure, or second to provide the participant with an attentional task. These

Affective Touch

125

two examples are reliant on the type of study being conducted. Haggarty and colleagues [33], for example, used an “oddball” task where participants had to distinguish between a target velocity stroke and a less common “oddball” stroke. The benefit of this task is that participants are thinking about the sensation of the touch and how it differs to the previous and next trials. Some researchers opt to occlude the participant’s arm from view completely (but see also Chapter 13, this volume, on somatosensory illusions involving hiding the arm). Again, the benefit of this is that the participant is focusing on the sensation and not the action of the touch being delivered or indeed who is delivering the touch. 4.4 Pleasantness Versus Intensity

During a first-hand stroking or vicarious touch procedure, the participant is asked how pleasant they find the touch. More recently the reference to the “intensity” of the stimulus has also been included. The intensity of a tactile stimulus is an important factor in the same way that relative arousal of an emotional stimulus is measured, the role of intensity should be linked to the sensation from the stimulus [43]. From personal experience, this is not a clear question to ask participants. In fact, when asked how “intense” a stimulus is, participants will ask what this means. It is important to consider a description that can be used across experiments for the term intensity to ensure that this is the same mechanism that is being measured across studies.

4.5 Bare Hands Versus Gloves

Stroking using a bare hand will likely not give the same experience in every trial and across participants. In fact, the skin on the palm is prone to sweating as a mechanism of physiological arousal, and this may result in different temperatures and textures, and changes in friction that alter the sensation (see Chapter 9, this volume). To overcome this, researchers have often opted for a soft glove, such as one made of silk. This has the benefit of running smoothly across the skin with a similar social intent but without the risk of altering perception between trials or participants.

4.6 Hairy Versus Glabrous Skin

The dorsal surface of the arm has a high density of hair follicles. When stroking this surface, it is possible that you will get a lot of activity in response to the activation of hair unit receptors and not of CT afferents. To control for this, it is better to stroke the ventral surface of the forearm which is smoother and less hairy.

4.7 Unusual Stimulus Velocities

More recently there has been discussion in the “CT community” about the validity of using unusual velocities of touch. By this we mean that the touch delivered at 30 cm/s has very little ecological validity, in that individuals would not experience this kind of touch in any naturalistic setting. It is therefore common to see 18 cm/ s being used as a control velocity. This raises some concerns. It is of course beneficial for top-down processing of tactile interactions

126

Connor J. Haggarty et al.

that a more naturalistic 18 cm/s is used, and so for cognitive or behavioral studies this may be the best option. However, when comparing a CT-optimal velocity, it is important to have a velocity likely to result in a greater frequency of Aβ activity as a comparison. Aβ afferents’ firing frequency increases linearly with stroking velocity, so the faster the stroking, the more these afferents fire. However, it is important to consider that the velocity should not be uncomfortable and should be sustainable for a researcher repeating it many times, so 30 cm/s is often used. It is therefore justified to include faster-stroking velocities for studies interested in the physiological responses of these afferents as opposed to their cognitive effects. 4.8 Securing Electrodes

To ensure that electrode cables do not pick up any electrical interference or that the electrodes are not pulled from the participant’s face or interfered with, electrodes are usually attached to the same side of the face with all cables running toward the participant’s ear. This means that the cables can be held in place by the participant’s ear and if necessary, attached to their chair. Make sure there is enough slack in the cables so that participants can move slightly (see also Chapter 19, this volume).

4.9 Confirming Electrode Placement

Before collecting any data, it is a good idea to get participants to frown and/or smile—or perform any other facial response of interest—to ensure that you are measuring from the appropriate muscles. These overt facial responses will be clearly recognizable on the channels in your recording software, and act as a good measure of electrode placement. Remember, however, that asking participants to do this before the experiment could result in priming or biasing them. If choosing to test these responses before the study, you may ask participants to make “priming-free” movements such as “draw your eyebrows together” and “raise the corners of your mouth” in this way you can measure the cleanliness of the data being collected prior to recording (see also Chapter 20, this volume).

4.10 Condition Blinding

It is wise to create a cover story for the procedure to ensure that participants are not primed to the purpose of the study (i.e., of implicit emotional arousal). One approach is to say to participants that the top sensors—those measuring the corrugator muscle activity and the reference, are measuring frontal lobe activity, and that the zygomaticus muscle electrodes on the cheek are reference points. This removes the consideration of emotional processes, and avoids drawing attention to participants’ muscle responses.

Affective Touch

127

References 1. Vallbo A, Olausson H, Wessberg J, Norrsell U (1993) A system of unmyelinated afferents for innocuous mechanoreception in the human skin. Brain Res 628(1–2):301–304. https:// doi.org/10.1016/0006-8993(93)90968-s 2. Lo¨ken LS, Wessberg J, Morrison I, McGlone F, Olausson H (2009) Coding of pleasant touch by unmyelinated afferents in humans. Nat Neurosci. https://doi.org/10.1038/nn.2312 3. McGlone F, Reilly D (2010) The cutaneous sensory system. Neurosci Biobehav Rev. https://doi.org/10.1016/j.neubiorev.2009. 08.004 4. Bessou P, Perl E (1969) Response of cutaneous sensory units with unmyelinated fibers to noxious stimuli. J Neurophysiol 32(6):1025–1043. https://doi.org/10.1152/ jn.1969.32.6.1025 5. McGlone F, Wessberg J, Olausson H (2014) Discriminative and affective touch: sensing and feeling. Neuron. https://doi.org/10.1016/j. neuron.2014.05.001 6. Ackerley R, Backlund Wasling H, Liljencrantz J, Olausson H, Johnson RD, Wessberg J (2014) Human C-tactile afferents are tuned to the temperature of a skin-stroking caress. J Neurosci. https://doi.org/10.1523/ jneurosci.2847-13.2014 7. Morrison I, Lo¨ken LS, Olausson H (2010) The skin as a social organ. Exp Brain Res. https:// doi.org/10.1007/s00221-009-2007-y 8. Morrison I, Lo¨ken LS, Minde J, Wessberg J, Perini I, Nennesmo I, Olausson H (2011) Reduced C-afferent fibre density affects perceived pleasantness and empathy for touch. Brain. https://doi.org/10.1093/brain/ awr011 9. Pawling R, Cannon PR, McGlone FP, Walker SC (2017) C-tactile afferent stimulating touch carries a positive affective value. PLoS One. https://doi.org/10.1371/journal.pone. 0173457 10. Mayo LM, Linde´ J, Olausson H, Heilig M, Morrison I (2018) Putting a good face on touch: facial expression reflects the affective valence of caress-like touch across modalities. Biol Psychol. https://doi.org/10.1016/j.bio psycho.2018.07.001 11. Ree A, Mayo LM, Leknes S, Sailer U (2019) Touch targeting C-tactile afferent fibers has a unique physiological pattern: a combined electrodermal and facial electromyography study. Biol Psychol. https://doi.org/10.1016/j.bio psycho.2018.11.006

12. Ebisch SJH, Perrucci MG, Ferretti A, Del Gratta C, Romani GL, Gallese V (2008) The sense of touch: embodied simulation in a Visuotactile mirroring mechanism for observed animate or inanimate touch. J Cogn Neurosci. https://doi.org/10.1162/jocn.2008.20111 13. Keysers C, Kaas JH, Gazzola V (2010) Somatosensation in social perception. Nat Rev Neurosci. https://doi.org/10.1038/nrn2833 14. Lamm C, Silani G, Singer T (2015) Distinct neural networks underlying empathy for pleasant and unpleasant touch. Cortex. https://doi. org/10.1016/j.cortex.2015.01.021 15. Morrison I, Bjornsdotter M, Olausson H (2011) Vicarious responses to social touch in posterior insular cortex are tuned to pleasant caressing speeds. J Neurosci. https://doi.org/ 10.1523/jneurosci.0397-11.2011 16. Kaplan JT, Iacoboni M (2006) Getting a grip on other minds: Mirror neurons, intention understanding, and cognitive empathy. Soc N e u r o s c i . h t t p s : // d o i . o r g / 1 0 . 1 0 8 0 / 17470910600985605 17. Vachon-Presseau E, Roy M, Martel M, Albouy G, Chen J, Budell L et al (2012) Neural processing of sensory and emotionalcommunicative information associated with the perception of vicarious pain. NeuroImage 63(1):54–62. https://doi.org/10.1016/j. neuroimage.2012.06.030 18. Bufalari I, Ionta S (2013) The social and personality neuroscience of empathy for pain and touch. Front Hum Neurosci. https://doi.org/ 10.3389/fnhum.2013.00393 19. Gordon I, Voos AC, Bennett RH, Bolling DZ, Pelphrey KA, Kaiser MD (2013) Brain mechanisms for processing affective touch. Hum Brain Mapp. https://doi.org/10.1002/ hbm.21480 20. Morelli SA, Lieberman MD (2013) The role of automaticity and attention in neural processes underlying empathy for happiness, sadness, and anxiety. Front Hum Neurosci. https://doi. org/10.3389/fnhum.2013.00160 21. Schirmer A, Reece C, Zhao C, Ng E, Wu E, Yen S-C (2015) Reach out to one and you reach out to many: social touch affects thirdparty observers. Br J Psychol. https://doi.org/ 10.1111/bjop.12068 22. Campagnoli RR, Krutman L, Vargas CD, Lobo I, Oliveira JM, Oliveira L et al (2015) Preparing to caress: a neural signature of social bonding. Front Psychol. https://doi.org/10. 3389/fpsyg.2015.00016

128

Connor J. Haggarty et al.

23. Essick G, James A, McGlone F (1999) Psychophysical assessment of the affective components of non-painful touch. Neuroreport 10(10):2083–2087. https://doi.org/10. 1097/00001756-199907130-00017 24. Lloyd DM, McGlone FP, Yosipovitch G (2015) Somatosensory pleasure circuit: from skin to brain and back. Exp Dermatol. https://doi.org/10.1111/exd.12639 25. Yu J, Yang J, Yu Y, Wu Q, Takahashi S, Ejima Y, Wu J (2019) Stroking hardness changes the perception of affective touch pleasantness across different skin sites. Heliyon. https:// doi.org/10.1016/j.heliyon.2019.e02141 26. Kirsch L, Krahe´ C, Blom N, Crucianelli L, Moro V, Jenkinson P, Fotopoulou A (2018) Reading the mind in the touch: neurophysiological specificity in the communication of emotions by touch. Neuropsychologia 116: 136–149. https://doi.org/10.1016/j. neuropsychologia.2017.05.024 27. Kirsch L, Besharati S, Papadaki C, Crucianelli L, Bertagnoli S, Ward N et al (2020) Damage to the right insula disrupts the perception of affective touch. elife 9. https://doi.org/10.7554/elife.47895 28. Bjornsdotter M, Loken L, Olausson H, Vallbo A, Wessberg J (2009) Somatotopic Organization of gentle touch processing in the posterior insular cortex. J Neurosci. https://doi.org/10.1523/jneurosci.0400-09. 2009 ˇ eko M, Gracely JL, Richards EA, 29. Case LK, C Olausson H, Bushnell MC (2016) Touch perception altered by chronic pain and by opioid blockade. Eneuro. https://doi.org/10.1523/ eneuro.0138-15.2016 30. Gordon I, Vander Wyk BC, Bennett RH, Cordeaux C, Lucas MV, Eilbott JA et al (2013) Oxytocin enhances brain function in children with autism. Proc Natl Acad Sci. https://doi.org/10.1073/pnas.1312857110 ´ F, Sampaio A (2020) 31. Miguel HO, Gonc¸alves O Behavioral response to tactile stimuli relates to brain response to affective touch in 12-monthold infants. Dev Psychobiol. https://doi.org/ 10.1002/dev.21891 32. Ackerley R, Saar K, McGlone F, Backlund Wasling H (2014) Quantifying the sensory and emotional perception of touch: differences between glabrous and hairy skin. Front Behav Neurosci 8. https://doi.org/10.3389/fnbeh. 2014.00034 33. Haggarty CJ, Malinowski P, McGlone FP, Walker SC (2020) Autistic traits modulate cortical responses to affective but not

discriminative touch. Eur J Neurosci. https:// doi.org/10.1111/ejn.14637 34. Hauser SC, McIntyre S, Israr A, Olausson H, Gerling GJ (2019) Uncovering human-tohuman physical interactions that underlie emotional and affective touch communication. In: 2019 IEEE world haptics conference (WHC), Tokyo, Japan, pp 407–412. https://doi.org/ 10.1109/WHC.2019.8816169 35. Sivamani RK, Goodman J, Gitis NV, Maibach HI (2003) Friction coefficient of skin in realtime. Skin Res Technol. https://doi.org/10. 1034/j.1600-0846.2003.20361.x 36. Triscoli C, Olausson H, Sailer U, Ignell H, Croy I (2013) CT-optimized skin stroking delivered by hand or robot is comparable. Front Behav Neurosci. https://doi.org/10. 3389/fnbeh.2013.00208 37. Walker SC, Trotter PD, Woods A, McGlone FP (2017) Vicarious ratings of social touch reflect the anatomical distribution & velocity tuning of C-tactile afferents: a hedonic homunculus? Behav Brain Res. https://doi.org/10.1016/j. bbr.2016.11.046 38. Haggarty CJ, Trotter PD, McGlone FP, Walker SC (2021) Children’s vicarious ratings of social touch are tuned to the velocity but not the location of a caress. PLOS One. https://doi. org/10.1371/journal.pone.0256303 39. Haggarty CJ, Moore DJ, Trotter PD, Hagan R, McGlone FP, Walker SC (2021) Vicarious ratings of social touch the effect of age and autistic traits. Sci Rep 11:19336. https://doi.org/10.1038/s41598-02198802-2 40. Croy I, Sehlstedt I, Wasling HB, Ackerley R, Olausson H (2019) Gentle touch perception: from early childhood to adolescence. Dev Cogn Neurosci 35:81–86. https://doi.org/ 10.1016/j.dcn.2017.07.009 41. Triscoli C, Ackerley R, Sailer U (2014) Touch satiety: differential effects of stroking velocity on liking and wanting touch over repetitions. PLoS One 9(11):e113425. https://doi.org/ 10.1371/journal.pone.0113425 42. Watkins R, Dione M, Ackerley R, Backlund Wasling H, Wessberg J, Lo¨ken L (2021) Evidence for sparse C-tactile afferent innervation of glabrous human hand skin. J Neurophysiol 125(1):232–237. https://doi.org/10.1152/ jn.00587.2020 43. Bensmaia S (2008) Tactile intensity and population codes. Behav Brain Res 190(2):165–173. https://doi.org/10.1016/j. bbr.2008.02.044

Chapter 7 Qualia, Brain Waves, and Spinal Reflexes: The Study of Pain Perception by Means of Subjective Reports, Electroencephalography, and Electromyography Elia Valentini, Sarah Vaughan, and Amanda Clauwaert Abstract Pain is defined as an inherently conscious multidimensional experience, which is traditionally studied using psychophysical methods. Pain threshold, rating, and tolerance are essential to the experimental and clinical assessment of pain. Yet, there is a long-standing tradition in physiological methods that complements the behavioral assessment, such as the study of spinal reflexes and brain activity. This chapter will focus on the quantitative assessment of pain experience using psychophysical as well as electrophysiological methods. We will review some of the main methodological approaches in the field and their advancements, both in healthy and clinical populations. We will also pinpoint some of the most relevant technical and ethical considerations. Moreover, researchers are invited to reflect on methodological demands and challenges that may be particularly relevant within the study of pain, such as confounding sampling variables (e.g., health status), design variables (e.g., motor arousal), and interpretational caveats (e.g., specificity of response measure) to ensure that the measurement is sufficiently robust and reliable as well as reproducible by other researchers and clinicians. Key words Electroencephalography, Electromyography, Electrophysiological methods, Nociception, Nociceptive withdrawal reflex, Pain assessment, Pain rating, Pain tolerance, Pain threshold, Psychophysics

1

Introduction The current chapter aims to provide the reader with a methodological and technical guidance through the historical and contemporary approaches to the study of pain using psychophysiological techniques, particularly those of psychophysics and electrophysiology. We strived to illustrate the most common methodologies and offer the basic information for beginners and those who will need to apply noxious stimulation in their professional work. By no means do we believe this chapter offers an exhaustive summary and description of the methodological know what and know how

129

130

Elia Valentini et al.

available in the field, and we therefore urge the reader to consult the references as well as engage in personal bibliographic research when embarking upon pain research. 1.1 Measuring Pain: A Methodological Conundrum

The study of pain has achieved a substantial sophistication over the past decades. The development of psychophysics fascinated researchers with the idea of crafting an “objective” and reproducible methodology to measure the complex, multifaceted experience of pain. This started with Hardy et al. [1], who developed the first psychophysical approach to the quantification of experience resulting from the noxious increase in radiant heat energy. Classically, the laboratory assessment of the pain threshold (i.e., painful vs. non-painful sensation) would depart from clinical qualitative assessment by administering a well-controlled quantitative protocol. The dependent variable (i.e., pain response) is conceived as the probability of judging a given stimulation as painful, thus entailing a mean stimulus intensity at which participants most likely feel pain, and a variability expressing the uncertainty of this relationship. In addition, researchers can collect reaction times of these responses, adding a “mental chronometry” nuance to the behavioral assessment. Further developments ensued from the application of signal detection theory (SDT) in the study of visual, auditory, and tactile perception [2]. Pain researchers attempted to characterize the sensation originating from thermal noxious stimuli using a range of SDT procedures [3, 4]. Classically, SDT entails two main measures, the sensitivity (d-prime or d′) and response bias (criterion or C). These are measured during judgment tasks whereby the participant is asked to either detect a stimulus or decide if two stimuli are different on a sensory dimension (e.g., intensity). In the context of pain, researchers ask participants to judge whether a stimulus is painful or non-painful. Thus, “hits” indicate correct classification of high-intensity stimuli as painful; “misses” indicate misclassification of high-intensity stimuli as non-painful; “false alarms” consist of low-intensity stimuli classified as painful; and “correct rejections” consist of low-intensity stimuli classified as non-painful. Importantly, both d′ and criterion are computed using the proportion of hit (hit rate; H) and false-alarm trials (false alarm rate; F, see also Chapters 1 and 20, this volume). In addition, researchers are often interested in knowing not only if a participant can detect and discriminate noxious stimuli, but also how much pain they felt and how unpleasant it was. This information is routinely acquired by means of direct magnitude estimation, using a pain rating scale [5]. This may require a verbal or written, numerical, or graphical procedure. For example, the visual analog scale (VAS; usually a 100 mm line) invites the participant to click on a horizontal (or vertical) bar on a computer screen or use a slider to move from a

Psychophysics and Electrophysiology of Pain

131

bottom extreme (e.g., 0, “no pain”) to a top extreme (e.g., 100 “the worst imaginable pain”). The combination of measurement methods is paramount in the context of pain research, for pain is an integrative phenomenon involving different processes and physiological systems at different temporal scales. To complicate matters, there is no direct relationship between nociception (i.e., the neural process of encoding noxious stimuli) and the experience of pain [6]. Crucially, researchers cannot directly measure the activity of nociceptors in humans, thus they rely mainly on indirect measures of nociception and subjective measures of perception. Therefore, researchers have used a range of different techniques, from skin biopsy to very advanced neuroimaging technology. Neuroimaging has been widely applied to the study of pain in the attempt to identify some sort of “pain center” or “pain network” within the brain (see Chapters 17, 18, and 19, this volume). However, research found that innocuous but salient stimuli can activate the brain similarly to painful nociceptive stimuli, thus impairing researchers’ ability to identify pain-specific physiological responses [7, 8], and even more for pain biomarkers [9]. Nonetheless, more advanced computational and integrative multimodal approaches could still provide researchers with a better understanding of the psychophysiological mechanisms of pain [10, 11]. This chapter will focus on the combination of the psychophysical methods with two main electrophysiological approaches to provide a better assessment of pain perception across different individuals, namely electroencephalography (EEG, see also Chapter 19, this volume) and electromyography (EMG, see also Chapters 6 and 20, this volume). Both methods index the activity of the nervous system and can be used either in the experimental or clinical setting. For example, Hardy showed that both the verbal report of pain and the nociceptive flexion reflex (recorded by means of EMG) can converge in substantiating a skin temperature of around 44 °C as a threshold for heat pain [12]. The use of nociceptive reflexes has the advantage of bypassing the confounding effect of response biases. However, it has become clear that, despite being an optimal measure of defensive spinal mechanisms [13], they are subject to substantial descending cognitive modulation (see Note 4.1). In fact, as demonstrated by a recent meta-analysis, the activation threshold for nociceptive reflexes is lower in clinical pain, but their size, onset latency, and duration do not seem to differentiate between patients and healthy individuals. The authors therefore concluded that EMG measures following spinal reflexes are a useful measure of defensive arousal within the central nervous system, but cannot be considered a biomarker of pain [14]. The same consensus has been reached about EEG responses to nociceptive and/or painful stimuli [15].

132

Elia Valentini et al.

1.2 Methodological Developments

We regret to disappoint the reader that, to date, pain remains impenetrable due to its inherently qualitative dimension. In this respect, the distinction between pain and nociception is important. While pain is a conscious multidimensional experience, nociception is the physiological encoding and processing of potentially noxious stimuli and does not require consciousness to take place. Crucially, not all noxious stimuli activate nociceptors or cause pain (e.g., radiation or carbon monoxide). In addition, both pain and nociception may dissociate and occur separately [16]. Moreover, pain can originate from either selective activation of nociceptive fibers, or from (co-)activation of non-nociceptive fibers. One main distinction can be used at a somatic level: researchers can tell apart the involvement of Aβ, Aδ, and C fibers in their separate somatosensory functions. The large-diameter myelinated Aβ fibers segregate within the dorsal column—medial lemniscal system, and convey information about mechanical stimulation of the skin and kinesthetic signals from the limbs that reach the brain and eventually contribute to touch perception and limb proprioception. The small-diameter myelinated Aδ and unmyelinated C fibers travel (mainly) through the spinothalamic tract—anterolateral system, and transfer polymodal information about thermal, mechanical, and chemical energy, which eventually contribute to the perception of itch, temperature, internal homeostasis, and pain (see Chapters 8 and 9, this volume). According to the differentiation of nerve fibers, researchers can administer different stimuli to activate different fibers and rely on the aforementioned methods to gather useful information about pain correlates (Table 1). Some psychophysical methods are fundamental in the clinical setting where a quantification of sensory dysfunction is paramount. Quantitative sensory testing (QST) precisely applies a series of different tests to evaluate central and peripheral nervous pathways that quantitatively assess sensory dysfunction (Fig. 1). Traditionally, sensory assessment in the clinic involved manual administration of brushes, cotton buds, sharp objects, and objects varying in temperature, on the skin of the patient according to a specified protocol. This approach is nowadays improved using devices able to deliver well-controlled and automated series of mechanical and thermal stimuli. A machinecontrolled QST renders the assessment more valid and reliable in the clinical setting, while dramatically extending the experimental reach in the study of nociception and pain [17]. Clinicians and researchers can measure detection threshold, pain threshold, and pain tolerance, regardless of whether they are applying simple mechanical stimuli exerting variable pressure, or complex lab-protocols delivering repeated thermal stimuli in specific patterns of duration and intensity. In the clinical setting, the use of graded intensity and temporally precise stimuli grants a quicker and standardized evaluation that aids the distinction between

Psychophysics and Electrophysiology of Pain

133

Table 1 Main types of controlled nociceptive stimulation/pain induction approaches for experimental and clinical purposes Stimulation type/ modality

Duration Sensory fibers

Devices

Thermal Laser emitted Tonic/ Aδ (initial prick) & radiant phasic C (burning sensation) heat

Carbon Dioxide (CO2) laser 10,600 nm, infrared CO2 laser stimulator (e.g., Laser Stimulation Device, SIFEC), Thulium (Tm) Fiber Laser Stimulator 2000 nm: see [47–50], ElEn YAP, Erchonia

Tonic/ Aδ (rapid skin heating) QST.lab, TSA2, Medoc Sense, Medoc Pathway Contact phasic & C (slow heating) thermal stimulators Cold and hot Tonic water baths

Aδ & C-fibers

RW-3025P Refrigerated & Heating Bath Circulator Programmable 30 L, Dancer Designs or a circulation pump: Reich 10 L/Min 0.5 bar

Pinprick Phasic stimulators

Aδ

MRC Systems

Pressure Tonic stimulators

(Aβ) Aδ & C

Nocitech, Somedic, ALGOMED, Wagner instruments, Itech medical

Mechanical

Electrical

Phasic/ Aβ & Aδ (low intensity) STMISOLA Biopac, Digitimer (DS7AH, DS5) Tonic & C (high intensity)

Chemical

Tonic

C

Capsaicin: topical or intradermal, or Hypertonic Saline PCCA Ltd

nociceptive and neuropathic pain [18]. Such a distinction can also benefit from the use of EEG and EMG recording. In particular, EEG potentials evoked by laser stimulation of the skin provide a useful tool for the clinician to assess the cortical representation of Aδ and (when possible) C fiber transmission [19]. For example, laser-evoked potentials (LEPs) were applied successfully to the assessment of nociceptive transmission in a patient with neuropathic pain who underwent brachial plexus avulsion [20]. LEPs were nearly abolished as a result of stimulation of the painful dermatome (C6), and were lower in amplitude following stimulation of the contiguous affected but non-painful dermatome (C5), as compared to the contralateral unaffected dermatome [21]. Although LEPs proved useful in the study of pain, as the radiant heat energy allows a selective and synchronous activation of cutaneous nociceptors [22], researchers and clinicians understood how synchronized EEG responses elicited by nociceptive

134

Elia Valentini et al.

Fig. 1 Example of stimulations implemented within standard Quantitative Sensory Testing (QST). (a) Thermal testing which comprises detection and pain thresholds for cold and warm or hot stimuli. (b) Mechanical pain threshold and wind-up ratio to pinprick stimuli. This can also be incorporated into dynamic mechanical allodynia and stimulus response functions with additional tactile stimulators (cotton wisp, Q-tip, and sensory brush). (c) Pressure pain threshold or tolerance testing utilizing an algometer, the only test for deep pain sensitivity. (d) Vibration detection threshold testing utilizing a tuning fork. (e) Electrical pain threshold and tolerance determination. (f) Cold or hot water bath for the determination of pain threshold and tolerance

and painful somatosensory stimuli were strongly influenced by the level of vigilance, alertness, and attention [19]. Even more, growing evidence indicated that the magnitude of activity recorded by neuroimaging techniques [8] including EEG could be dissociated from the intensity of pain perception (i.e., from subjective experience). Over the last ten years, experimental research on the neural correlates of pain has shifted toward a greater focus on identifying neural markers of the experience of pain by enhancing the focus on repeated [23], prolonged [24], and inflammatory [25] pain stimulation, as well as implementing advanced computational approaches to data analysis [26]. However, these more sophisticated methods, while being difficult (if not impossible) for the clinician to apply, still rely on the subjective ratings of pain to determine their validity as “objective” measures of pain perception.

Psychophysics and Electrophysiology of Pain

1.3 Experimental Versus Clinical Pain

135

While experimentally induced pain is a valuable tool to study some of the dynamics of acute pain, it is not sufficient for a characterization of chronic pain mechanisms. Note, however, that experimental pain can never be directly compared to the experiences of pain patients, as clinical and experimental pain often rely on the activation of different peripheral nociceptors and neural pathways. Nevertheless, experimentally induced pain can provide valuable information on how pain is processed and what behavioral responses are associated with it, in both healthy and clinical populations (see Note 4.2). Moreover, acute pain can be provoked in therapeutic and training settings in the context of developing coping skills in the patient or client who is confronted with a painful condition [17], as in pain exposure therapy [28]. Importantly, responses to experimental pain can be predictive of postoperative pain experiences (see [29] for a systematic review). The main advantage of the laboratory setting is that it allows standardization of the intensity of the nociceptive input, and to control other variables [30]. However, the controlled environment of laboratory pain can be problematic in terms of ecological validity. For example, participants know that the unpleasant sensations are (somewhat) more predictable, no actual tissue damage will occur, and that they always have the possibility to stop the pain [27]. These factors are also typically disclosed to the participants prior to any experiment and conform to the ethical guidelines which will be discussed later in this chapter (see Note 4.3). Furthermore, the experience of pain is a complex interplay between many factors, and is shaped by our prior experiences [31]. As a result, the behavioral and neural mechanisms that are triggered by experimental pain are not necessarily the same as those that have been shaped by long-term pain experience. For example, individuals with chronic pain may have adopted new, but perhaps maladaptive behaviors (accompanied by changes in the underlying brain processes), to cope with the continuous presence of pain. Similar to what Arendt-Nielsen and Andersen emphasized [32], only a multi-modal and multi-assessment approach can yield the necessary (but perhaps not sufficient) sensitivity to detect pharmacologically and clinically relevant effects. To conclude, while experimental pain in clinical and pharmacological contexts can lead to both theoretical and applied progress, they can never be directly compared to naturally occurring pain. Therefore, it is key to consider the best paradigm for your research question, to use more than one pain-induction and recording technique, and to keep up to date with consensus-based theoretical and empirical models for interpreting the functional significance of the responses you collect.

136

Elia Valentini et al.

1.4 Application to Different Populations

2

Experimental pain research has provided better understanding of pain in special populations [33–39] and has contributed to greater understanding in the complex relationship between pain, depression, anxiety, and sleep disorders [40–42], above and beyond the more commonly investigated individual differences in age and sex [43] as well as gender and race [44]. A special population (e.g., children, or individuals suffering from psychiatric or medical conditions) could be thought of as any given group of people whose defining characteristics require special or careful consideration and attention when designing an experimental pain procedure. For example, an older adult with age-related visual, auditory or cognitive impairment may find it difficult to maintain concentration for a full experimental protocol; they may be unable to see or hear instructions and questions (see Chapter 12, this volume). Therefore, both the experimental method and the tool being used to measure the magnitude of pain must be carefully considered (see Note 4.4, [45]). If not, these factors can hinder completion of assessment protocols and affect responses, besides causing unnecessary discomfort. In such cases, parent or carer report, recording behavioral responses, and/or vocalizations are recommended. One study which highlights these modifications used pain behaviors, namely vocal, facial, and body movements, and the faces pain scale-revised, when taking responses to pressure pain in adults with moderate Alzheimer’s [46]. Additionally, testing was completed in the care home rather than a laboratory to facilitate care of the individuals and reduce anxiety. Pain assessment in special populations is therefore an interplay of the population characteristics, the setting, and the assessment and measurement techniques being used, more so than in other populations.

Materials

2.1 Psychophysics and Behavioral Responses During Pain Assessment

In general, studies of pain require application of a nociceptive stimulus (or a supra-threshold stimulus in a different modality) to create an experience of pain. The experience obtained because of controlled stimulation can be evaluated and measured in several ways, for example, ratings or questionnaires. Physical units, ratings, and responses such as withdrawal of the hand from the highest tolerable hot water (i.e., pain tolerance) can all inform the assessment of the pain experience. This section outlines the equipment and materials most used in both experimental and clinical studies of pain. It then provides two more noted examples on how these materials may be combined.

Psychophysics and Electrophysiology of Pain 2.1.1 Stimulation Apparatus

137

Controlled stimulation is achieved using hardware able to control the physical parameters of stimulation over time (e.g., mA, KPa, °C). • Laser-emitted radiant heat stimulators • Contact thermal stimulators • Cold and hot water baths • Mechanical pinprick stimulators • Mechanical pressure stimulators (also known as algometers/ dolorimeters) • Electrical stimulators • Chemical stimulators

2.1.2 Recording Apparatus and Consumables

Each of the following materials is either equipment that can be used with the stimulators, or is consumable required to run the hardware. • Biosignal recording systems for physiological readings (e.g., BIOPAC, AD Instruments, CED) such as skin conductance • Sphygmomanometer for systolic and diastolic blood pressure • Electrodes • EEG cap • Conductive gel • Sterile syringe (0.3 mL) • Double-sided adhesive stickers • Tape measure—to measure the scalp

Software

There is a great deal of software available for the preparation and presentation of stimuli. For example, companies selling the stimulators also provide programs that ensure precise and timely delivery of stimuli. Crucially, this software is often easily connected with the recording apparatus, and allows the synchronization of, for example, EEG and psychophysical responses with the onset of the physical stimulus using time stamps. Clearly, some studies may require more advanced stimulus and response patterns and procedures, which are usually set out and executed through custom software based on major computing platforms such as Matlab (MathWorks Inc.) or LabView (National Instruments).

2.1.4 Ratings and Behavioral Responses to Pain

Individuals’ judgments about their own perception are the primary measure of pain. Numerical, visual analog, and verbal rating scales (NRS, VAS, VRS) can provide a quantitative estimate of the severity or magnitude of the subjective experience both in the experimental and clinical setting [47, 48]:

2.1.3

138

Elia Valentini et al.

• Numerical pain rating scale (NRS) • Visual analog scale (VAS) • Verbal rating scale (VRS) 2.1.5

Questionnaires

Pain questionnaires can provide a quantitative estimate of components of pain or a measure of a specific experience [49, 50]. Here are some of the most used instruments: • McGill pain questionnaire [51] • Brief pain inventory [52] • Pain catastrophizing scale [53] • Fear of pain questionnaire [54] • Pain anxiety symptoms scale [55] • The Leeds assessment of neuropathic symptoms and signs pain scale [56] • Regional pain scale [57] • Neuropathic pain scale [58]

2.1.6 Observational Measures

In some contexts, for example, when working with special populations, you might need to rely on quantitative measures of observed pain behaviors, including verbal and non-verbal expressions [59–61]: • Pain behavior checklist [62] • Non-communicating children’s pain checklist [63] • Observational scale of behavioral distress [64] • Facial action coding system [65] • Child facial action coding system [66] • Dalhousie everyday pain scale [67]

2.2 Example Material 1: Transcutaneous Electrical Stimulation

Electrical stimulation can typically be performed on a range of skin sites, although it can also be performed on intracutaneous tissue, muscle, or the viscera [68, 69] (see also Chapters 8, 15, 16 and 20, this volume). High voltage (HV) current stimulators such as the DS7AH (Digitimer), which allows currents up to 1A with a maximum pulse duration of 200 μs, can be used to generate stimuli. Such devices can deliver a range of stimulation patterns though unipolar constant current pulses with a rectangular shape, which is the most common stimulation pattern [68]. Electrodes are chosen dependent on the site being tested. For electrical stimulation within the muscle, for example, small needle electrodes with uninsulated tips would be used [70]. Conversely, surface electrodes are placed on skin sites with conductive gel through which the current passes

Psychophysics and Electrophysiology of Pain

139

[71], activating Aβ fibers at the lowest currents. Increasing the current leads to activation of Aδ and eventually C fibers at higher intensities. Of note is that different stimulation methods and patterns activate different fibers (see [72] for triangular pulses as an example and [73] for intra-epidermal preferential activation of Aδ fibers). As well as manual operation, most stimulators can be paired with software for controlled command options, such as E-prime, via either parallel port, serial, or USB control. Outcome variables for pain could be obtained using NRS, VAS or VRS, or electrical threshold can be measured as a physical unit (mA). 2.3 Example Material 2: Capsaicin Application

Intradermal, intramuscular, or topical capsaicin (the compound naturally found in chili peppers [6]) is a chemical method of nociceptor activation [74]. This model is interesting in that it produces primary heat hyperalgesia, secondary mechanical allodynia, and hyperalgesia, which are unique and not evoked by the other methods described here. Capsaicin injections of 10 or 100 μg have been shown to produce robust pain experience, although 100 μg is most commonly used [75, 76]. There are two formulations for Capsaicin injections: Tween and Hydroxypropyl-β-Cyclodextrin (HP-β-CD). These are classes of emulsifiers and are biocompatible surfactants whose job is to improve the solubility and skin penetration of capsaicin. Formulations have been shown to be comparable in a dose range of 1-30 μg, but at the more common 100 μg, HP-β-CD is preferable [77]. Outcome measures include NRS, VAS, or VRS, and can be taken at different time points. Capsaicin can be paired with other hardware, for example, pinprick stimulators, producing physical units of measurement for the paired hardware—in the case of pinpricks, in micro newtons (μN).

2.4 EEG and EMG Responses During Pain Assessment

We have discussed a wide range of approaches to nociceptive stimulation and pain induction. Most of these allow researchers to complement the psychophysical assessment with recording of EEG and EMG. It is recommended that readers would make themselves familiar with guidelines and recommendations for the appropriate use of both techniques [78, 79]. Concerning the recording of defensive reflexes, the preferable approach to the stimulation of peripheral nerves is by means of electrical stimulation of the skin to activate a specific muscle group. This type of stimulation requires technical expertise but is generally safe and easily reproducible. The most studied reflex in pain research is the nociceptive withdrawal reflex (NWR), which is a polysynaptic spinal reflex elicited by the activation of (mainly) Aδ afferents conveying into the ipsilateral sural nerve and monitored with EMG over the biceps femoris muscle. Practically, the experimenter is required to apply one electrode over the biceps femoris muscle of one leg, 10 cm superior to the popliteal fossa, and a reference electrode over the lateral protuberance of the femur.

140

Elia Valentini et al.

Standard commercial stimulators can then be used to apply direct current over the retromalleolar pathway of the sural nerve of the leg [80]. Co-recording of EEG and NWR is rare but can be achieved [81].

3

Methods

3.1 Screening and Instructions 3.1.1 General Recommendations

3.1.2

Clinical Populations

As pain studies may not be attractive for study participants, researchers often use cover stories to avoid reluctance to sign up an experiment. For example, as fear of pain can be an important participant characteristic, one does not want to recruit only the participants that do not experience any. This deception, however, should be approved by the ethics committee, disclosed immediately upon arrival at the lab, and participants should still have the choice to opt out of the experiment (see Note 4.3). In addition, to decrease the threatening value of the stimulus, one might opt not to refer to the painful stimulus as “painful”, but rather as “unpleasant” during the study. Unpleasantness is the crucial motivational aspect of the multidimensional experience of pain. Concerning recruitment, there is substantial inter-individual variability in the experience of experimental and clinical pain, depending on participant characteristics [82]. A standard EEG study in a laboratory pain study would involve having the participants being screened for health and safety purposes and to ensure that they have no history of neurological, psychiatric, or pain disorders that could interfere with the study [83]. Individual differences can be found on biological levels, such as genetics, but also on behavioral levels, such as pain expression. In general, research shows that female participants, for example, are more susceptible to pain than men [84], and women tend to report higher pain intensities and express lower levels of tolerance [85]. As a result, women may respond differently to pain treatments in comparison to men [86]. Moreover, differences in experimental or clinical pain have also been reported across different ethnicities and races and different age groups [87–89]. Importantly, biases may arise when assessing and treating pain in different populations [90, 91], often based on stereotypical attributions [92]. Researchers should therefore be aware of any possible biases, as well as to carefully select and balance their study sample, depending on the research question (see Note 4.5). Clinical pain patients may differ in several aspects of their experience. For example, there is a substantial range in the magnitude of chronic pain. When recruiting patients, you can consider recruiting participants directly from the general population (e.g., via flyers) or via clinical settings (e.g., specialized medical centers). However, it is possible that those referred by a clinician may report more severe

Psychophysics and Electrophysiology of Pain

141

pain symptoms than individuals recruited from the general population, who in turn may not have sought specialized help. On the contrary, a convenience sample through public adverts may expose the researcher to self-selection of candidate participants who are interested in the study for specific reasons (e.g., financial retribution, promise of a treatment, psychological factors). In addition, chronic pain conditions can often be sub-classified depending on the occurrence of the symptoms. For example, in the case of chronic low back pain, researchers are now considering the existence of an intermediate clinical subgroup—a recurrent low back pain group— of which the symptoms are considered as less severe in comparison with the chronic low back pain population [93]. As a result, before embarking in recruiting clinical samples, it is highly recommended to use well-defined pathophysiological categories of pain, and to thoroughly assess participants’ pain symptoms (e.g., occurrence, duration, severity) and comorbidities. 3.2 Pain Assessment and Calibration

While with some nociceptive stimulation such as contact-heat temperature, scientists may apply standard settings for every participant, some other stimulation approaches may need to be calibrated to the experience of the participant. This is particularly tricky as the pain intensity should be tolerable for the participants, and the tolerance level for each individual may differ greatly. Moreover, there is no one-to-one relationship between the nociceptive input and the perceived pain, and as a result, participants do not necessarily rate a specific pain intensity as equally painful [94]. Researchers should therefore wisely consider how they want to go about the intensity of the stimulation in their study (see Note 4.6). Does the study benefit from an intensity calibrated around participants’ pain thresholds, or instead would it rather be established according to the physical parameters expected to cause similar perceptual experiences in the participants? In other words, the researchers may want to ask if they are more interested in a stimulus pattern that elicits similar levels of threat, pain, or discomfort across participants, or if they want to evaluate differences in response to specific nociceptive input. In most cases, to deal with perceptual differences, researchers and clinicians investigate the pain threshold, which is typically determined for each individual participant right before the start of a study, to select the stimulus intensities to be used in the study itself. For example, researchers would present a series of thermal stimuli of varying intensity and ask the individual to rate or judge the sensations evoked by each stimulus. First, the absolute threshold can be obtained by repeatedly presenting a series of fixed stimuli of different temperature in random order and asking the subject to report if they detect it or not (see also Chapter 9, this volume. The sensory threshold is determined as the value that was detected 50% of the time. Then,

142

Elia Valentini et al.

a pain threshold can be examined using the very same approach (i.e., method of constant stimuli) and ask participants to rate each stimulus using, for example, a VAS ranging from “not painful at all” to “extremely painful” [95]. Alternatively, the methods of limits can be used, which entails the delivery of alternating increasing and decreasing intensity series of stimuli. Pain researchers may use only ascending series to reduce participants’ discomfort associated with surprising intense stimulation. However, they may vary the starting intensity of a series and the step size between sensory events to reduce response bias [18]. More advanced approaches may mitigate response bias using adaptive staircase procedures whereby an algorithm will vary the intensity increase or decrease step size of the forthcoming stimulus on the basis of the participant’s prior response(s) [96]. A further approach is to let the participants determine the intensity of the stimulus. For example, the researcher presents the participants with a low-intensity stimulus and asks them to increase the intensity until they have reached their personal tolerance limit or a lower predefined value [97, 98]. 3.3 Nociceptive Stimulation and Pain Induction

This section gives an overview of most of the nociceptive and pain induction methods shown in Fig. 1 (see also Table 1). Before starting an experiment, researchers should ask how many and which noxious stimuli are needed (e.g., thermal, mechanical, electrical) and if sub-, near-, or supra-threshold stimulus intensities are required. The answer depends on the aims of the study. With that in mind, experimental may differ from clinical aims. For example, a clinician may be interested in peripheral and/or central sensitization as well as the type of tissue investigated (cutaneous, muscle, and/or viscera). They might be exploring clinical pain states including allodynia (pain to otherwise innocuous stimuli). In this instance, stimulus response functions such as that measured using pinpricks and as described below may be more informative (Fig. 1b). Therefore, careful consideration of the technique being adopted is important. Cutaneous thermal stimulation can be achieved via contact heat stimulation (Fig. 1a), radiant heat via lasers (described in other sections of this chapter), and cold or hot water baths (Fig. 1f), in which participants immerse a part of their body in water maintained at a predefined temperature. Contact heat can be applied to a region of choice (e.g., the dorsum of the hand or foot) via a thermode of a specified size, using the method of limits (e.g., ramping of the temperature until the participant reports pain) or an adaptive staircase as discussed in the previous section. Mechanical stimulation can be achieved by application of a rigid thin filament or metal rod on the skin (Fig. 1b), or on the muscle via pressure algometry (Fig. 1c) and a tuning fork (Fig. 1d) to assess

Psychophysics and Electrophysiology of Pain

143

the function of Aβ fibers. Electrical stimulation can be adapted for all tissue types and body parts, although the method shown in Fig. 1e gives an example of cutaneous application above the median nerve. The tests can be performed in multiple areas and repeated several times. If adopting QST, it is recommended that researchers make themselves aware of the published guidelines generated by the Neuropathic Pain Special Interest Group (NeuPSIG) of the International Association for the Study of Pain as there were designed with both clinical and research application in mind [99]. 3.4 Psychophysics and Behavioral Responses

The simplest approach to the quantification of pain experience is the pain rating. Rating scales vary in format, acquisition modality, and intervals (see Subheadings 1.1 and 2.1.4) but they all provide researchers with at least an ordinal, or even with a ratio scale measurement [100] that allows researchers to compute parametric statistics to analyze differences between conditions. Nevertheless, whether the researcher is going to apply parametric or non-parametric statistics says nothing about the latent theoretical framework used to interpret the mechanisms of perceptual decision. Historically, signal detection theory (SDT) proved successful in quantifying perceptual judgments in auditory and visual modalities (Subheading 1.1). Its former application to pain seemed to be able to disentangle information about the sensory experience and the report of pain [101, 102] (see Note 4.7 for further critical considerations). In a simple binary or alternative forced-choice model, high and low-intensity stimuli (e.g., a painful and non-painful stimulus) are presented randomly to participants. Presentation can also be at two locations or in one of two temporal intervals during a trial. The task of the participant is to decide the classification of the information sampled (i.e., the stimulus presented was high or low, painful, or non-painful, left or right) and to provide a response. According to the theory, noise is embedded within the decision and in a simple binary model such as this, there are then 4 possible classifications of a response. More than two intensities can be used, but this model provides ease of understanding. The behavioral measure of reaction time latency is an additional useful method to measure pain indirectly, particularly as it has been shown to be related to stimulus intensity [103]. Reaction time is the interval between the application of a stimulus and the objective response of interest. Its measurement is straightforward when tightly time-locked stimuli are delivered under computer control. In its simplest form, these tasks require participants to make a response, for example, a movement or the pressing of a key, whenever any stimulus is presented. The main dependent measure is then the response speed. This methodology has been extensively used in the assessment of clinical pain states [104]. However, there are multiple techniques in which response time can be used. The

144

Elia Valentini et al.

following is based on the procedure described by Mouraux & Plaghki [105] in which response time was used for determination of individual threshold estimates, as well as to measure the response to co-activated Aδ and C-nociceptors to investigate early and ultralate LEPs. Heat stimuli were delivered to the dorsum of the left foot using a CO2 laser in two conditions (“first pain” representing Aδ activation and “second pain” representing C-nociceptor activation). Each condition contained 6 blocks of 10 trials with an interstimulus interval randomized between 30 and 40 s. In one condition, participants were asked to pay attention to the “first feeling” of pain, while in the other condition they were asked to pay attention to the “second pain”. Not only were response times measured to determine the point at which this first and second pain were detected, but participants were also asked about intensity and quality of sensation (e.g., itchy, burning, tingling). This procedure highlights the utility of the combination of measurements. Specifically, the inclusion of a qualitative component can help achieve further understanding of the subjective experience associated with the nociceptive stimuli above and beyond the quantification of pain. 3.5 EEG and EMG Recording

The EEG set-up for a pain study is akin to methods for touch (see Chapter 19, this volume). However, it is important to note that in the context of pain studies, a clean signal is paramount, as additional sources of noise can be generated by voluntary muscle activity during nociceptive stimulation. In this respect, establishing a positive rapport with the participant is critical, as their compliance can significantly improve the quality of the recording and the overall goodness of the specific performance required during any of the aforementioned stimulation contexts. We suggest the experimenter or clinician is proactive in ensuring the participant will move into a position that is most comfortable and that they carefully attend to the instructions. For other technical information about EEG recording neurophysiology, hardware, and software see Chapter 19, this volume. If the study requires a fine-grained characterization of the regional distribution of electrical activity on the scalp, we recommend an EEG set-up with more than 60 scalp electrodes. Pain psychophysiology and neuroscience is increasingly allowing (and requiring) more data gathering, and more complex data analysis as a result. However, the researcher or clinician must be conscious of the type of question to be addressed and down- or up-scale the EEG recording system accordingly. For example, the assessment of small fiber function would not require a high-density EEG system, whereas EEG network inference analysis of different pain-related perceptual states would certainly require a higher density set-up.

Psychophysics and Electrophysiology of Pain

145

We specified above the procedure to implement the recording of the NWR (Subheading 2.4). In comparison to EEG, the set-up for the recording of peripheral spinal reflexes via EMG is simpler and less variable. However, although the classical approach to elicit spinal reflexes would require placing the stimulating electrode at the ankle and the recording electrode at the distal portion of the ipsilateral biceps femoris muscle [106], there is indication that an equally or more effective approach would involve electrical stimulation of the sural nerve via the sole of the foot, and recording the response of the ipsilateral tibialis anterior muscle [107]. A similar approach can be used even in infants [108]. 3.6

Data Processing

3.7 How to Conduct a Short-Lasting Acute Pain Study

Most of the data acquired with the methods discussed here would not require intensive computational processing or extensive resources. However, advanced statistical analyses of psychophysical and even more physiological data might require some computational crunching. We recommend the reader becomes familiar with mixed models (particularly for repeated measures designs) [109, 110] and with Bayesian approaches [111, 112] to the analyses of all the dependent variables involved in pain studies. The advantages associated with these approaches are widely discussed elsewhere. Specific considerations for physiological recordings are also due. The advantage of having a good scalp electrode coverage is to be able to address the topography of both phase- and non-phaselocked EEG responses by computing either standard averaging of time-locked deflections and time–frequency decomposition of EEG oscillatory activity associated with painful stimulation. In addition, high-density coverage allows average reference of scalp activity, which is a necessary step for proceeding to source (or generator) level analysis (see also Chapter 19, this volume). The following example is based on the procedure described in [113], in which the authors examined how nociceptive laser heat pulses are processed in chronic low back pain patients in comparison with healthy controls. The room was temperature controlled (21–23 °C), and the participants and investigators were acoustically isolated and wore protective goggles. The authors assessed individual pain thresholds by presenting stepwise increasing laser pulse intensities (according to the method of limits) that the participants were asked to categorize as a sensation in the range between “just perceived” to “very severe pain”. The mean of the intensities rated as “medium pain” was selected to use during the experiment. This resulted in an intensity that felt like a pinprick sensation. Note that, as a result, each participant received a different stimulation level, but provided a similar rating of their sensations.

146

Elia Valentini et al.

During the experimental procedure, participants underwent 2 blocks of 30 laser stimuli each. In the first block, all stimuli were presented on the painful location—the lower back—whereas in the second block, all stimuli were presented on a non-painful location, the abdomen. Importantly, to minimize sensitization (a reduced pain threshold) or habituation (a higher pain threshold) of the nociceptors due to the many painful stimuli, the researchers slightly displaced the laser beam on the skin after each stimulus. Moreover, between each stimulation, there was an inter-trial interval of 20–25 s. This is a long interval for most experiments, but as a rule of thumb the reader may want to consider a minimum of 5 s as an appropriate interval to avoid sensitization, and to ensure a variable interval is used to minimize habituation, unless a constant repetitive or steady-state stimulation is required. However, see elsewhere for indications on nociceptive-related steady-state stimulation approaches [23]. 3.8 How to Conduct a Prolonged Acute Pain Study

The following example is based on the procedure described in [24], whereby the authors examined ongoing pain and EEG activity during tonic painful heat stimuli delivered to the left-hand dorsum of 41 healthy participants for a duration of 10 min. For the responses, the authors used a potentiometer apt to allow the participant to continuously rate pain from no pain to worst tolerable pain (0–100 VAS) using a slider with their fingers. Participants and investigators were acoustically isolated and wore protective goggles. The study included a phasic pain condition (not reported here) and did not rely on individual pain threshold assessment but rather used a predefined starting temperature for the tonic stimulation. Importantly, the authors developed software able to adjust the stimulator intensity online to match the individual pain rating profile in order to keep the felt pain in a VAS range between 30 and 70. Another procedural detail of this study highlights the importance of thinking carefully about the design and paradigm with respect to the data analysis. The initial increase and final decrease of stimulus and pain intensity were not included in the analysis to maximize relationship stability between the two variables (Fig. 2b). The design implied a visual control condition meant to separate out sensory, motor, and attentional modulations associated with the rating process, rather than with the experience of pain itself. Participants were instructed to continuously rate the length of a vertical red bar while no painful stimulation was applied. The data analysis entails substantial detail for which we refer the reader to the original paper [24]. We emphasize here that linear mixed models were fitted to the data to capitalize on the role of within- and between-subjects variability in accounting for the relationship between EEG oscillatory activity and continuous pain rating.

Psychophysics and Electrophysiology of Pain

Output

100

Pain intensity Stimulus intensity

80

45 44.5

60 40

44

20 0 0

1

2

3

4 5 6 Time (min)

7

8

9

43.5 10

Stimulus intensity (°C)

Pain intensity (VAS)

Stimulation

B 4

147

84 Hz

5 2 0 (t) (t) 0 –2

Pain intensity

–5

Stimulus intensity

15 Hz

–4 0

20

40 60 80 Frequency (Hz)

100

Fig. 2 Tonic heat pain testing example. This graphical sketch is according to the tonic pain study reported in [24]. Outcome and stimulation graphs are adapted from the cited paper

4

Notes

4.1 The Role of Cognitive Factors and Their Impact on Design and Interpretation

Researchers should be aware of the current theoretical consensus on the functional significance of brain responses recorded during pain experiences and/or elicited by nociceptive stimuli. These represent a combination of affective and cognitive processes that are not causally linked to pain but rather generic, shared with other sensory modalities and mental functions. It is argued that these processes contribute to the detection of the painful stimulus and motivate adaptive responses [15]. While doing so they modulate

148

Elia Valentini et al.

the experience of pain [126, 127] as well as the magnitude of brain responses [128]. The influence of attention is probably the most commonly studied cognitive aspect in the processing of pain stimuli [129]. Attention operates by re-weighting the importance of all incoming information and selecting the information that is relevant for the cognitive and behavioral goals [15]. This can be achieved via two different processes. First, since pain is a motivationally relevant state, painful stimuli will involuntarily draw cognitive resources toward, for example, the location of pain (i.e., bottom-up attention). As a result, such redirection of resources can interfere with other ongoing tasks [130]. It is postulated that the ability of painful stimuli to capture attention depends on their physical salience (including novelty and predictability), and relevance to the participants’ goal or internal regulation, rather than their painfulness or sensory modality [8]. Second, we can also actively guide our attention, and typically, we tend to actively attend to, or look for, cues that are related to threat (i.e., top-down attention), according to our contextual motivations and goals. It has been argued that an extreme form of this top-down attention may maintain clinical pain symptoms [131]. By contrast, researchers are currently exploring the possibilities of using top-down attention to modulate the responses to pain, for example, by distraction or sensory monitoring [132–135]. With this in mind, researchers should at all times be aware that the use of painful stimuli will automatically trigger co-occurring, additional, and potentially confounding (psychological) factors that may not always be the focus of the research question. As a result, one should always carefully consider what exactly the observed brain responses may represent. In addition, while it is not always possible to exclude or manipulate these confounds, they should be addressed by including, for example, appropriate control conditions in an experimental design. 4.2 Design and Procedural Issues

Design and procedural considerations vary substantially between an exploratory laboratory study on healthy individuals and a confirmatory clinical trial on patients. Concerning the latter, it is even more crucial that methodological heterogeneity is reduced in favor, for example, of standardized participant selection, trial procedure and duration, treatment groups, and dosing regimens, as recommended in the IMMPACT framework [120]. In this respect, proof of concept studies are an essential component that help clinical researchers to decide whether to proceed to more comprehensive and expensive phase [121]. There is no doubt that piloting a study is an essential element. For all the other methodological matters it is paramount the researchers pilot the study on themselves, including the stimulation pattern and procedure. Aiming for simple designs and short (as

Psychophysics and Electrophysiology of Pain

149

possible) procedures is general advice that applies even more soundly to pain studies, as they can be more exhausting and stressful for participants (and researchers) than many other observational studies. The rule of thumb here is that if the participants became exhausted or distressed by the experimental procedure, not only would the researchers perform poorly from an ethical point of view, but they would also risk jeopardizing the quality of the data and hence their interpretation. In this vein, long and complex experiments increase the likelihood of poor performance. Pain studies are, however, susceptible to complexity and long durations as compared to other research domains. Therefore, it is important to simplify the design and keep the research questions to a minimum within each study or experiment. With such complexity, procedural consistency across participants is vital. To achieve it, senior researchers can write down detailed protocols to follow during experimental sessions and to be handed out to junior colleagues (who may need even more guidance). Written protocols can then help ensure consistency across experimenters, and as a result, across participants as well. Technical considerations can significantly improve the study outcome. For example, if oscillatory EEG responses are of interest the researcher may opt for a contact-heat stimulator, instead of a water bath, as it will deliver short-lasting bursts of temperature increases and decreases. In contrast, the water bath technology would not allow fast changes in temperature and thus the generation of periodic patterns of stimulation. If instead the researcher is interested in a constant experience of pain versus no-pain, a tonic fixed temperature would perfectly suit. 4.3 Ethical Considerations

One of the most obvious critical elements of ethical procedures is informed consent. In fact, informed consent can only be achieved if the participant fully comprehends the procedure. Therefore, the information provided to the participant must be an accurate and fair representation of the procedures they will undergo. A vital point is how much information is essential for the participant to be considered as “informed”, and whether the information can affect the results of the study. This is of particular importance in pain research because the words and sentences used in the study description could build expectations in the participant that can then affect the results; research has highlighted that increased expectancy is correlated with an increase in pain ratings. An accurate description of the pain induction procedure is therefore imperative. It is also imperative to consider who the population is when writing this text—you would not use the same language for adults as you would for children, similarly you might want to carefully consider the choice of words when dealing with those who might have a different thinking style such as those with autism spectrum conditions. Understandably, using pain stimulators in any study requires careful planning and a thorough examination by your local ethical

150

Elia Valentini et al.

committee. While ethical considerations certainly limit the research that can be done using painful stimulation, the safety and wellbeing of all participants should always have the highest priority. Researchers should be fully aware of procedural safety measures for all tests being used or considered in their design. For example, the time taken for vasoconstriction to occur for the temperature being used in the water bath immersion. It is therefore important to consider the following recommendations when conducting an experiment with painful stimuli. First, the number of pain stimuli should be limited and within reason, and the use of painful stimuli in any study must be justified. Participants must give their full consent after being informed about the goals, procedures, and possible risks. Upon arrival at the lab, whenever possible, it is also recommended to let the participant experience an example of a painful stimulus that will be used during the study at first and then ask if they still agree to continue. Second, the devices that are used for presenting painful stimulation should be completely safe and should not cause any long-term damage or cause great distress to the participant. They should not be used in any other way than described in the user handbook and should undergo regular technical checks. In addition, all researchers who operate these devices should have received appropriate training prior to testing participants. Sterilization or cleaning of equipment and the experimental environment needs to be considered in the protocol design. Equipment should be cleaned in between one participant and another, hence an appropriate amount of time must be available between sessions to have the entire apparatus properly set-up. This is to ensure the maximal safety of the participant—particularly when there may be complex medical issues to consider, and/or complex psychiatric issues. While these general recommendations are applicable to all pain research, there are additional considerations when doing research with specific populations. In what follows, we will discuss points of attention for several specific populations. 4.4 Children and Individuals Who May Lack Capacity (Including Cognitive Impairment)

Including children or individuals who lack capacity in experimental pain research can be a daunting process and is preferably avoided. However, there is a range of circumstances in which children or those lacking capacity would be a potential sample population for pain research. In such instances, typical practice is to inform either the parents or the legal tutor of the person lacking capacity. However, as a good practice, there should be tailored information presented to the participant themselves, if the guardian of the participant concurs. When including these populations in pain research it is important to think about if the chosen test is the most valid way of gaining the data, and if it is the only way in which this data could be obtained. A test should be chosen that ensures tissue damage is of minimal risk, and that there is no chance

Psychophysics and Electrophysiology of Pain

151

of psychological impacts. Again, the balance of risk and benefit should be considered. Children and legal guardians should still feel in control in the same way that an adult participant would, so this control should be a careful consideration of the protocol. For example, reminding children or legal guardians more often that the procedure can stop at any time they wish. Asking children or legal guardians if they are happy to continue with the experiment more frequently. Breaks should also be considered more frequently for either population. Protecting children or those who lack capacity from harm may be impossible by the very nature of pain research. For this reason, it is also important to include in protocols, discussions around levels of distress or behavioral responses to novel environments or to novel stimuli with the legal guardian or parent, both prior and during the experiment. One remedy, to distress caused by novelty could be to show the pain equipment or a virtual tour of the laboratory, prior to the day of the experiment. It is also recommended that protocols incorporate a recovery period after data collection is complete. It can be a useful practice to have a playful activity at the start and end of the experiment for them to engage with or to discuss their favorite topic. This can help to assess carefully if they are prepared to participate in the study and can also help to mitigate the painful experience. 4.5 Recruitment and Screening

Recruitment and sample size considerations are particularly relevant to clinical trials. However, because of the replicability crisis observed in many research areas of life and social sciences, increased interest in this aspect of methodology has been accruing for crosssectional, observational, correlational, and experimental studies in the last years. Research has shown that recruitment methods may influence participant heterogeneity, particularly from a personality point of view, for both clinical and experimental studies [114, 115]. Therefore, the recruitment method should consider potential selection and self-selection issues and tackle them through diversification of sampling methods (e.g., online public adverts, through support groups, physical flyers, and posters, mailing lists, and local newspapers). Concerning screening, a different approach is required in studies involving healthy participants versus chronic pain patients or special populations. In addition, depending on the type of stimulation involved in the study, the reader might want to consider further detail in the level of screening. For experimental or correlational studies with healthy volunteers, there could be several approaches to recruitment and screening with different degrees of stringency. In most cases, the researcher may want to ask volunteers to at least report whether they have medical conditions (e.g., neurological, psychiatric, systemic) that may threaten their safe participation in the study for the individual, or a more specific series

152

Elia Valentini et al.

of questions about drug intake, medications and specific conditions (e.g., dermatological symptoms, sleep disturbances, migraine) that may bias the measurements. We recommend to consult the reference literature for standards, starting with [83]. When it comes to studies involving clinical patients, one of the most common issues is the researcher being able to rely only on screening questionnaires and self-report measures to diagnose and quantify any pain conditions and missing a clinical assessment of the patient. Having a collaboration with a clinician and/or a hospital can clearly solve both the recruitment bias and the screening deficit. Moreover, psychological comorbidities such as anxiety and depression are highly prevalent in these clinical patients [116, 117], and may contribute to maintaining the pain symptoms [118]. In any case, studies involving patients may benefit from very clear and detailed inclusion and exclusion criteria, as well as a detailed report of health status of the participants at enrolment (e.g., co-morbidities, by means of well-validated questionnaires such as the Hospital Anxiety and Depression Scale and Patient Health Questionnaire9). 4.6 Stimulus Material and Stimulation Methodology

We already discussed the type of stimulation one could potentially adopt within a study. Remember that we distinguished between stimulation that is phasic (short lasting, such as a pinprick) or tonic (sustained for a longer time, such as immersion in a cold or hot bath). You may want to keep in mind that some devices delivering phasic stimulation can also be adapted into a tonic pattern, but not the other way around. For example, while electrical stimulation is usually short-lasting, it can be transformed into a tonic stimulus by reducing the intensity and extending the presentation time (using either a continuous pulse or by creating a train of pulses). Moreover, depending on the research question, the onset and offset of the phasic (or tonic) stimulus may need to be timed very precisely. This is, for example, the case when measuring somatosensory event-related potentials elicited by a brief stimulus. To achieve this, preference is given to devices that can be controlled or triggered via programming software. Finally, depending on the research question, one might also consider the mobility of the stimulation devices and the body parts of interest. For example, the electrodes of an electrical stimulator can technically be placed anywhere on the body, from the fingers to the lower back, and this makes it possible to stimulate that body part during movement execution. The researcher may want to pay attention when using this type of stimulus that the electrodes are better not placed over muscles to avoid muscular contractions during stimulation. Water baths, by contrast, only allow stimulation of the limbs. In any case, two major physiological processes will very likely occur during pain provocation studies, and more generally in any study applying sensory stimulation, that is habituation and

Psychophysics and Electrophysiology of Pain

153

sensitization. They are both particularly relevant in pain studies both as an object of investigation and as a confound to control during the study. If control is required, it is worth considering minimization of stimulation intensities below tolerance level as that will prevent build-up of sensitization in the short term within the first part of a study session (hence not only for ethical reasons). Likewise, near-threshold and mild painful stimuli can lead to relatively rapid habituation of both a peripheral and central nature. It is worth noting that serious consideration must be given to repeated stimulation sessions over time as research has shown the two phenomena are coupled and have different temporal functions [119], as well as repeated stimulation to different body sites within the same session, because site non-specific sensitization to thermal pain has been reported [106]. 4.7 Data Analysis and Interpretation

Price and colleagues suggested VAS is superior to NRS for it reveals ratio scale properties [100]. However, it is worth noting that the numerical scale may be more quickly adopted and more efficient than the visual scale in the clinical setting [122]. Nonetheless, their statistical comparability seems to be well accepted [123]. We proposed SDT as a major approach to the study of pain psychophysics; however, SDT application to pain was not devoid of criticism. In fact, Rollman summarized the problem by emphasizing the reduced efficiency of the theory when moving from the context of sensory detection to the discrimination between intense but diverse stimuli [124]. As emphasized in Chapman’s rebuttal letter to Rollman [125], “It (SDT) must not be looked upon as a final solution to the problem of measuring pain [..]”. Indeed, notwithstanding the mathematical appeal of SDT measures, not all the experimental questions and settings allow for their application (e.g., SDT requires a larger number of trials than conventional threshold estimations). Concerning EEG recordings during nociceptive stimulation, it is worth noting that even when selective stimulation of nociceptors is achieved, most of the subsequent synchronized brain activity cannot be considered as a fingerprint of the experience of pain [8]. However, we still recommend to care about the timing of stimulation, especially if phasic stimuli are used, in order to obtain a good signal-to-noise ratio and reduce the amount of variability (already particularly large) associated with brain and cortical responses concomitant with the experience of pain.

154

Elia Valentini et al.

References 1. Hardy JD, Wolff HG, Goodell H (1940) Studies on pain. A new method for measuring pain threshold: observations on spatial summation of pain. J Clin Invest 19:649–657 2. Hillyard SA, Squires KC, Bauer JW, Lindsay PH (1971) Evoked potential correlates of auditory signal detection. Science 172:1357– 1360. https://doi.org/10.1126/science. 172.3990.1357 3. Clark WC (1969) Sensory-decision theory analysis of the placebo effect on the criterion for pain and thermal sensitivity. J Abnorm Psychol 74:363–371. https://doi.org/10. 1037/h0027509 4. Clark WC, Mehl L (1973) Signal detection theory procedures are not equivalent when thermal stimuli are judged. J Exp Psychol 97:148–153. https://doi.org/10.1037/ h0033866 5. Campbell JA, Lahuerta J (1983) Physical methods used in pain measurements: a review. J R Soc Med 76:409–414 6. Merskey H, Bogduk N (1994) Classification of chronic pain. In: Merskey H, Bogduk N (ed.) Part III: pain terms, a current list with definitions and notes on usage. IASP task force on taxonomy. IASP Press, Seattle, pp. 209–214 7. Mouraux A, Iannetti GD (2009) Nociceptive laser-evoked brain potentials do not reflect nociceptive-specific neural activity. J Neurophysiol 101:3258–3269. https://doi.org/ 10.1152/jn.91181.2008 8. Legrain V, Iannetti GD, Plaghki L, Mouraux A (2011) The pain matrix reloaded: a salience detection system for the body. Prog Neurobiol 93:111–124. https://doi.org/10.1016/ j.pneurobio.2010.10.005 9. Mouraux A, Iannetti GD (2018) The search for pain biomarkers in the human brain. Brain 141:3290–3307. https://doi.org/10.1093/ brain/awy281 10. Lee I-S, Necka EA, Atlas LY (2020) Distinguishing pain from nociception, salience, and arousal: how autonomic nervous system activity can improve neuroimaging tests of specificity. NeuroImage 204:116254. https://doi. org/10.1016/j.neuroimage.2019.116254 11. Weber Ii KA, Wager TD, Mackey S et al (2019) Evidence for decreased Neurologic Pain Signature activation following thoracic spinal manipulation in healthy volunteers and participants with neck pain. Neuroimage Clin 24:102042. https://doi.org/10.1016/ j.nicl.2019.102042

12. Hardy JD (1953) Thresholds of pain and reflex contraction as related to noxious stimulation. J Appl Physiol 5:725–739. https://doi. org/10.1152/jappl.1953.5.12.725 13. Graziano MSA, Cooke DF (2006) Parietofrontal interactions, personal space, and defensive behavior. Neuropsychologia 44: 845–859. https://doi.org/10.1016/j. neuropsychologia.2005.09.009 14. Wallwork SB, Grabherr L, O’Connell NE et al (2017) Defensive reflexes in people with pain - a biomarker of the need to protect? A metaanalytical systematic review. Rev Neurosci 28: 381–396. https://doi.org/10.1515/ revneuro-2016-0057 15. Legrain V, Mancini F, Sambo CF et al (2012) Cognitive aspects of nociception and pain: bridging neurophysiology with cognitive psychology. Neurophysiol Clin 42:325–336. https://doi.org/10.1016/j.neucli.2012. 06.003 16. Loeser JD, Treede R-D (2008) The Kyoto protocol of IASP Basic Pain Terminology. Pain 137:473–477. https://doi.org/10. 1016/j.pain.2008.04.025 17. Mu¨cke M, Cuhls H, Radbruch L et al (2016) Quantitative sensory testing (QST). English version. Schmerz. https://doi.org/10.1007/ s00482-015-0093-2 18. Eliav E, Gracely RH (2008) Chapter 3 - Measuring and assessing pain. In: Sharav Y, Benoliel R (eds) Orofacial pain and headache. Mosby, Edinburgh, pp 45–56 19. Treede RD, Lorenz J, Baumgartner U (2003) Clinical usefulness of laser-evoked potentials. Neurophysiol Clin 33:303–314 20. Cruccu G, Garcı´a-Larrea L (2004) Clinical utility of pain--laser evoked potentials. Suppl Clin Neurophysiol 57:101–110. https://doi. org/10.1016/s1567-424x(09)70348-4 21. Garcia-Larrea L (2012) Objective pain diagnostics: clinical neurophysiology. Neurophysiol Clin 42:187–197. https://doi.org/10. 1016/j.neucli.2012.03.001 22. Carmon A, Mor J, Goldberg J (1976) Evoked cerebral responses to noxious thermal stimuli in humans. Exp Brain Res 25:103–107 23. Colon E, Legrain V, Mouraux A (2012) Steady-state evoked potentials to study the processing of tactile and nociceptive somatosensory input in the human brain. Neurophysiol Clin. https://doi.org/10.1016/j.neucli. 2012.05.005 24. Schulz E, May ES, Postorino M et al (2015) Prefrontal gamma oscillations encode tonic

Psychophysics and Electrophysiology of Pain pain in humans. Cereb Cortex 25:4407– 4414. https://doi.org/10.1093/cercor/ bhv043 25. Furman AJ, Prokhorenko M, Keaser ML et al (2020) Sensorimotor peak alpha frequency is a reliable biomarker of prolonged pain sensitivity. Cereb Cortex. https://doi.org/10. 1093/cercor/bhaa124 26. Nickel MM, Ta Dinh S, May ES et al (2020) Neural oscillations and connectivity characterizing the state of tonic experimental pain in humans. Hum Brain Mapp 41:17–29. https://doi.org/10.1002/hbm.24784 27. Edens JL, Gil KM (1995) Experimental induction of pain: utility in the study of clinical pain. Behav Ther 26:197–216. https:// doi.org/10.1016/S0005-7894(05)80102-9 28. Staal JB, Klomp FP, Nijhuis-van der Sanden MWG (2019) Pain exposure physical therapy in complex regional pain syndrome: promising enough to warrant further investigation. Can J Anesth/J Can Anesth 66:115– 116. https://doi.org/10.1007/s12630-0181173-4 29. Sangesland A, Støren C, Vaegter HB (2017) Are preoperative experimental pain assessments correlated with clinical pain outcomes after surgery? A systematic review. Scand J Pain 15:44–52. https://doi.org/10.1016/j. sjpain.2016.12.002 30. Staahl C, Drewes AM (2004) Experimental human pain models: a review of standardised methods for preclinical testing of analgesics. Basic Clin Pharmacol Toxicol 95:97–111. https://doi.org/10.1111/j.1742-7843. 2004.950301.x 31. Linton SJ, Shaw WS (2011) Impact of psychological factors in the experience of pain. Phys Ther 91:700–711. https://doi.org/10. 2522/ptj.20100330 32. Arendt-Nielsen L, Andersen OK (2006) Electrophysiological assessment of pain. Suppl Clin Neurophysiol 59:241–249. https://doi.org/10.1016/s1567-424x(09) 70037-6 33. Lautenbacher S, Krieg JC (1994) Pain perception in psychiatric disorders: a review of the literature. J Psychiatr Res 28:109–122. https://doi.org/10.1016/0022-3956(94) 90023-X 34. Vaughan S, McGlone F, Poole H, Moore DJ (2019) A quantitative sensory testing approach to pain in autism spectrum disorders. J Autism Dev Disord 0:0. https://doi. org/10.1007/s10803-019-03918-0 35. de la Fuente-Sandoval C, Favila R, Go´mezMartı´n D et al (2012) Neural response to

155

experimental heat pain in stable patients with schizophrenia. J Psychiatr Res 46:128–134. https://doi.org/10.1016/j.jpsychires.2011. 09.008 36. Cascio C, McGlone F, Folger S et al (2008) Tactile perception in adults with autism: a multidimensional psychophysical study. J Autism Dev Disord 38:127–137. https:// doi.org/10.1007/s10803-007-0370-8 37. Treister R, Eisenberg E, Demeter N, Pud D (2015) Alterations in pain response are partially reversed by methylphenidate (Ritalin) in adults with attention deficit hyperactivity disorder (ADHD). Pain Pract 15:4–11. https:// doi.org/10.1111/papr.12129 38. Failla MD, Moana-Filho EJ, Essick GK et al (2018) Initially intact neural responses to pain in autism are diminished during sustained pain. Autism 22:669–683. https://doi.org/ 10.1177/1362361317696043 39. Stubbs B, Thompson T, Solmi M et al (2016) Is pain sensitivity altered in people with Alzheimer’s disease? A systematic review and meta-analysis of experimental pain research. Exp Gerontol 82:30–38 40. Lichtner V, Dowding D, Esterhuizen P et al (2014) Pain assessment for people with dementia: a systematic review of systematic reviews of pain assessment tools. BMC Geriatr 14:138. https://doi.org/10.1186/14712318-14-138 41. Birnie KA, Petter M, Boerner KE et al (2012) Contemporary use of the cold pressor task in pediatric pain research: a systematic review of methods. J Pain 13:817–826 42. Thompson T, Correll CU, Gallop K et al (2016) Is pain perception altered in people with depression? A systematic review and meta-analysis of experimental pain research. J Pain 17:1257–1272. https://doi.org/10. 1016/j.jpain.2016.08.007 43. Magerl W, Krumova EK, Baron R et al (2010) Reference data for quantitative sensory testing (QST): refined stratification for age and a novel method for statistical comparison of group data. Pain 151:598–605. https://doi. org/10.1016/j.pain.2010.07.026 44. Kim HJ, Yang GS, Greenspan JD et al (2017) Racial and ethnic differences in experimental pain sensitivity. Pain 158:194–211. https:// doi.org/10.1097/j.pain. 0000000000000731 45. Mody L, Miller DK, McGloin JM et al (2008) Recruitment and retention of older adults in aging research. J Am Geriatr Soc 56:2340– 2 3 4 8 . h t t p s : // d o i . o r g / 1 0 . 1 1 1 1 / j . 1532-5415.2008.02015.x

156

Elia Valentini et al.

46. Beach PA, Huck JT, Miranda MM, Bozoki AC (2015) Autonomic, behavioral, and subjective pain responses in Alzheimer’s disease. Pain Med 16:1930–1942. https://doi.org/ 10.1111/pme.12769 47. Williamson A, Hoggart B (2005) Pain: a review of three commonly used pain rating scales. J Clin Nurs 14:798–804 48. Karcioglu O, Topacoglu H, Dikme O, Dikme O (2018) A systematic review of the pain scales in adults: which to use? Am J Emerg Med 36:707–714 49. Younger J, McCue R, Mackey S (2009) Pain outcomes: a brief review of instruments and techniques. Curr Pain Headache Rep 13:39– 43 50. Glover M (1998) Outcome measures. Hand. Therapy 3:24–25. https://doi.org/10.1177/ 175899839800300407 51. Melzack R (1975) The McGill pain questionnaire: major properties and scoring methods. Pain 1:277–299. https://doi.org/10.1016/ 0304-3959(75)90044-5 52. Cleeland CS (1989) Measurement of pain by subjective report. Measurement of pain by subjective report, pp 391–403 53. Sullivan MJL, Bischop S, Pivik J (1995) The Pain Catastrophizing Scale: development and validation. Psychol Assess 7:524 54. McNeil DW, Rainwater AJ (1998) Development of the fear of pain questionnaire-III. J Behav Med 21:389–410. https://doi.org/ 10.1023/A:1018782831217 55. McCracken LM, Zayfert C, Gross RT (1992) The Pain Anxiety Symptoms Scale: development and validation of a scale to measure fear of pain. Pain 50:67–73. https://doi.org/10. 1016/0304-3959(92)90113-p 56. Bennett M (2001) The LANSS Pain Scale: the Leeds assessment of neuropathic symptoms and signs. Pain 92:147–157. https://doi. org/10.1016/S0304-3959(00)00482-6 57. Wolfe F (2003) Pain extent and diagnosis: development and validation of the regional pain scale in 12,799 patients with rheumatic disease. J Rheumatol 30:369–378 58. Galer BS, Jensen MP (1997) Development and preliminary validation of a pain measure specific to neuropathic pain: the Neuropathic Pain Scale. Neurology 48:332–338. https:// doi.org/10.1212/WNL.48.2.332 59. von Baeyer CL, Spagrud LJ (2007) Systematic review of observational (behavioral) measures of pain for children and adolescents aged 3 to 18 years. Pain 127:140–150. https://doi. org/10.1016/j.pain.2006.08.014

60. Breau LM, McGrath PJ, Camfield CS, Finley AG (2002) Psychometric properties of the non-communicating children’s pain checklist-revised. Pain 99:349 61. Kunz M, Meixner D, Lautenbacher S (2019) Facial muscle movements encoding pain - a systematic review. Pain 160:535–549 62. Anciano D (1986) The pain behaviour checklist: factor analysis and validation. Br J Clin Psychol 25:301–302. https://doi.org/10. 1111/j.2044-8260.1986.tb00709.x 63. McGrath PJ, Rosmus C, Canfield C et al (1998) Behaviours caregivers use to determine pain in non-verbal, cognitively impaired individuals. Dev Med Child Neurol 40:340– 3 4 3 . h t t p s : // d o i . o r g / 1 0 . 1 1 1 1 / j . 1469-8749.1998.tb15386.x 64. Elliott CH, Jay SM, Woody P (1987) An observation scale for measuring children’s distress during medical procedures. J Pediatr Psychol 12:543–551. https://doi.org/10. 1093/jpepsy/12.4.543 65. Ekman P, Friesen WV (1978) Facial action coding system: a technique for the measurement of facial movement. California, CA: Consulting Psychologists Press, Palo Alto 66. Chambers CT, Cassidy KL, McGrath PJ (1996) Child facial coding system: a manual. Dalhousie University and University of British Columbia, Nova Scotia and Vancouver 67. Fearon I, McGrath PJ, Achat H (1996) “Booboos”: the study of everyday pain among young children. Pain 68:55–62. https://doi. org/10.1016/S0304-3959(96)03200-9 68. Andersen OK (2006) Pain in humans, electrical stimulation (skin, muscle and viscera). In: Encyclopedia of pain. Springer, Berlin Heidelberg, pp 1679–1681 69. Sunil Kumar Reddy K, Naidu MUR, Usha Rani P, Ramesh Kumar Rao T (2012) Human experimental pain models: a review of standardized methods in drug development. J Res Med Sci 17:587–595 70. Laursen RJ, Graven-Nielsen T, Jenser TS, Arendt-Nielsen L (1997) Quantification of local and referred pain in humans induced by intramuscular electrical stimulation. Eur J Pain 1:105–113. https://doi.org/10.1016/ S1090-3801(97)90068-9 71. Keller T, Kuhn A (2008) Electrodes for transcutaneous (surface) electrical stimulation. J Automat Contr 18:35–45. https://doi.org/ 10.2298/jac0802035k 72. Accornero N, Bini G, Lenzi GL, Manfredi M (1977) Selective activation of peripheral nerve fibre groups of different diameter by triangular shaped stimulus pulses. J Physiol 273:539–

Psychophysics and Electrophysiology of Pain 560. https://doi.org/10.1113/jphysiol. 1977.sp012109 73. Inui K, Kakigi R (2012) Pain perception in humans: use of intraepidermal electrical stimulation. J Neurol Neurosurg Psychiatry 83: 551–556 74. O’Neill J, Brock C, Olesen AE et al (2012) Unravelling the mystery of capsaicin: a tool to understand and treat pain. Pharmacol Rev 64: 939–971 75. Scanlon GC, Wallace MS, Ispirescu JS, Schulteis G (2006) Intradermal capsaicin causes dose-dependent pain, allodynia, and hyperalgesia in humans. J Investig Med 54:238–244. https://doi.org/10.2310/6650.2006. 05046 76. Simone DA, Baumann TK, LaMotte RH (1989) Dose-dependent pain and mechanical hyperalgesia in humans after intradermal injection of capsaicin. Pain 38:99–107. https://doi.org/10.1016/0304-3959(89) 90079-1 77. Gustafsson H, Åkesson J, Lau CL et al (2009) A comparison of two formulations of intradermal capsaicin as models of neuropathic pain in healthy volunteers. Br J Clin Pharmacol 68: 511–517. https://doi.org/10.1111/j. 1365-2125.2009.03489.x 78. Keil A, Debener S, Gratton G et al (2014) Committee report: publication guidelines and recommendations for studies using electroencephalography and magnetoencephalography. Psychophysiology 51:1–21. https://doi.org/10.1111/psyp.12147 79. Blumenthal TD, Cuthbert BN, Filion DL et al (2005) Committee report: guidelines for human startle eyeblink electromyographic studies. Psychophysiology 42:1–15. https:// doi.org/10.1111/j.1469-8986.2005. 00271.x 80. Rhudy JL, France CR (2007) Defining the nociceptive flexion reflex (NFR) threshold in human participants: a comparison of different scoring criteria. Pain 128:244–253. https:// doi.org/10.1016/j.pain.2006.09.024 81. Torta DM, Jure FA, Andersen OK, Biurrun Manresa JAB (2019) Intense and sustained pain reduces cortical responses to auditory stimuli: implications for the interpretation of the effects of heterotopic noxious conditioning stimulation in humans. Eur J Neurosci 50: 3934–3943. https://doi.org/10.1111/ejn. 14546 82. Fillingim RB (2017) Individual differences in pain: understanding the mosaic that makes pain personal. Pain 158(Suppl 1):S11–S18.

157

h t t p s : // d o i . o r g / 1 0 . 1 0 9 7 / j . p a i n . 0000000000000775 83. Price DD, Milling LS, Kirsch I et al (1999) An analysis of factors that contribute to the magnitude of placebo analgesia in an experimental paradigm. Pain 83:147–156 84. Etherton J, Lawson M, Graham R (2014) Individual and gender differences in subjective and objective indices of pain: gender, fear of pain, pain catastrophizing and cardiovascular reactivity. Appl Psychophysiol Biofeedback 39:89–97. https://doi.org/10.1007/ s10484-014-9245-x 85. Ruau D, Liu LY, Clark JD et al (2012) Sex differences in reported pain across 11,000 patients captured in electronic medical records. J Pain 13:228–234. https://doi. org/10.1016/j.jpain.2011.11.002 86. Pieretti S, Di Giannuario A, Di Giovannandrea R et al (2016) Gender differences in pain and its relief. Ann Ist Super Sanita 52:184– 189. https://doi.org/10.4415/ANN_16_ 02_09 87. Kim HJ, Yang GS, Greenspan JD et al (2017) Racial and ethnic differences in experimental pain sensitivity: systematic review and metaanalysis. Pain 158:194–211. https://doi.org/ 10.1097/j.pain.0000000000000731 88. Bell BA, Ruscheweyh R, Kelley BJ et al (2018) Ethnic differences identified by pain sensitivity questionnaire correlate with clinical pain responses. Reg Anesth Pain Med 43:200– 204. https://doi.org/10.1097/AAP. 0000000000000689 89. El Tumi H, Johnson MI, Dantas PBF et al (2017) Age-related changes in pain sensitivity in healthy humans: a systematic review with meta-analysis. Eur J Pain 21:955–964. https://doi.org/10.1002/ejp.1011 90. Hoffman KM, Trawalter S, Axt JR, Oliver MN (2016) Racial bias in pain assessment and treatment recommendations, and false beliefs about biological differences between blacks and whites. Proc Natl Acad Sci U S A 113:4296–4301. https://doi.org/10.1073/ pnas.1516047113 91. Hampton SB, Cavalier J, Langford R (2015) The influence of race and gender on pain management: a systematic literature review. Pain Manag Nurs 16:968–977. https://doi. org/10.1016/j.pmn.2015.06.009 92. Wandner LD, Scipio CD, Hirsh AT et al (2012) The perception of pain in others: how gender, race, and age influence pain expectations. J Pain 13:220–227. https:// doi.org/10.1016/j.jpain.2011.10.014

158

Elia Valentini et al.

93. Stanton TR, Latimer J, Maher CG, Hancock MJ (2010) How do we define the condition ‘recurrent low back pain’? A systematic review. Eur Spine J 19:533–539. https://doi.org/10. 1007/s00586-009-1214-3 94. Coghill RC, McHaffie JG, Yen Y-F (2003) Neural correlates of interindividual differences in the subjective experience of pain. Proc Natl Acad Sci U S A 100:8538–8542. h t t p s : // d o i . o r g / 1 0 . 1 0 7 3 / p n a s . 1430684100 95. Vanden Bulcke C, Crombez G, Durnez W, Van Damme S (2015) Is attentional prioritization on a location where pain is expected modality-specific or multisensory? Conscious Cogn 36:246–255. https://doi.org/10. 1016/j.concog.2015.07.003 96. Doll RJ, Buitenweg JR, Meijer HGE, Veltink PH (2014) Tracking of nociceptive thresholds using adaptive psychophysical methods. Behav Res 46:55–66. https://doi.org/10. 3758/s13428-013-0368-4 97. Clauwaert A, Torta DM, Danneels L, Van Damme S (2018) Attentional modulation of somatosensory processing during the anticipation of movements accompanying pain: an event-related potential study. J Pain 19:219– 227. https://doi.org/10.1016/j.jpain.2017. 10.008 98. Notebaert L, Crombez G, Van Damme S et al (2010) Looking out for danger: an attentional bias towards spatially predictable threatening stimuli. Behav Res Ther 48:1150–1154. https://doi.org/10.1016/j.brat.2010. 07.013 99. Misha BM, Attal N, Baron R et al (2013) Value of quantitative sensory testing in neurological and pain disorders: NeuPSIG consensus. Pain 154:1807–1819. https://doi.org/ 10.1016/j.pain.2013.05.047 100. Price DD, Bush FM, Long S, Harkins SW (1994) A comparison of pain measurement characteristics of mechanical visual analogue and simple numerical rating scales. Pain 56: 217–226. https://doi.org/10.1016/03043959(94)90097-3 101. Jones B (1979) Signal detection theory and pain research. Pain 7:305–312. https://doi. org/10.1016/0304-3959(79)90086-1 102. Beck B, Cook S, Iannetti G, Haggard P (2019) Neural markers of nociceptive input and pain intensity coding: a signal detection approach. https://doi.org/10.31234/osf. io/dxeyn 103. Kenshalo DR, Anton F, Dubner R (1989) The detection and perceived intensity of noxious thermal stimuli in monkey and in human.

J Neurophysiol 62:429–436. https://doi. org/10.1152/jn.1989.62.2.429 104. Vlaeyen JWS, Crombez G, Linton SJ (2016) The fear-avoidance model of pain. Pain 157: 1588–1589. https://doi.org/10.1097/j. pain.0000000000000574 105. Mouraux A, Plaghki L (2007) Are laserevoked brain potentials modulated by attending to first or second pain? Pain 129:321– 331. https://doi.org/10.1016/j.pain.2006. 10.018 106. Willer JC (1977) Comparative study of perceived pain and nociceptive flexion reflex in man. Pain 3:69–80. https://doi.org/10. 1016/0304-3959(77)90036-7 107. Jensen MB, Biurrun Manresa J, Andersen OK (2015) Reliable estimation of nociceptive withdrawal reflex thresholds. J Neurosci Methods 253:110–115. https://doi.org/10. 1016/j.jneumeth.2015.06.014 108. Hartley C, Goksan S, Poorun R et al (2015) The relationship between nociceptive brain activity, spinal reflex withdrawal and behaviour in newborn infants. Sci Rep 5:12519. https://doi.org/10.1038/srep12519 109. Bagiella E, Sloan RP, Heitjan DF (2000) Mixed-effects models in psychophysiology. Psychophysiology 37:13–20 110. Bendayan R, Ramı´rez-Maestre C, Ferrer E et al (2017) From acute to chronic back pain: using linear mixed models to explore changes in pain intensity, disability, and depression. Scand J Pain 16:45–51. https://doi.org/10.1016/j.sjpain.2017. 02.009 111. Tabor A, Burr C (2019) Bayesian learning models of pain: a call to action. Curr Opin Behav Sci 26:54–61. https://doi.org/10. 1016/j.cobeha.2018.10.006 112. Hoskin R, Berzuini C, Acosta-Kane D et al (2019) Sensitivity to pain expectations: a Bayesian model of individual differences. Cognition 182:127–139. https://doi.org/ 10.1016/j.cognition.2018.08.022 113. Franz M, Ritter A, Puta C et al (2014) Laser heat hyperalgesia is not a feature of non-specific chronic low back pain. Eur J Pain 18:1501–1508. https://doi.org/10. 1002/ejp.535 114. Nijs J, Inghelbrecht E, Daenen L et al (2011) Recruitment bias in chronic pain research: whiplash as a model. Clin Rheumatol 30: 1481. https://doi.org/10.1007/s10067011-1829-8 115. Karos K, Alleva JM, Peters ML (2018) Pain, please: an investigation of sampling bias in

Psychophysics and Electrophysiology of Pain pain research. J Pain 19:787–796. https:// doi.org/10.1016/j.jpain.2018.02.011 116. Bener A, Verjee M, Dafeeah EE et al (2013) Psychological factors: anxiety, depression, and somatization symptoms in low back pain patients. J Pain Res 6:95–101. https://doi. org/10.2147/JPR.S40740 117. McWilliams LA, Goodwin RD, Cox BJ (2004) Depression and anxiety associated with three pain conditions: results from a nationally representative sample. Pain 111: 77–83. https://doi.org/10.1016/j.pain. 2004.06.002 118. Lerman SF, Rudich Z, Brill S et al (2015) Longitudinal associations between depression, anxiety, pain, and pain-related disability in chronic pain patients. Psychosom Med 77: 333–341. https://doi.org/10.1097/PSY. 0000000000000158 119. May A, Rodriguez‐Raecke R, Schulte A et al (2012) Within-session sensitization and between-session habituation: a robust physiological response to repetitive painful heat stimulation. Eur J Pain 16:401–409. https://doi.org/10.1002/j.1532-2149. 2011.00023.x 120. Dworkin RH, Turk DC, Peirce-Sandner S et al (2010) Research design considerations for confirmatory chronic pain clinical trials: IMMPACT recommendations. Pain 149: 177–193. https://doi.org/10.1016/j.pain. 2010.02.018 121. Campbell CM, Gilron I, Doshi T, Raja S (2019) Designing and conducting proof-ofconcept chronic pain analgesic clinical trials. PAIN Rep 4:e697. https://doi.org/10. 1097/PR9.0000000000000697 122. Brunelli C, Zecca E, Martini C et al (2010) Comparison of numerical and verbal rating scales to measure pain exacerbations in patients with chronic cancer pain. Health Qual Life Outcomes 8:42. https://doi.org/ 10.1186/1477-7525-8-42 123. Hjermstad MJ, Fayers PM, Haugen DF et al (2011) Studies comparing numerical rating scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. J Pain Symptom Manag 41:1073–1093. https:// doi.org/10.1016/j.jpainsymman.2010. 08.016 124. Rollman GB (1977) Signal detection theory measurement of pain: a review and critique. Pain 3:187–211. https://doi.org/10.1016/ 0304-3959(77)90002-1 125. Chapman CR (1977) Sensory decision theory methods in pain research: a reply to Rollman.

159

Pain 3:295–305. https://doi.org/10.1016/ 0304-3959(77)90062-8 126. Van Damme S, Legrain V, Vogt J, Crombez G (2010) Keeping pain in mind: a motivational account of attention to pain. Neurosci Biobehav Rev 34:204–213. https://doi.org/10. 1016/j.neubiorev.2009.01.005 127. Villemure C, Bushnell MC (2002) Cognitive modulation of pain: how do attention and emotion influence pain processing? Pain 95: 195–199. https://doi.org/10.1016/s03043959(02)00007-6 128. Valeriani M, Restuccia D, Le Pera D et al (2002) Attention-related modifications of ultra-late CO(2) laser evoked potentials to human trigeminal nerve stimulation. Neurosci Lett 329:329–333. https://doi.org/ 10.1016/s0304-3940(02)00671-7 129. Torta DMM, Legrain V, Mouraux A, Valentini E (2017) Attention to pain! A neurocognitive perspective on attentional modulation of pain in neuroimaging studies. Cortex 89: 120 130. Crombez G, Baeyens F, Eelen P (1994) Sensory and temporal information about impending pain: the influence of predictability on pain. Behav Res Ther 32:611–622. https://doi.org/10.1016/0005-7967(94) 90015-9 131. Crombez G, Van Damme S, Eccleston C (2005) Hypervigilance to pain: an experimental and clinical analysis. Pain 116:4–7. https://doi.org/10.1016/j.pain.2005. 03.035 132. Seminowicz DA, Mikulis DJ, Davis KD (2004) Cognitive modulation of pain-related brain responses depends on behavioral strategy. Pain 112:48–58. https://doi.org/10. 1016/j.pain.2004.07.027 133. Veldhuijzen DS, Kenemans JL, de Bruin CM et al (2006) Pain and attention: attentional disruption or distraction? J Pain 7:11–20. https://doi.org/10.1016/j.jpain.2005. 06.003 134. Moseley GL, Zalucki NM, Wiech K (2008) Tactile discrimination, but not tactile stimulation alone, reduces chronic limb pain. Pain 137:600–608. https://doi.org/10.1016/j. pain.2007.10.021 135. Roelofs J, Peters ML, van der Zijden M, Vlaeyen JWS (2004) Does fear of pain moderate the effects of sensory focusing and distraction on cold pressor pain in pain-free individuals? J Pain 5:250–256. https://doi. org/10.1016/j.jpain.2004.04.001

Chapter 8 The Many Challenges of Human Experimental Itch Research Henning Holle and Donna M. Lloyd Abstract Itch has long been a neglected sense within somatosensory research, and with good reason: acute itch, although relatively easy to trigger, is notoriously difficult to control experimentally. Its time course and behavior cannot easily be predicted and participants find it difficult to quantify (and indeed qualify) the sensation. Even scratch behavior in response to itch is weakly correlated to the amount of itch someone is experiencing. This chapter will focus on the three main methods of acute, experimental itch induction: chemically evoked itch (through histamine and cowhage), mechanically or electrically evoked itch, and psychologically evoked itch (through visual and auditory means). The basic materials and experimental designs will be described along with our personal experiences of trying to study itch using these methods. Itch research is not for the faint-hearted; there are more failures along the road than successes. We do it because itch remains one of the most elusive and fascinating areas of somatosensory research. It can give excruciating pain or intense pleasure with just a single scratch and unlocking its mysteries will help the countless thousands who experience debilitating pruritic skin conditions such as eczema, atopic dermatitis, and psoriasis. Key words Acute itch, Auditory evoked itch (AEI), Chemically evoked itch, Cowhage, Electrical/ mechanically evoked itch, Experimental induction, Histamine, Pruritus, Psychologically evoked itch, Visually evoked itch (VEI)

1

Introduction Itch is one of our most basic bodily sensations and serves a vital protective function. Even those who lose their sense of touch through large fiber neuropathy still have the ability to feel itch [1]. The first published studies on human experimental itch research date back to the beginning of the twentieth century [2]. Although relatively easy to trigger (using itching powder— mucuna pruriens—or mechanical stimulation with wool), little was known about the underlying mechanisms at the time. Historically, itch was conceived as a form of pain (see also Chapter 7, this volume). As early as 1922, von Frey wrote a short paper devoted to the “problem of pruritus” [3]. In it, he postulated that itch and

161

162

Henning Holle and Donna M. Lloyd

pain resulted from the same stimulus and were served by the same nerve network. The weaker stimulus led to itch while the stronger stimulus led to pain, and this could be manipulated in a linear way such that increasing the pressure on the stimulus—a plant bristle— led to increasing intensity of sensation up to prickling and even burning. The average delay period between application of the stimulus and itching to occur was 10 s, and this delay was necessary for the release of chemicals to trigger the sensation. Scratching or rubbing relieved itch by diluting or removing the stimulating substance released in the skin. While von Frey’s original intensity hypothesis has since been replaced by more complex interactive accounts [4], his succinct summary not only provides a description of itch that we would recognize today, but he clearly identifies one of the main challenges of human experimental itch research: namely, the slow time course of the itch sensation to develop from delivery of the stimulus. In this chapter, we will focus on the many challenges associated with experimentally inducing itch and the variety of different methods for eliciting itch. We focus on those we have had some firsthand experience of, including chemically evoked itch (using histamine and cowhage), mechanically evoked itch (using electrical stimuli), and psychologically evoked itch (using visual and/or auditory stimuli). We will discuss some of the best ways to assess itch and the key methodological challenges of studying acute itch.

2

Materials

2.1 Chemically Evoked Itch

Researchers wishing to study acute itch in humans can choose between a variety of chemical itch induction methods, each with their own set of advantages and disadvantages (Table 1). A wellestablished substance in this context is histamine, often employed in the form of 1% histamine dihydrochloride in aqueous solution. Since histamine cannot cross the intact skin barrier, several methods have been developed to deliver histamine to its site of action at the junction of epidermis and dermis, where the terminals of itchrelated C-fibers are located [7, 8]. One such method is the histamine prick test, which is well established both as a research tool and as a control stimulus in routine allergy diagnosis. In a histamine prick test, a drop of histamine solution is placed on the target site and the skin is then pricked through the drop using the tip of a sterile lancet. With this method, tiny amounts (1–2 μl maximum; [9]) of the histamine solution are delivered into the upper layers of the skin. Due to the construction of the lancet (a 1 mm tip followed by a broad shoulder), penetration depth is very limited. After a latency of about 35 s, an itch sensation like a mosquito-bite starts to develop, peaking at around 120 s after the onset of the skin prick [10] followed by a

Itch

163

Table 1 Overview of materials required for induction and assessment of chemically induced itch and their respective advantages and disadvantages Method

Required materials

Advantages (+) and disadvantages (-)

Histamine prick Histamine vials (concentrations of 1%, test 0.1%, and 0.01%) Control vials (saline solution) Skin prick lancets (all available from Allergy Therapeutics, Worthing, UK)

+ simple, does not require any special equipment + histamine solutions can be bought ready-made + allows assessment of skin response - limited options for manipulating dose intensity

Histamine Histamine gel (e.g., 1% histamine iontophoresis dissolved in 2.5% methylcellulose), needs to be created on site either by a lab or a pharmacist Iontophoresis device Delivery and dispersive electrodes (both device and electrodes are available, for example from Perimed Instruments, J€arf€alla, Sweden)

+ more degrees of freedom (dose, infusion time, current) for manipulating dose intensity than prick test method + allows assessment of skin response - requires special equipment - solutions need to be prepared on site

Cowhage (rubbing method)

Cowhage spicules (available upon request from Dr Ethan Lerner, Massachusetts General Hospital, or from Zandu Pharmaceuticals, Mumbai, India) Scotch tape to demarcate target area Cotton cloth (to remove stray spicules and reduce skin irritation)

+ simple method to stimulate non-histaminergic pathway + no special equipment required - limited control over dose intensity - cowhage-induced itch does not produce measurable skin reactions or changes in blood perfusion

Cowhage (spicule insertion method)

Cowhage spicules (see above) Cotton tab (as applicator) Glue (to attach spicules to applicator)

+ allows more control over dose intensity than rubbing method + no special equipment required - cowhage-induced itch does not produce measurable skin reactions or changes in blood perfusion

Assessment of skin reaction (wheal & flare)

Translucent paper and pen [5] Or alternatively, use kajal eyeliner to draw outline directly on skin, follow by digital photograph with ruler added for scale, see [6]

Assessment of blood perfusion

Laser Doppler Flowmetry (available, for example from Moor Instruments)

slow decay (see Note 4.1). Due to the vasodilatory effect of histamine, the area surrounding the skin prick is raised (wheal), and is encircled by an area of reddened skin (flare). Itch, wheal, and flare tend to completely subside 30–60 min after the histamine prick.

164

Henning Holle and Donna M. Lloyd

The intensity of the itch sensation varies as a function of the histamine dose up to a limit [7] (see Note 4.2). Another method is application of histamine by iontophoresis using an electric current. This typically involves creating a gel by dissolving histamine dihydrochloride in methylcellulose [5, 11]. The gel is then placed into a delivery electrode (the anode) and a larger reference cathode is fixed nearby. Subsequently, a small electrical current is applied which causes the positively charged histamine ions to be driven into the skin by repulsion of the charges from the anode. As with the histamine prick test, itch intensity varies in a dose-dependent manner during iontophoresis by manipulating either the strength of the histamine solution or aspects of the electrical stimulation, i.e., current and length of stimulation [12]. Finally, there are some slightly more invasive methods to ensure transepidermal delivery of histamine for an induction of itch. Van de Sand and colleagues [13] aimed for an intense, long-lasting itch stimulus for their research. To achieve this, they slightly abraded the target skin site and then covered these pre-treated sites with a histamine gel. In the past (e.g., [14]), researchers have also injected histamine. However, this approach has been discontinued, partly because injection of histamine tends to elicit a mixture of pain and itch, rather than a pure itch sensation [7]. In addition to the histamine itch pathway, there is also a separate histamine-independent itch pathway that has been increasingly better understood during the past decade [15]. This alternative pathway is, for example, stimulated by the tropical plant cowhage, in particular when the tiny hair-like spicules covering the seed pods of the plant become lodged in the skin. Mucunain, the itchinducing agent of the cowhage spicules, binds to proteinaseactivated receptors 2/4 (PAR 2/4) in the epidermis [16] and induces itch that differs from histamine in terms of its local skin reaction (little or no flare for cowhage; [6, 17]), the type of nociceptive C-fibers involved (cowhage: mechanosensitive, histamine: mechanoinsensitive; [18, 19]), and the quality of the itch response (cowhage: more pricking, stinging and burning; [6, 20, 21]). Some authors have argued that PAR2/PAR4-mediated itch may be a more appropriate model for understanding pathological itch than models based on histamine, since antihistamines have limited effectiveness in treating chronic itch [22]. Since mucunain is not yet widely available as a synthesized compound, itch researchers have so far relied on manual insertion of the cowhage spicules to activate the PAR2/PAR4 pathway (see Note 4.3). In one technique, a small number of individual spicules are glued to an applicator (e.g., a small cotton stick) so that the spicules protrude from it perpendicularly [20, 21]. The spicules are then inserted at a 30° angle into the skin so that approximately 0.2 mm of the tip of the spicule enters the skin. Other research

Itch

165

groups place a small number of spicules on the skin and then rub for a period of 30–45 s. This results in some of the spicules becoming lodged in the skin [23, 22]. Both approaches yield an itch sensation peaking around 1–2 min after the beginning of the stimulation, followed by a slow decay. 2.2 Electrical/ Mechanically Evoked Itch

Electrical or mechanical stimulation has been shown to produce controlled levels of itching in early studies [8, 24, 25]. Constant monophasic pulsations of direct positive current at 50 Hz and 10 ms on/off cycles were passed through non-invasive electrodes placed on the skin and a linear response of itch intensity (as measured by mean response times to detect itching at differing levels of stimulation) was found [24]. However, the reproducibility and intensity of the itch sensation was not high. More recent studies have used highly innovative methods for delivering electrical or mechanical stimulation to produce a more definitive and reproducible itch sensation (see below).

2.2.1 How to Create an Electrically Evoked Itch Stimulus

Transcutaneous electrical stimulation is typically delivered through a stainless-steel wire attached to the skin controlled by a programmable stimulator such as Digitimer (see also Chapter 7, this volume). Saline-soaked gauze pads act as the reference electrode (anode). Constant current stimulation of different durations and frequencies is then applied. For example, [26] used durations ranging between 0.08–8 ms and frequencies between 2–200 Hz with stimulation applied to the left wrist. We have also tried transcutaneous electrical nerve stimulation (TENS) to deliver an itch-inducing stimulus (see also Chapter 20, this volume). TENS is a method of pain relief involving the use of a mild electric current (frequency 50–100 Hz, pulse width 50–200 μs). It is delivered through a battery-operated device that has leads connected to electrode pads through which current is transmitted directly to the skin layers activating Aβ fibers. When the machine is switched on, small electrical impulses are delivered to the affected area of the body, which is often felt as a tingling sensation. It was our hope that we could modify this tingling sensation by increasing the electric current to elicit the sensation of itch in a controlled way; however, this was not the case, and we were unable to produce a reliable itch sensation in our participants who mainly reported just tingling sensations (although previous studies have produced a reliable itch sensation with increasing current, e.g., [25]).

2.2.2 How to Create a Mechanically Evoked Itch Stimulus

A gentler stimulus is used to create the sensation of itch using mechanical rather than electrical stimulation. For example, in a study by Fukuoka et al. [27] a probe for mechanical stimulation was created. One end of the 30 cm probe had an electrically controlled piezoelectric actuator that horizontally vibrated in the

166

Henning Holle and Donna M. Lloyd

range of 0° to 0.2° at a frequency of 1 to 50 Hz. This led to the horizontal vibration of a stainless-steel wire loop with amplitude 0 to 1 mm at the other end of the probe. The wire loop could touch and vibrate only vellus hairs (i.e., the barely noticeable hair that develops on most of a person’s body during childhood) and not the skin surface. The frequency and amplitude were fixed at 8 Hz and 1 mm for most testing sessions. 2.3 Psychologically Evoked Itch

In addition to itches that arise due to mechanical or chemical stimulation of the skin, it is possible for a non-tactile stimulus to induce or increase itch sensations through psychological suggestion, that is, with no physical stimulation of the skin. These effects can be created or manipulated in an experimental setting using visual or auditory stimuli (see Note 4.4). For example, people typically feel itch sensations which they want to scratch, when presented with videos, static pictures or sounds of itch-related stimuli like insects, or when seeing other people scratching (see also Chapter 6, this volume). In contrast to chemically or mechanically evoked itch, this type of itch-inducing stimulation is non-invasive and no skin manipulation is needed. This makes it the ideal stimulus for investigating experimentally induced itch in both skin healthy controls and those with a pruritic skin condition where stimulation of the skin might worsen their condition.

2.3.1 How to Create an Itch-Inducing Stimulus Set for Visually Evoked Itch (VEI)

We have typically used static images to trigger visually evoked itch in the laboratory. An itch-evoking event can take many forms and some events may act as a more potent trigger for some people than others. It therefore helps to first generate a reasonably large set of well-matched itch and non-itch images and categorize the content of the images as much as possible. Itch content was first considered in [28], which used three image categories: insects (context), insects touching the skin (skin contact), and people scratching the skin (skin response). Skin response was subsequently divided into two separate categories to differentiate between the action of scratching and the result of skin irritation [29], resulting in four image categories: (i) Skin contact: images of itch-related vs. non–itch-related objects in contact with the skin (e.g., insects crawling on the hand vs. marbles touching the skin). (ii) Skin response: images of human responses to itch (i.e., scratching) or non–itch-related touching of the skin (i.e., washing the hands). (iii) Context: images where itch or non-itch stimuli were seen in the environment but not on the body (e.g., ants crawling on the ground or butterflies flying). (iv) Skin condition: images of hives or freckles.

Itch

167

As well as itch and non-itch, our images were also balanced by body part [30]. We have previously used arms and hands, legs and feet, head and neck, and torso [29, 31]. Their inclusion allows for the systematic investigation of body location differences in VEI. Next, it’s important to match the itch and non-itch images as much as possible on attributes such as appearance and composition. For example, brightness and color saturation should be as similar as possible so that one image is not simply more “attention grabbing” than another. Similarly, the position of the models, the angle of view, and the exact body locations should be as closely matched as possible, and the contents cover approximately the same proportion of both image areas (see Note 4.5). As a final step, it can be useful to carry out a test survey to ensure itch and non-itch images are matched in every aspect except how itchy they make people feel [31]. Previously we had simply asked people to rate how stimulating the image appeared to them, using a seven-point scale illustrated with Self-Assessment Manikin pictograms [32]. This is a non-verbal pictorial assessment technique that directly measures the pleasure, arousal, and dominance associated with a person’s affective reaction to a wide variety of stimuli, and this assists with participants’ interpretation of stimulation. 2.3.2 How to Create an Itch-Inducing Stimulus Set for Auditory Evoked Itch (AEI)

Contagious itch has also been investigated in the auditory modality by [33] who found that scratching sounds, particularly at higher frequencies, led to increased itching in both psoriatic and skin healthy participants compared to rubbing sounds. A particular challenge in auditory evoked itch is to obtain good quality scratching recordings. Due to the very limited loudness of a scratching sound, recordings often have a poor signal-to-noise ratio. We recommend using professional recording equipment to minimize the influence of background noise. Similar to the considerations already made for VEI, the auditory experimental items and control items should only differ on the variable of interest (e.g., scratch vs. rubbing as a control sound), but be identical in all other respects. Particularly helpful tools in this respect are Audacity (www.audacityteam.org) for editing and Praat (http://www.fon.hum.uva.nl/praat/) for advanced processing. Both packages are available free of charge. For example, Praat allows matching stimuli for overall loudness, or decomposing sounds into frequency bands [33]. When presenting scratch sounds to the participants, care must be taken to maintain the natural faint loudness of scratching, as too high a volume can lead to a hyperrealistic impression of scratching, reducing the overall validity of the experiment.

2.4 How Can We Assess Itch?

The assessment of itch can take place on at least four different levels: (1) via self-reports, (2) through the observation of behavior, (3) via physiological correlates in the peripheral nervous system, and (4) through measuring neural correlates in the central nervous system.

168

Henning Holle and Donna M. Lloyd

2.4.1 Self-Report Measures

Self-report measures have a long-standing tradition in human itch research. For quantitative itch measurements (e.g., of acute itch intensity), there are a variety of available rating scales including classic visual analog scales, verbal rating scales, and hybrid forms (i.e., analog scales with additional verbal anchor points, [12, 21], see Note 4.6). There are also instruments available to assess the qualitative aspect of acute itch, such as the adapted version of the Eppendorf Itch Questionnaire [20].

2.4.2 Behavioral Observation

Given that itch is usually defined as an unpleasant behavior associated with the urge to scratch, one possibility to assess the level of itch an organism is currently experiencing is to count the number of spontaneous scratching movements they show in response to pruritic stimulation. While such observational measurements are widely used in non-human itch research, they have only recently found more widespread adoption in studies involving human participants [28–30]. Analysis of the frequency of spontaneous scratching can provide insights into the intensity of the itch experience but can also be used to determine which aspects of an observed scratching behavior are spontaneously produced by a participant [34, 35]. It is important, though, that researchers make clear a priori coding rules about which behavior to code (e.g., to distinguish scratching from habitual self-touches). It is also highly desirable to make video recordings of the behavior, as opposed to on-the-fly coding of behavior, to enable analysis of inter-rater agreement and demonstrate reliability of coding decisions [36].

2.4.3 Physiological Correlates

Studying the physiological correlates of itch in the peripheral nervous system has been a particularly productive approach in understanding the neural fibers involved in itch processing. While most of this work has been carried out with animals, some methods are also applicable for research involving human participants. One of the most direct ways to quantify the neural signals causally contributing to itch in humans is to obtain extracellular recordings of single C-fibers via microneurography [37] (see Chapter 15, this volume). In this approach, thin tungsten needle electrodes are placed inside nerve fascicles, which enable the recording of a single-unit discharge from myelinated and unmyelinated fibers. This has allowed researchers to demonstrate separate peripheral pathways for cowhage and histamine-induced itch in humans [19]. The skin reactions following histamine application (either via skin prick or iontophoresis) provide another physiological correlate of itch in the peripheral nervous system. Skin reactions to histamine are known as a triple response: First, an initial and faint localized transient skin reddening (i.e., initial local vasodilation), followed by a wheal surrounded by a flare. A wheal is a vascular leakage response

Itch

169

to histamine, observed as a raised, often pale and circumscribed dermal edema whereas a flare is an area of reddened skin, reflecting increased superficial perfusion following an axon reflex [38]. The size of wheal and flare can easily be quantified, and serve as an objective indicator that itch induction has been successful. Flare size shows a moderate correlation with subjective itch intensity, at least when histamine is applied via the prick test [7]. The changes in blood perfusion following histamine application can also be continuously monitored using laser doppler flowmetry [39]. The application of cowhage, in contrast, does not result in an axon flare reflex [6]; therefore, no wheal or flares are measurable for this type of itch. 2.4.4 Central Nervous System Correlates

3

Finally, itch can be assessed through measuring neural correlates in the central nervous system. The brain network involved in the processing of acute itch (for review, see [40, 41]) consists of contralateral somatosensory cortices (S1 and S2), bilateral supplementary motor area, insula and anterior cingulate cortices (ACC) as well as ipsilateral inferior frontal gyrus (IFG). It has been argued that functional specialization exists within this cortical network, with a sensory-discriminative role for somatosensory cortices, in particular S1. Consistent with this idea, neural activation in this area varies as a function of histamine concentration [42] and perceived itch intensity is reduced when inhibitory brain stimulation is applied to S1 [40, 43].

Methods

3.1 Chemically Evoked Itch

The focus here will be on those chemical induction methods for which we have the most experience, that is, the histamine prick test and the cowhage rubbing method.

3.1.1 Test

The histamine vials should be stored in a fridge when not in use and be regularly checked for expiration dates. On the day of testing, they should be taken into the testing room well in advance of the test so that the solution is warmed up to room temperature.

Histamine Prick

Preparation Induction

It is helpful to show the lancet to the participant before the actual prick test, in particular, to point out that due to the construction of the lancet (small tip, followed by a broad shoulder), it is impossible to penetrate deep into the skin. In our studies, we have used the volar aspect of the forearm as the target site, about 3 cm proximal to the wrist crease. The target site does not require any preparation, but should be clean and free from wounds, rashes, or other impairments of the skin barrier. Participants should be instructed to keep their arms still and should aim to keep their arm sufficiently supinated so that the target area is level (to avoid the solution running

170

Henning Holle and Donna M. Lloyd

off to the side). Following this, the applicator is taken out of the vial, ensuring that no air bubbles are contained. The experimenter then places a single drop of the solution on the forearm of the participant. It is important to not touch the forearm of the participant with the tip of the applicator while doing so, otherwise the solution in the vial will no longer be sterile. Further advice on standards and safety guidelines for the histamine prick is given in [44]. The experimenter then removes a new sterile lancet from the packaging, taking care to not touch the tip of the lancet. They then prick the skin through the drop in one swift but gentle motion. A small red dot at the end of the trial at the application site (indicating capillary bleeding) usually indicates that too much force was used. Presence of a wheal and flare can be used as visual confirmation that the prick test has been successful. The used lancet should be disposed of immediately into a sharp’s disposal container, and the drop of histamine wiped off. Data Collection

Following the prick test data collection can begin, for example, by asking participants for itch intensity ratings. While the individual time course of a histamine trial can vary greatly from one person to the next, most participants experience at least 5 min of itch following histamine prick. If the skin reaction is mapped, this should occur at a fixed interval (e.g., 10 min after skin prick onset).

Tips and Tricks for the Histamine Prick Test

The histamine prick test is a good way to get started with chemically induced itch research since it does not require preparation of substances. The fact that wheal and flare give visual feedback about whether a trial has been successful is very useful for novice experimenters, as it allows one to repeat failed trials (when no wheal or flare are visible). It should be noted, though, that flares can be difficult to demarcate in darker skin types. When multiple pricks are performed in a single experimental session, great care should be taken to leave sufficient time (at least 30 min) between trials to minimize carry-over effects. The prick test can be administered in a double-blind fashion by having an external person tape a random code over the labels of vials with either histamine or control solution.

3.1.2

Before the experimental session, the researcher needs to count out a dose of spicules (40–50) using tweezers and, if required, a magnifying glass. In our lab, we have placed these doses into small, folded pieces of paper, which are then secured with a Biro clip in advance of testing.

Cowhage

Preparation

Induction

The target site (volar aspect of forearm) is first demarcated by creating a rectangular shape using cellulose tape, with an edge length of 4 cm. The experimenter then places a dose of cowhage into the center of the target area. Subsequently, the experimenter

Itch

171

rubs the cowhage spicules into the skin of the participant in small circular movements for 45 s. Throughout the induction period, it is important to remind the participant to keep the forearm sufficiently supinated to avoid the spicules falling off to the side. Data Collection

After the end of the induction period, data collection can begin, for example, in the form of repeated ratings (see Histamine). At the end of the data collection period, any spicules can be removed from the arm using scotch tape.

Tips and Tricks for the Cowhage Test

We have good experience with wiping the target area using a plain cotton cloth at the end of the trial, which helps to soothe irritated skin. As is the case with histamine, it is important to leave enough time between individual cowhage trials (at least 30 min) to avoid carry-over effects. A placebo control for cowhage can be achieved by autoclaving the spicules, which inactivates the itch eliciting protease mucunain. If one does not have access to an autoclave device, the same effect can also be achieved by steaming the spicules in a pressure cooker. Since a single stray spicule can elicit itch, great care must be taken in the lab to remove all stray cowhage spicules at the end of each experimental session.

3.2 Electrical or Mechanically Evoked Itch

Historically, electrical stimulation was seen as by far the most controllable means of eliciting itch. Electrical stimulation offered the promise of a terminable sensation of itch linearly related to the amount of stimulation, unlike histamine or other chemical inducers, which can be highly variable. Electrically evoked itch has been delivered by insertion of a copper wire into the skin, or via wires or electric plates placed on the skin (using standard metal electrocardiography plate electrodes; [8, 24, 25]). However, the findings of these studies were difficult to replicate, likely because of the differing methods used, and the itch sensation produced could last anywhere between 1 and 12 min, preventing multiple trials. More recently, Ikoma and colleagues [26] have systematically studied electrically evoked itch using transcutaneous electrical stimulation (see also Chapter 7, this volume), which allowed multiple repetitions of trials. For example, in their basic experiment, 50 pulses with 2 ms duration at a frequency of 50 Hz were applied to the left wrist over 30 s. This produced a pure itch sensation in 80% of those tested with a delay between stimulation and sensation of 1 s. Electrical stimulation was shown to be most effective for pulse durations of more than 2 ms and frequencies of more than 50 Hz. They then demonstrated that itch intensity increased in a linear fashion with increasing pulse duration or frequency. In one experiment, pulse durations varied between 0.08 and 8 ms (0.08, 2, 4, and 8 ms) with a fixed frequency of 50 Hz, while in another experiment, stimulus frequencies varied between 2 and 200 Hz (2, 10, 20, 50, 100, and 200 Hz) with a fixed pulse duration of

3.2.1 Basic Paradigms for Electrically Evoked Itch

172

Henning Holle and Donna M. Lloyd

2 ms. These permutations were applied in a random order at intervals of 30 s. Each test stimulus was compared with a reference stimulus (50 Hz, 2 ms pulses) and participants rated itch intensity relative to the initial reference stimulus on a 10-point numerical rating scale (NRS). In addition to relative itch intensity, the extent of alloknesis (i.e., itch evoked by light brushing) and hyperknesis (i.e., itch evoked by pricking) were tested by application of a cotton bud or pinprick to the tested area of skin, respectively. 3.2.2 Basic Paradigms for Mechanically Evoked Itch

In the study by Fukuoka et al. [27] mechanical stimulation was applied to the vellus hairs of the face (chin, cheek, and forehead) and the arm (midpoint between the wrist and elbow on the volar aspect of the forearm). Light touching with a cotton swab acted as the control. Itch intensity on a 10-point NRS was measured (using the fingers if the face was the stimulation site) and the maximum intensity of itch during a 90 s period was assessed. Participants were also asked if they wanted to scratch the site and whether it had the following characteristics: crawling, tickling, stinging, burning, stabbing, and pricking.

3.3 Psychologically Evoked Itch

One of the earliest methods for inducing itch through psychological means was via a short lecture on itch [45]. Unsurprisingly, the audience scratched significantly more during the itch lecture than during a subsequent lecture on relaxation, and itch ratings were highest after the itch lecture for both participants with healthy skin and those with self-reported skin conditions. More recent studies have sought to separate the relative contributions of vision and sound on psychologically evoked itch (see below). The effectiveness of psychological methods of itch induction can be further enhanced via attentional and expectancy-based manipulations [46].

3.3.1 Basic Paradigms for VEI

Itch can be elicited by purely visual means using either static images or video presentation of moving images. Stimulus presentation followed by an itch rating is the most basic paradigm, to which further measures can be added.

Static Images

In our previous studies we have typically grouped stimuli into blocks of 4 trials that are either all itch or all non-itch images, with image content and body part represented equally. Blocks, and the trials within those blocks, are then presented randomly. Practice trials can be used to familiarize participants with the task followed by 64 experimental trials with no repetition [31]. Each image is on screen for 8 s, after which participants are presented with a simple VAS scale showing a horizontal line running left to right from 1 (not itchy at all) to 9 (very itchy) on which participants rate how itchy they feel at that moment. Although we tend to use a mouse button to move a cursor along the scale, you can also use the left/right arrow keys on the keyboard or simply type the number

Itch

173

from 1 to 9 to record the rating. A further adaptation to the basic paradigm is to determine the location of any itches felt. A subsequent screen can be used to display two body outlines (based on the McGill Pain Questionnaire) showing the front and back to enable participants to click on the body where they feel itchy. These are marked with a red circle after each click and then categorized by body location (arm, leg, head, torso) along with the trial in which they occurred. The itch rating forms one part of the outcome measure for VEI paradigms. The second part is formed by the observation and recording of participants’ scratching behavior. This is much trickier, and the method for identifying what constitutes a scratch can vary across different studies, so it’s important to have a systematic approach and be consistent. We have previously used the following criteria to determine what actions are recorded as a scratch: Scraping the skin surface with fingernails and/or rubbing either the skin directly or clothing against the skin in a way that causes friction. Actions not recorded as a scratch include rubbing that would simply move the skin or massage the underlying tissue and tucking back hair or adjusting clothing. If only the sum total of scratches is being recorded (i.e., with no details about the location or duration of the scratch) then we have typically had the experimenter record these in real time during the experiment. The experimenter is in the room with the participant sitting at a distance and out of the participant’s eye-line but still able to clearly see if the participant is scratching. If the location and/or duration of the scratch is being measured, then it makes sense to record scratch behavior using a webcam positioned so that the participant cannot see it directly (as it’s important not to disrupt the participant’s natural scratch behavior) but the experimenter can still see all the participant’s body. Understandably, this is a very tricky thing to control. For the purposes of ethical approval, you need to be explicit that you will be monitoring people’s behavior either directly or through recording their scratch behavior on video, as they need to consent to this. However, in doing so you also draw attention to the fact that their scratch behavior is being monitored, and this may cause some people to suppress or amplify their normal scratch behavior (see Note 4.7). In addition to recording how itchy the participant feels, where on the body they feel itchy and the number and location of any scratches, as a final outcome measure, we have previously asked participants to rate how itchy they thought the person in the picture felt [28, 29]. These results correlated with one another, indicating that empathy with another person’s itch may be influential in creating the VEI effect.

174

Henning Holle and Donna M. Lloyd

Moving Images

Further evidence that VEI may be an empathic experience based on simulating the bodily experience of others comes from [30]. In the first study to use functional magnetic resonance imaging (fMRI, see Chapter 18, this volume) to identify the neural correlates of contagious itch, participants were shown 20 s video clips of bodily scratching or a control condition (Fig. 1). The bodily scratching movement consisted of continuous scraping of the left forearm, left upper arm, chest, right forearm or right upper arm, using four curled fingers of one hand. The control condition was continuous tapping of one of these locations. One male and one female model were filmed with only the waist to the neck visible. The stimulus set consisted of 20 videos in total (2 conditions, scratch vs. no scratch × 5 body locations × 2 models). The experiment used a blocked design. Each block consisted of one 20 s video, followed by a fixation cross presented for 3.3 s. Next, participants were asked to rate the intensity of itchiness induced by the video using a button press on a scale from 0 (not at all) to 7 (extremely). An additional verbal label “moderately” was placed at the mid-point of the scale. This screen was followed by the fixation cross again before the start of the next block. One experimental run consisted of 20 blocks, and participants completed 4 experimental runs during the fMRI part of the study. Viewing scratching activated the “‘itch matrix” in a way similar to when people experience actual bodily itch sensations. The itch matrix includes the anterior insula, primary somatosensory, prefrontal and premotor cortices. These areas are associated with mirroring and simulation of actions (e.g., premotor cortex), sensory aspects of itch (e.g., S1), and top-down predictions of interoceptive signals, which may enable simulation of the feeling of itching (e.g., anterior insula). Other researchers have also used videos of people scratching to induce itch in participants with atopic dermatitis (AD, [47]). Interestingly, they found activation of the supplementary motor area, left ventral striatum and right orbitofrontal cortex, areas of the frontostriatal circuit, which is associated with the urge to scratch. Participants would have to strongly suppress scratch behavior while in the MRI scanner to prevent distortion of the image from movement artifacts. This is a particular problem for MRI studies of itch and especially those studies in populations with pruritic skin conditions. Previous studies have demonstrated that a combination of watching videos of people scratching while participants received either histamine or saline administration caused self-reported itch intensity to increase in all participants. This resulted in a doubling of spontaneous scratching episodes in those with AD, who also appeared to scratch a more widespread area for longer [22].

Itch

175

Fig. 1 A selection of images taken from [30]. (a) Upper arm scratch (left) and control touch (right). (b) Chest scratch (left) and control touch (right). (c) Lower arm scratch (left) and control touch (right) 3.3.2 Basic Paradigms for AEI Auditory Only

Auditory-evoked itch can be investigated using either a categorical (e.g., scratch sounds vs. control sounds) or a parametric manipulation (e.g., effect of linear increase of scratch loudness). One study [33] realized both experimental manipulations by presenting scratching sounds (as experimental items) in addition to rubbing sounds, which served as a control condition. Furthermore, the amplitude of high-frequency sounds (i.e., frequencies above 1000 Hz) was either attenuated, enhanced, or unchanged. The effect of this amplitude manipulation was that the sounds either had an edge (in the case of enhanced high-frequency amplitude) or a slightly muffled quality (in the case of attenuated high-frequency amplitude). These sounds were presented to a group of patients with psoriasis as well as a group of healthy controls who were asked to rate the amount of induced itch. Results showed that, in healthy controls, scratching sounds induced greater itch than rubbing

176

Henning Holle and Donna M. Lloyd

sounds. Additionally, the magnitude of itch was found to vary as a function of the high-frequency amplitude. Finally, there was a group by amplitude interaction, with patients showing greater vulnerability to the high-frequency enhanced sounds relative to controls. In summary, we hope to have laid bare some of the challenges of conducting itch research but it was not our intention to dissuade anyone from undertaking this type of research! If basic science is to be of any benefit to society it needs scientists to take up the challenge and translate the findings from studies of acute experimental itch into effective clinical interventions to provide a better understanding of pruritic disease states and how best to treat them.

4 4.1

Notes Slowness of Itch

Probably the biggest methodological challenge, at least when chemical itch-induction methods are used, is the inherent slowness of itch. The sluggish nature of the itch response, where one trial lasts at least 5 min followed by at least 20 min of waiting before another trial can commence, has a number of adverse consequences for researchers. First, the length of time makes it impossible to determine individual sensory thresholds for itch, as is standard procedure in other sensory research domains (see Chapters 1, 2, 3, 4, 5, 6, 7, 9 and 10, this volume). Because the pruritic dose cannot be adjusted to a person’s threshold, the same dose (e.g., 1% histamine) elicits an itch response with considerable inter-individual variation, with some participants experiencing only a very mild itch that has completely subsided after 3 min, whereas others experience an intense itch lasting 15 min or longer. A second direct consequence of the slowness is that it severely limits the number of data points available for analysis. The combination of both factors results in considerable random unexplained variation in each itch study, which severely limits the power of the statistical analysis.

4.2 When Itch Becomes Pain

Drezga et al. [42] used different concentrations of histamine solution to determine the relationship between dose concentration and itch intensity. Up to a level of 1% concentration, they observed a linear and positive relationship between dose and itch intensity. However, at very high concentrations, or when injected directly into the skin, other authors have reported that histamine tends to induce pain rather than itch (e.g., [7]).

4.3 Dose Variability of Cowhage

Induction methods based on cowhage face an additional random source of variation since a recombinant version of the itch-inducing protease mucunain is not yet commercially available. The amount of mucunain in a cowhage spicule varies depending on the origin of the plant [48]; furthermore, only a fraction of the number of

Itch

177

spicules applied to the skin become lodged into the skin and subsequently start to induce itch [34]. Thus, unlike in the histamine prick test, where the researcher can have a relatively good estimate of how much of the pruritogen went beyond the skin barrier [9], this is unknown for cowhage. Development of a commercially available recombinant mucunain would be a big step forward here [48]. 4.4 Technical Aspects of Producing Auditory and Visual Itch Stimuli

From our experience of auditory itch, it can be a technical challenge to make good sound recordings of scratching sounds and there were some surprising insights into which sounds are most itch inducing. Similarly, it is also very challenging to produce a wellcontrolled stimulus set for VEI and several factors such as the novelty, arousing, or unpleasant nature of the images need to be accounted for.

4.5 Controlling the Visual Stimulus

You may also have to pixelate the model’s face to remove cues to emotional reactions.

4.6 Prior Itch Experience and Choice of Rating Scales

For studies that rely on itch intensity ratings as a major outcome variable, there are some important considerations regarding prior itch experience of participants as well as the choice and implementation of the rating scale to be used. Itch ratings rely on verbal anchor points to guide participants in their use of the rating space. For example, many itch studies use a visual analog scale where the top end of the scale is labeled “worst itch imaginable”. However, healthy volunteers coming into the lab to take part in a study may not have had previous intense itch experiences. For example, we have found in pilot studies that many UK students have never experienced a mosquito bite. This lack of previous experiences makes it impossible to use an itch rating scale in a meaningful way because the verbal anchor points cannot serve their purpose as a reference point. To mitigate this problem, we recommend additional time be spent at the beginning of a session to familiarize the participant with the rating scale, including rating of familiar events (e.g., itchy scalp, mosquito bite, wound itch). If the paradigm involves chemically induced itch, it is useful to have participants experience and rate an initial familiarization stimulus before data collection begins [49, 50]. For studies focusing on chemically induced itch, we recommend hybrid scales, such as the General Labelled Magnitude Scale [21], since they provide more reliable itch intensity estimates than classic visual analog scales [49]. If participants are asked to make repeated ratings (e.g., over the course of a whole itch trial) then care should be taken to ensure that the ratings are independent of each other. In particular, participants should not be able to see previous ratings as this increases the risk that participants only provide ordinal level data (e.g., has itch intensity increased or has it decreased relative to the last rating

178

Henning Holle and Donna M. Lloyd

provided?) as opposed to the intended ratio level data. This is an issue, for example, when continuous rating devices are used (e.g., continuous computerized visual analog scales, CoVAS). 4.7

Other Factors

There may be several other factors that contribute to how a person experiences itch:

4.7.1 External Environment

For example, testing in a cold room dulls the sensation whereas heat increases it. Ideally, the temperature and humidity of the testing room should be recorded.

4.7.2 Demand Characteristics

Unfortunately, simply asking people about itch can often make them itchy by focusing attention on normally sub-perceptual sensations and thereby bringing them into conscious awareness. This may be particularly important in those with an ongoing pruritic skin condition [29], although there is evidence to suggest individual differences in a non-pruritic person’s capacity to experience itch (the so-called “itchish” person; [51]). Tomasch and colleagues [51] showed an increased ratio of small C-fibers to large A-fibers (3–12 times as many) in those people who rate itch more intensely. Obviously, evidence of an existing pruritic skin condition should be checked during recruitment. Finally, it is sometimes difficult to differentiate self-touch from scratch in video recordings, and participants may produce more self-touch because they feel uncomfortable rather than a genuine itch sensation that prompts the behavior.

References 1. Cole J (2016) Losing touch: a man without his body. Oxford University Press, UK € 2. To¨ro¨k L (1907) Uber das Wesen der Juckempfindung. Zeitsch Psychol 46:23–35 3. von Frey M (1922) Zur Physiologie der Juckempfindung. Arch Neerl Physiol 7:142–145 4. Handwerker H (2014) Itch hypotheses: from pattern to specificity and to population coding. In: Carstens ATE (ed) Itch: mechanisms and treatment. CRC Press/Taylor & Francis, Boca Raton 5. Heyer G, Hornstein OP, Handwerker HO (1989) Skin reactions and itch sensation induced by epicutaneous histamine application in atopic dermatitis and controls. J Invest Dermatol 93:492–496. https://doi.org/10. 1111/1523-1747.ep12284051 6. Sikand P, Shimada SG, Green BG, LaMotte RH (2009) Similar itch and nociceptive sensations evoked by punctate cutaneous application of capsaicin, histamine and cowhage. Pain 144:

66–75. https://doi.org/10.1016/j.pain. 2009.03.001 7. Darsow U, Ring J, Scharein E, Bromm B (1996) Correlations between histamineinduced wheal, flare and itch. Arch Dermatol Res 288:436–441. https://doi.org/10.1007/ BF02505231 8. Shelley WB, Arthur RP (1957) The neurohistology and neurophysiology of the itch sensation in man. AMA Arch Derm 76:296–323. https://doi.org/10.1001/archderm.1957. 01550210020004 9. Pepys J (1975) Skin testing. Br J Hosp Med 14: 412–417 10. Pfab F, Valet M, Sprenger T et al (2006) Shortterm alternating temperature enhances histamine-induced itch: a biphasic stimulus model. J Invest Dermatol 126:2673–2678. https://doi.org/10.1038/sj.jid.5700577 11. Papoiu AD, Coghill RC, Kraft RA et al (2012) A tale of two itches. Common features and notable differences in brain activation evoked

Itch by cowhage and histamine induced itch. NeuroImage 59:3611–3623. https://doi.org/10. 1016/j.neuroimage.2011.10.099 12. Magerl W, Westerman RA, Mohner B et al (1990) Properties of transdermal histamine iontophoresis: differential effects of season, gender, and body region. J Invest Dermatol 94:347–352. https://doi.org/10.1111/ 1523-1747.ep12874474 13. Van de Sand M, Sprenger C, Buchel C (2015) BOLD responses to itch in the human spinal cord. NeuroImage 108:138–143. https://doi. org/10.1016/j.neuroimage.2014.12.019 14. Hsieh JC, Hagermark O, Stahle-Backdahl M et al (1994) Urge to scratch represented in the human cerebral cortex during itch. J Neurophysiol 72:3004–3008. https://doi.org/10. 1152/jn.1994.72.6.3004 15. Davidson S, Zhang X, Yoon CH et al (2007) The itch-producing agents histamine and cowhage activate separate populations of primate spinothalamic tract neurons. J Neurosci 27: 10007–10014. https://doi.org/10.1523/ JNEUROSCI.2862-07.2007 16. Reddy VB, Iuga AO, Shimada SG et al (2008) Cowhage-evoked itch is mediated by a novel cysteine protease: a ligand of protease-activated receptors. J Neurosci 28:4331–4335. https:// doi.org/10.1523/JNEUROSCI.0716-08. 2008 17. Johanek LM, Meyer RA, Hartke T et al (2007) Psychophysical and physiological evidence for parallel afferent pathways mediating the sensation of itch. J Neurosci 27:7490–7497. https://doi.org/10.1523/JNEUROSCI. 1249-07.2007 18. Johanek LM, Meyer RA, Friedman RM et al (2008) A role for polymodal C-fiber afferents in nonhistaminergic itch. J Neurosci 28:7659– 7 6 6 9 . h t t p s : // d o i . o r g / 1 0 . 1 5 2 3 / JNEUROSCI.1760-08.2008 19. Namer B, Carr R, Johanek LM et al (2008) Separate peripheral pathways for pruritus in man. J Neurophysiol 100:2062–2069. https://doi.org/10.1152/jn.90482.2008 20. Kosteletzky F, Namer B, Forster C et al (2009) Impact of scratching on itch and sympathetic reflexes induced by cowhage (Mucuna pruriens) and histamine. Acta Derm Venereol 89: 2 7 1 – 2 7 7 . h t t p s : // d o i . o r g / 1 0 . 2 3 4 0 / 00015555-0624 21. LaMotte RH, Shimada SG, Green BG et al (2009) Pruritic and nociceptive sensations and dysesthesias from a spicule of cowhage. J Neurophysiol 101:1430–1443. https://doi.org/ 10.1152/jn.91268.2008

179

22. Papoiu ADP, Wang H, Coghill RC et al (2011) Contagious itch in humans: a study of visual ‘transmission’ of itch in atopic dermatitis and healthy subjects. Br J Dermatol 164:1299– 1 3 0 3 . h t t p s : // d o i . o r g / 1 0 . 1 1 1 1 / j . 1365-2133.2011.10318.x 23. Lewis Z, George DN, Cowdell F et al (2019) Effects of short-term temperature change in the innocuous range on histaminergic and non-histaminergic acute itch. Acta Derm Venereol 99:188–195. https://doi.org/10. 2340/00015555-3077 24. Edwards AE, Shellow WV, Wright ET, Dignam TF (1976) Pruritic skin diseases, psychological stress, and the itch sensation. A reliable method for the induction of experimental pruritus. Arch Dermatol 112:339–343 25. Tuckett RP (1982) Itch evoked by electrical stimulation of the skin. J Invest Dermatol 79: 3 6 8 – 3 7 3 . h t t p s : // d o i . o r g / 1 0 . 1 1 1 1 / 1523-1747.ep12529734 26. Ikoma A, Handwerker H, Miyachi Y et al (2005) Electrically evoked itch in humans. Pain 113:148–154 27. Fukuoka M, Miyachi Y, Ikoma A (2013) Mechanically evoked itch in humans. Pain 154:897–904. https://doi.org/10.1016/j. pain.2013.02.021 28. Lloyd DM, Hall E, Hall S, McGlone FP (2013) Can itch-related visual stimuli alone provoke a scratch response in healthy individuals? Br J Dermatol 168:106–111. https://doi.org/10. 1111/bjd.12132 29. Lloyd DM, Dodd R, Higgins C et al (2017) Are sex and history of pruritic skin conditions factors which affect the phenomenon of visually evoked itch? An exploratory study. Itch 2: e 1 0 . h t t p s : // d o i . o r g / 1 0 . 1 0 9 7 / i t x . 0000000000000010 30. Holle H, Warne K, Seth AK et al (2012) Neural basis of contagious itch and why some people are more prone to it. Proc Natl Acad Sci U S A 109:19816–19821. https://doi.org/10. 1073/pnas.1216160109 31. Young M (2018) Breaking the itch-scratch cycle. PhD thesis, University of Leeds 32. Bradley MM, Lang PJ (1994) Measuring emotion: the self- assessment manikin and the semantic differential. J Behav Ther Exp Psychiat 25:49–59. https://doi.org/10.1016/ 0005-7916(94)90063-9 33. Swithenbank S, Cowdell F, Holle H (2016) The role of auditory itch contagion in psoriasis. Acta Derm Venereol 96:728–731. https://doi. org/10.2340/00015555-2320 34. Papoiu AD, Tey HL, Coghill RC et al (2011) Cowhage-induced itch as an experimental

180

Henning Holle and Donna M. Lloyd

model for pruritus. A comparative study with histamine-induced itch. PLoS One 6:e17786. https://doi.org/10.1371/journal.pone. 0017786 35. Ward J, Burckhardt V, Holle H (2013) Contagious scratching: shared feelings but not shared body locations. Front Hum Neurosci 7:122. https://doi.org/10.3389/fnhum.2013. 00122 36. Holle H, Rein R (2015) EasyDIAg: a tool for easy determination of interrater agreement. Beh Res Methods 47:837–847. https://doi. org/10.3758/s13428-014-0506-7 37. Schmelz M, Schmidt R (2010) Microneurographic single-unit recordings to assess receptive properties of afferent human C-fibers. Neurosci Lett 470:158–161. https://doi.org/ 10.1016/j.neulet.2009.05.064 38. Andersen HH, Elberling J, Arendt-Nielsen L (2015) Human surrogate models of histaminergic and non-histaminergic itch. Acta Derm Venereol 95:771–777. https://doi.org/10. 2340/00015555-2146 39. Hovell CJ, Beasley CR, Mani R et al (1987) Laser Doppler flowmetry for determining changes in cutaneous blood flow following intradermal injection of histamine. Clin Allergy 17:469–479. https://doi.org/10.1111/j. 1365-2222.1987.tb02041.x 40. Jones O, Schindler IC, Holle H (2017) Assessing acute itch intensity: general labelled magnitude scale is more reliable than classic visual analogue scale. Acta Derm Venereol 97:375– 3 7 6 . h t t p s : // d o i . o r g / 1 0 . 2 3 4 0 / 00015555-2584 41. Mochizuki H, Kakigi R (2015) Central mechanisms of itch. Clin Neurophysiol 126: 1650–1660. https://doi.org/10.1016/j. clinph.2014.11.019 42. Drzezga A, Darsow U, Treede RD et al (2001) Central activation by histamine-induced itch: analogies to pain processing: a correlational analysis of O-15 H2O positron emission tomography studies. Pain 92:295–305. https://doi.org/10.1016/s0304-3959(01) 00271-8

43. Nakagawa K, Mochizuki H, Koyama S et al (2016) A transcranial direct current stimulation over the sensorimotor cortex modulates the itch sensation induced by histamine. Clin Neurophysiol 127:827–832. https://doi.org/ 10.1016/j.clinph.2015.07.003 44. Heinzerling L, Mari A, Bergmann KC et al (2013) The skin prick test - European standards. Clin Transl Allergy 3:3. https://doi. org/10.1186/2045-7022-3-3 45. Niemeier V, Kupfer J, Gieler U (2000) Observations during an itch-inducing lecture. Dermatol Psychosom 1:15–18 46. van Laarhoven A, Holle H (2020) Optimizing audiovisual itch induction: the role of attention and expectancy. Br J Dermatol 182:1088– 1089. https://doi.org/10.1111/bjd.18596 47. Schut C, Mochizuki H, Grossman SK et al (2017) Brain processing of contagious itch in patients with atopic dermatitis. Front Psychol 8:1267. https://doi.org/10.3389/fpsyg. 2017.01267 48. Andersen HH, Sørensen AR, Nielsen GA et al (2017) A test-retest reliability study of human experimental models of histaminergic and non-histaminergic itch. Acta Derm Venereol 97:198–207. https://doi.org/10.2340/ 00015555-2502 49. Jones O, Schindler I, Holle H (2019) Transcranial magnetic stimulation over contralateral primary somatosensory cortex disrupts perception of itch intensity. Exp Dermatol 28:1380– 1384. https://doi.org/10.1111/exd.13803 50. Phan NQ, Blome C, Fritz F et al (2012) Assessment of pruritus intensity: prospective study on validity and reliability of the visual analogue scale, numerical rating scale and verbal rating scale in 471 patients with chronic pruritus. Acta Derm Venereol 92:502–507. https://doi.org/ 10.2340/00015555-1246 51. Tomasch J, Britton WA (1956) On the individual variability of fibre composition in human peripheral nerves. J Anat 90:337–349

Chapter 9 Experimental Framework and Methods for the Assessment of Skin Wetness Sensing in Humans Charlotte Merrick, Rochelle Ackerley, and Davide Filingeri Abstract The study of the human ability to both detect the presence and estimate the amount of wetness on the skin has grown in scientific interest over the last century, due to the implication of wetness in comfort and skin health. In 1900, Bentley demonstrated that skin wetness is detected based on touch and temperature stimuli combining to produce sensations of liquidity, and that wetness perception increases with cold touch. It has since been demonstrated that, in the absence of a skin hygroreceptor (i.e., wetness receptor) in humans, the biophysical effects of moisture on the skin—conductive heat transfer and mechanical interaction—excite specific cutaneous mechanoreceptors and thermoreceptors. The resulting afferent signals are centrally integrated to generate our perception of skin wetness. As well as providing a theoretical foundation for understanding this aspect of somatosensation, these insights have helped develop a methodological framework for the study of human skin wetness sensing, which relies on assessing the independent and interactive effects of thermo-tactile stimulation of the skin in the presence of a liquid. This chapter will provide an overview of the experimental framework and methods available to evaluate the biophysical and psychophysical responses to controlled dry and wet stimuli applied to skin, and the resulting wetness perception. We will use example scenarios of skin-moisture interactions (e.g., arising from contact with a wet surface or from sweat production), to critically evaluate the methods, noting their accuracy, reliability, and efficiency, and discuss their limitations and commonly encountered difficulties. It is hoped that these considerations will guide and further develop research of this relatively little-investigated, yet fundamental, aspect of somatosensation. Key words Wetness, Moisture, Hygroreception, Temperature, Thermosensation, Touch, Mechanosensation, Biophysics, Psychophysics, Skin

1

Introduction Humans experience the perception of wetness every day, for example, when having a shower in the morning, sweating during exercise, or coming into contact with a wet object. The ability to sense wetness enriches our perception, bringing a new depth to sensations, which enhances the way we interpret the world around us. Furthermore, it provides an important signal for behavioral responses, aimed at maintaining both thermoregulatory

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_9, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

181

182

Charlotte Merrick et al.

homeostasis (e.g., sweat-induced skin wetness drives thermal discomfort and triggers cool-seeking behaviors) [1] and skin health (e.g., incontinence-associated wetness drives discomfort and triggers drying behaviors) [2]. This ability to detect wet stimuli on the skin is termed wetness perception. Wetness is currently believed to be a percept generated in the central nervous system, as humans do not seem to possess hygroreceptors, which are specific receptors for encoding wetness [3, 4]. Such specialized neurons are found in certain insects, and detect and quantify relative humidity. This is in contrast to other sensations, such as touch and temperature, where different types of mechanoreceptor and thermoreceptor encode specific aspects of skin stimulation [5, 6]. Given that humans, as well as other animals, frequently come into contact with moisture, and that this interaction is essential in our lives, it is perhaps surprising that hygroreceptors do not appear to exist in mammals [7]. Given the high sensitivity that humans have for detecting wetness, it is of interest to probe how touch and temperature signals convey such a sensation. 1.1 History of Wetness Perception

In 1900, Bentley [8] proposed that the sensation of liquidity was made up of specific components that formed a perception of wetness when combined in specific ways. To understand how the sensation was formed, Bentley attempted to reconstruct the sensation of touching a liquid (wetness perception) through synthetic experimentation, where he manipulated the substance touched and its temperature to create a wetness illusion. At one point, Bentley writes, “molasses, benzine and even mercury passed under certain thermal conditions for water: an indication of how widely the organism is obliged to draw upon its resources for the completion of so simple a perception as that of a liquid” (1900, pp. 415–416). Although the use of dangerous liquids is now not acceptable, this work is still very relevant in modern-day wetness perception research, where we vary specific touch and temperature factors to increase or decrease perceived wetness. Bentley also demonstrated the difference between passive and active wetness perception, where passively applied liquids were always felt as wet, but with active touch when there was an interaction between liquid on a solid surface and the skin, other percepts such as oiliness appeared. Bentley concluded that actual moisture on the skin is not even sufficient to generate the perception of wetness, as a liquid applied with little change in pressure and/or temperature did not evoke wetness. Further, a wetness illusion could be created where a thin rubber sheath was placed on the finger, which was then lowered into a liquid: the participants truly felt wetness, especially when the water was cold. A study by Sullivan in 1923 found that, in general, the perception of liquidity was evoked by combinations of pressure and temperature, while the perception of solidity was transmitted by

Wetness Perception in Humans

183

pressure alone [9]. Further, the perception of “semi-liquidity” (viscous liquid) was typically a blend of intense pressure and temperature, whereas liquidity was produced from weak pressure and temperature. In related work in the same period, explorations were made into other concepts associated with liquidity such as stickiness [10], clamminess [11], and oiliness [12]. These were linked more to tactile aspects such as the stick-slip phenomenon, in which friction varies during an interaction due to changes in surface tension and adhesive forces, although clamminess was also associated with coolness, and oiliness with warmth. Later work added other related concepts, such as spreadability and viscosity, occurring from specific interactions between wet surfaces and the skin that are particularly modified by the thermal conductivity of fluids [13]. Work by Gagge and colleagues [14] introduced ideas of thermal comfort, using perceptual scales for both ratings of comfort and temperature sensations. This linked together the concepts of sweating, heat transfer, and thermoregulatory physiology. Sweeney and Branson later took this further by investigating moisture sensation in sensorial comfort, by using both absolute and difference thresholds [15] and magnitude estimation to rate moisture sensation [16], including considering comfort in-wear factors. The exploration of wetness has evolved from perceptual descriptions to psychophysical ratings. These studies have allowed us to develop the investigation of wetness on different levels, such as by accurately manipulating temperature and also in the perception of sweating, more of which we cover below. 1.2 Directions in Wetness Perception Research

There are many directions in current wetness perception research, which can be discussed most simply when considering the source of moisture (e.g., from an external source, sweating) and the research question. For example, one may be interested in investigating the minimum amount of skin wetness that can be detected by the finger pad, as this is relevant to optimize the design of absorbent products that involve individuals touching a wet material to determine its state—consider a parent touching an infant’s diaper to check its dryness. In addition, one may be interested in investigating regional differences across the body in the sensitivity to wet fabrics applied on the skin at different moisture saturations. These differences are relevant to optimize the design of sports clothing to maintain comfort during physical activity—consider an individual running and experiencing a sweaty t-shirt sticking to different parts of their torso. Different biophysical and psychophysical assessment methods can therefore be derived depending on whether a detection threshold or magnitude estimation approach is required, that is, whether one aims to determine the sensitivity of the skin to moisture, or the participant’s ability to differentiate among moisture levels and assess their relative salience.

184

Charlotte Merrick et al.

For the purpose of devising appropriate testing methodologies, the interaction between the skin and moisture can broadly be divided into two categories: one where the source of moisture is external to the body such as touching a cloth to determine if it is wet, or when we feel rain on our skin; and one where the source of moisture is internal to the body, such as when sweat, urine, teardrops, or other bodily fluids are produced. This initial differentiation is essential, as it determines the scenario in which skinmoisture interactions will occur, and the resulting thermo-tactile inputs that are likely to take place at the skin interface. Ultimately, this initial evaluation will inform the development of specific hypotheses as well as the appropriate experimental methods. In short, thermal factors may be more relevant for externally generated and mechanical factors than for internally generated moisture (Fig. 1). When moisture is externally generated (e.g., when touching a wet object), one may expect that thermal factors like moisture temperature and the related heat transfer, may be highly variable, and therefore play a larger role in determining variance in our wetness perception. By contrast, when moisture is internally generated (e.g., when sweat is produced), the thermal conditions are more constrained—sweat temperature is likely to be equal to skin temperature). Accordingly, the biophysical and psychophysical assessment of thermal cues when moisture is externally generated may take priority over the evaluation of mechanical components of the stimulus. Conversely, when moisture is internally generated (e.g., incontinence) one may expect that mechanical factors, such as the adhesion and friction of wet continence pads to the skin, may play a greater role in surface wetness perception than thermal factors such as elevated temperature and humidity in the microclimate surrounding the skin and the incontinence pad. In this case, the biophysical and psychophysical assessment of tactile cues when moisture is internally generated may take priority over the evaluation of thermal components. The methodological framework outlined above is presented schematically in Fig. 1, which provides a flow diagram for method selection. In this chapter, we use three examples to highlight potential methodological pathways. These examples are not exhaustive, and we emphasize that, as this field expands, alternative methodologies and approaches could be developed. Yet, we believe that the theoretical foundation of these examples and related methods cover a broad spectrum of conditions and could form the basis for further methods development. The first of these examples focuses on evaluating wetness detection thresholds from touch interactions between the index finger pad and externally wet stimuli, and assessing threshold modulations at different moisture temperatures. The second focuses on comparing the wetness

Wetness Perception in Humans

185

Physical skin wetness Mechanical component

Physical

Physical Mechanical interaction

Biophysical

Biophysical

Change in skin temperature

Skin deformation

Neurophysiologic

Neurophysiologic

Afferent inputs from thermoreceptors

Afferent inputs from mechanoreceptors

Thermal sensory cue

Processes

Heat transfer

Skin wetness perception

Mechanical sensory cue

THEORETICAL FRAMEWORK

Thermal component

Fig. 1 The methodological framework outlined in the chapter presented schematically as a flow diagram

186

Charlotte Merrick et al.

perception of stimuli applied to various locations across the body and the production of associated body wetness perception maps. Finally, post-exercise skin wetness perception will be examined, considering sweat production and different tactile inputs. The production of a consistently implemented methodology is critical in this research field as it not only allows for a wider comparison of scientific constructs, but also presents possibilities to develop specific informative and diagnostic tests. This is relevant in industrial applications, for example, the improvement of moisture management products [17], within clinical contexts such as the design of early diagnostic tests for individuals with sensory disorders [18], and in leisure and performance applications such as the management of sweat patterns in sports clothing [19].

2 2.1

Materials Hardware

In a typical experiment exploring wetness perception, we may need to control and measure various factors, such as air temperature/ humidity, object/skin temperature, as well as mechanical aspects of stimuli, such as pressure applied to the skin. Climate chambers can be particularly useful to control such conditions. The typical equipment that may be used in wetness perception experiments [18, 20– 22] include the following: – Force plate (e.g., 0–5 N). – Thermal plate and controller (e.g., 10–45 °C). – Thermal probe (consider the size and temperature of the probe, e.g., 10–45 °C) and controller (e.g., Physitemp Instruments Inc., Clifton, NJ, USA). – Water bath(s) to change the temperature of applied liquids (e.g., 15–45 °C) (e.g., Techne FRB2D, Cole-Parmer, Stone, UK). – Thermocouples to measure surface temperature (e.g., 0.08 mm wire diameter, 40 Gauge) (e.g., 5SRTC-TT-TI-40-2 M, Omega, Manchester, UK). – Indoor air quality monitor or combined thermometer and hygrometer to measure air temperature and humidity (e.g., 440 dP, Testo, Lenzkirch, Germany). – Infrared thermometer to measure surface temperature remotely (e.g., TG56, FLIR Systems, Wilsonville, OR, USA). – Items to obscure the stimulation from the participant, depending on the type of perceptual assessment scales used (e.g., screen, glasses, headphones). – Scale and stadiometer or tape measure (for measuring weight and height of the participant).

Wetness Perception in Humans

187

– Climatic chamber. – Water-perfused suit. 2.2 Consumables [18]

– Skin tape (e.g., 25 mm width Transpore, 3 M, Loughborough, UK). – Graduated plastic syringes. – Tissues.

2.3 Environmental Conditions

3

Depending on the experiment, specific thermal and humidity conditions may be required, but typical, standard, or neutral conditions would be a room between 20 and 25 °C and roughly 50% relative humidity depending on clothing, with atmospheric pressure around 100 kPa.

Methods Investigations of wetness perception rely heavily on the use of single-blind psychophysical testing (see Notes 4.1 and 4.2, see also Chapters 1, 2, 3, 4, 5, 6, 7, and 8, this volume). This is a branch of sensory examination which focuses on the quantification of specific modalities such as haptic touch, thermal inputs, and auditory cues. It relies on a stimulus-response paradigm (see Note 4.3), in which a series of specific stimuli are prepared by an experimenter and are introduced to a participant wearing a blindfold or otherwise visually obscured from the stimulus (see Note 4.4). The participant’s sensory responses are recorded, and the collation of these responses across various conditions allows the target sensory modality to be further assessed according to the chosen psychophysical test [23]. Psychophysical testing can be broadly divided into two categories depending on the desired nature of the outcome: threshold determination and magnitude estimation (see Notes 4.3, 4.4, and 4.5). The first of these is threshold determination, which uses a series of stimuli varying across a range of predetermined characteristics. In classical psychophysics, the stimuli can be introduced to the participant in several ways: the method of limits that increase the magnitude of a stimulus until a specific response is reached; the method of constant stimuli in which the stimuli are introduced in a random or balanced order; and the method of adjustment, in which the participant adjusts the stimulus magnitude to a certain detection level, or to match another stimulus. Thresholds can also be estimated using adaptive staircase procedures, in which changes to the stimulus are made according to the participant’s previous response(s), thus allowing the progressive and specific testing of a particular threshold [24, 25]. The establishment of the detection threshold can either be in its absolute state, showing the minimum

188

Charlotte Merrick et al.

magnitude of a stimulus required to elicit a positive response, or as a differential threshold, which denotes the minimum difference between stimuli for them to be perceived as different [23]. The threshold determination methodologies in this chapter will focus on absolute values. A second form of psychophysical testing is magnitude estimation. This can also be approached in two ways, the most common being the sequential presentation and subsequent rating of a stimulus. This involves, for example, a continuous visual analog scale on which the participant must mark a point to reflect their perception, an unbounded rating scale where the participant uses their own ratings to quantify a percept that can subsequently be normalized, or a graduated Likert scale forming of a series of descriptors with numerical counterparts. Alternatively, a participant may be presented with a stimulus and be asked to generate one of equal magnitude using the method of adjustment [26], but this technique is likely to be of less use in wetness perception due to the methods required to create the stimulus. 3.1

Participants

Participants are typically screened for exclusion criteria to ensure the generalizability of findings to the healthy population, unless a different population is specifically required. For example, it is preferable to recruit individuals from a specified age range, such that they will understand and comply with testing protocols but not have any degenerative sensory disorders [27]. For example, participants under 30 years old may be preferred, as both touch and temperature detection sensitivity decrease with age [28, 29]. Lifestyle factors can also be considered, such as the recruitment of non-smoking individuals. Prolonged smoke exposure can make individuals more susceptible to dermatological diseases and actively contributes to peripheral neuropathy, which would interfere with sensation [30]. Body mass can be considered, as participants having a body mass index (BMI) above 30 kg.m-2, may have some degree of peripheral nerve impairment [31]. The same applies to alcohol consumption above the recommended weekly alcohol intake, as exceeding this limit can result in both physical and psychological interference [32]. It is also preferable for individuals to not be taking any medications long-term, as the somatosensory effects of these may be unknown or hard to quantify. Finally, participants should not have any long-term somatosensory disorder, such as peripheral neuropathy, which may interfere with perceptual tasks. While exclusion criteria are generally quite uniform and similar between tests, participants may also be recruited for specific characteristics. For example, in clinically focused studies, there may be a need to recruit target individuals along with matched controls. Participants may also be required for sensory comparison studies, such as identifying perceptual differences between males and

Wetness Perception in Humans

189

females, from different countries or environmental factors, or women at pre- and post-partum time points. Where appropriate, prior to the scheduled testing, participants’ body mass and height should be recorded with a digital scale and stadiometer in order to determine their BMI and hence confirm eligibility for the study. This also applies to the completion of a health screen questionnaire to highlight any matches with the exclusion criteria. The number of participants required can be established using a sample size calculation based on data from previous or pilot studies. 3.2 Experimental Protocol 3.2.1 Example 1: Wetness Detection— External Moisture and Active Touch

The investigation aimed to establish the wetness detection threshold of the human index finger pad during active touch as a function of moisture temperature [17]. This was achieved using a singleblind repeated measures design. Stimuli consisted of textile samples varying in the applied liquid content across the range of 0–50 mL, which was determined during pilot studies. Each of the six wetness levels were repeated multiple times and introduced in a counterbalanced order. Each participant attended four experimental sessions, with each session representing a different temperature condition. At the beginning of each experimental session, participants were familiarized with the protocols and a calibration was done to show the extremes of the perceptual scales and provide experimental context. During experimental sessions, participants touched the sample using only the index finger pad and provided responses using a forced-choice task in which they had to select from either “dry” or “wet” descriptors on a digital screen. Prior to the start of each experimental session, 1 L of water was placed into a small manually controlled thermal chamber which maintained the temperature at either 25.1 °C, 29.2 °C, 33.4 °C, or 37.7 °C ± 0.1 °C. These temperatures were established during pilot studies to account for thermal changes which occur during sample preparation, such that the initial participant contact with the substrate would be at 25 °C, 29 °C, 33 °C, or 37 °C. Different volumes were applied to individual stimuli to moisten them prior to participant interaction using 0 mL, 10 mL, 20 mL, 30 mL, 40 mL, or 50 mL of water applied with a graduated syringe. When participants arrived at the laboratory, their height was recorded with a stadiometer, and weight recorded with a scale to confirm eligibility for the study. Participants were seated for the duration of the experiments. Participants were familiarized with the study protocols and rating scales to ensure they were correctly and consistently used. The practice protocol involved four different stimulus combinations; cold-wet, warm-wet, cold-dry, and warmdry, which demonstrated the extremes of each possible wetness and thermal combination. Each stimulus was presented under standard test conditions and the most appropriate response on the

190

Charlotte Merrick et al.

m

m 00

450 mm

6

m 0m

160

mm

500 mm

21

Fig. 2 L-shaped obscuring screen used to limit visual cues available to participants. The screen has two apertures to allow for variations in participant hand dominance (left) and a modifiable thermal plate used to maintain participants’ skin temperature between stimuli (right)

Fig. 3 A square thermal probe which was wrapped in wetted cotton fabric and applied to participants’ skin

psychophysical form was shown, providing a frame of reference as well as acquainting participants with procedures. Following the familiarization, a single thermocouple was affixed to the center of the index finger pad using surgical skin tape, ensuring the thermocouple tip was in contact with the skin but not covered by tape. Participants then placed their non-dominant hand through an aperture in an L-shaped screen which obscured the experimental setup from sight and hence limited visual cues (Fig. 2). The base of the screen included a foam mat to reduce conductive heat transfer and could be inverted to allow for left- or right-hand dominance. Participants placed their index finger on a fixed-position thermal plate (Fig. 3) which maintained a neutral skin temperature (Tsk) at 33 °C and hence established a thermal baseline. When the stimulus had been prepared, the participant was given the

Wetness Perception in Humans

191

command “contact” and moved their hand from the thermal plate to contact the stimulus at a static resting pressure. Participants had previously been informed of the commands and the positioning and movement of the finger had been demonstrated and practiced. The stimulus was always positioned correctly below the finger, and was slightly adjusted before contact if necessary. Upon transient contact participants completed a digital form to record their perceptions, taking no more than 3 s. Both a dichotomous response method (dry/wet) and a 100 mm visual analog scale (very dry to very wet) were used. The responses associated with the dichotomous method were assigned binary scores for subsequent analyses, with a “dry” response as 0 and a “wet” response as 1. After a contact period of 3 s, the participant was prompted to remove their finger from the stimulus using the command “lift.” Post-contact perceptual assessments, analogous to those used during contact, were completed again within 3 s. When finished, participants used the command “done,” at which point the stimulus would be replaced with a cotton towel. The participant was instructed “dry,” and would press their index finger statically onto the dry towel to collect residual water for 5 s. This was repeated regardless of wetness to prevent any learning effect or bias. Before the next stimulus, the participant’s index finger was returned to the thermal plate to maintain Tsk at 33 °C. This also served as a perceptual refractory period with a minimum duration of 20 s, during which time the next stimulus was prepared before cyclically repeating the protocol. All stimuli were repeated and presented in a balanced order. Each participant attended separate experimental sessions for the different temperature conditions. Overall, detection threshold tests are slightly more timeconsuming than magnitude estimation tests as they require a large number of repetitions, both within and between participants, to provide a high resolution and low noise output. However, they have been shown to be a reliable methodology that participants readily complete [23]. 3.2.2 Example 2: Wetness Magnitude Estimation—External Moisture and Passive Touch

The method aimed to establish differences in regional wetness sensitivity across the body, which was achieved using a singleblind psychophysical approach [33]. A series of sites were mapped across the left and medial areas of the body. Participants underwent a brief familiarization detailing the study protocols and a calibration using the volar (palm) side of the forearm. The stimuli consisted of a square thermal probe with a contact surface of 25 cm2 with an attached cotton square, to which 0.8 mL of water was applied. The temperature of the probe was regulated to be either Tsk - 5 °C, Tsk, or Tsk + 5 °C, as established with an infrared thermometer for a specific site, with each temperature condition presented in a separate experimental session. The stimuli were contacted sequentially

192

Charlotte Merrick et al.

onto the mapped sites, during which time participants reported the local wetness sensation on a digital visual analog scale, from dry to completely wet. At the start of an experimental session, participants arrived at the laboratory and changed into specified test clothing. Their body mass was recorded using a digital scale and their height using a stadiometer. A washable marker pen was then used to indicate the stimulation sites across the left and medial body. Following preparation, participants sat at rest for 15 min to adjust to environmental conditions, during which time they were familiarized with the experimental procedures. Calibration procedures consisted of six stimuli varying in both temperature and wetness (dry, Tsk - 5 °C; wet, Tsk - 5 °C; dry, Tsk; wet, Tsk; dry, Tsk + 5 °C; wet, Tsk + 5 °C), which were applied to the left volar forearm in a randomized order. Participants were instructed to associate each stimulus with the anchor points on the visual analog scale. Participants were instructed to stand throughout the experiments. The local Tsk at the first test site was recorded using an infrared thermometer. A square thermal probe (Fig. 3) was set at the corresponding test temperature (either Tsk - 5 °C, neutral Tsk, or Tsk + 5 °C). A 100% cotton fabric swatch was then applied to the thermal probe and wetted with a pipette using 0.8 mL of water to ensure full saturation. Participants were notified just before stimulus application, at which point the stimulus was applied statically on the skin site for 5 s. During this time the participant completed their perceptual rating, and the stimulus was removed before proceeding to the next skin region. The order of testing was counterbalanced between thermal conditions, and the order of body regions was counterbalanced between and within participants. Overall, the use of visual analog scales in magnitude estimation studies is simple and efficient. The basis of the psychophysical rating scales is easy for participants to understand, perhaps due to many similar real-life scenarios in which they are used, and is a rapid way to collect large data sets. This rapidity is mostly due to the use of digital rating scales, as hard copies require manual measurements that are time-consuming and can easily lead to errors. However, consistent results are highly reliant on the correct explanation and calibration of rating scales, such as giving examples for anchor points, stating if there is a midpoint to the scale, or clarifying whether it is linear. 3.2.3 Example 3: Wetness Magnitude Estimation—Self-Produced Moisture and the Interaction Between Wetness and Tactile Cues

The study aimed to establish if the perception of skin wetness can be significantly altered by manipulating tactile cues, independently from the level of physical skin wetness [34]. This was achieved using clothing of different tightness during an incremental walking protocol which caused participants to consistently generate physical skin wetness in the form of sweat. Prior to beginning, participants were familiarized with the experimental protocols and allowed to

Wetness Perception in Humans

193

acclimate to the surroundings. The walking protocol was conducted under two conditions, one using a tight-fitting shirt and the other using a loose-fitting shirt. In each condition, wetness perception was recorded using verbal responses to a 7-point Likert scale every 4 min (-3 dripping wet; -2 wet; -1 slightly wet; 0 neutral; +1 slightly dry; +2 dry; +3 very dry). On experimental days, participants were requested to arrive at the laboratory 30 min prior to their scheduled start time to allow for the preparation of procedures and bodily acclimation. Before beginning, participants were asked to void their bladder and their semi-nude body mass was recorded on a digital scale. Participants then wore the first layer of either tight- or loose-fitting clothing and were asked to rate their wetness in order to establish a baseline of sensation. Participants then put on the second layer of clothing, consisting of an impermeable jacket and trousers to prevent sweat dissipation. Participants were moved to the treadmill where they began the 45 min incremental walking protocol (5 km.h-1; gradient: +2 to +16%). During the protocol, participants were instructed to rate their wetness sensations every 5 min. As soon as a slightly wet response was given, participants were requested to detail whether this included chest, back, arms or thighs, and which of these was the wettest. Subsequent to finishing the walking protocol, participants removed their clothing, at which point their semi-nude body mass was again recorded. With a minimum of 48 h separating the trials, the protocol was repeated, this time using whichever of the tightor loose-fitting clothing options that was not previously worn. Like visual analog scales, the use of Likert scales to collect perceptual data is simple and efficient. It is easy for participants to comprehend and allows them to rate their perceptions while undergoing other continuous uninterrupted tasks, for example during exercise. An issue which can occur in such scenarios is the size and position of the scale, as it needs to be clearly legible for participants. Again, consistent results rely greatly on the correct explanation and calibration of rating scales. 3.3

Analysis

3.3.1 Example 1: Wetness Detection— External Moisture and Active Touch

In this study, the independent variables were temperature and water volume. Wetness perception was the dependent variable. Coding dry and wet responses as 0 and 1 respectively allowed a perceptual response ratio to be calculated for any given volume. For example, 5 dry responses and 7 wet responses would generate a value of 0.42 (5/12). Perceptual ratios and corresponding volumes were plotted against each other and fitted with a sigmoidal curve. The point at which the response exceeded 0.5 (50%) was considered the threshold for wetness detection. The threshold values were established for every participant by reading the volume at the 0.5 intersection, and means calculated for each participant and each temperature condition (Fig. 4).

194

Charlotte Merrick et al.

Fig. 4 An example of the sigmoidal curves fitted for each respective temperature within a single participant. Each point represents the ratio of responses from repeated stimuli with 95% confidence intervals 3.3.2 Example 2: Wetness Magnitude Estimation—External Moisture and Passive Touch

The dependent variable in the study was wetness perception, while the independent variables were stimulus location and temperature. Changes in perceptual responses can be mapped to show regional variations in wetness perception and how these perceptions are correlated with stimulus temperature (Fig. 5). This gives insights into regional sensitivity at different temperatures as well as intraand inter-individual variability. Similarly, if a sufficient number of participants are recruited, the information can reflect differences in perception according to sex, age, and other demographics, or under different conditions, such as pre- and post-exercise.

3.3.3 Example 3: Wetness Magnitude Estimation—Self-Produced Moisture and Differences Between Loose and Tight Clothing

In this study, the independent variables were clothing condition (loose- and tight-fitting clothing) and time (10 levels, in 5 min intervals). The dependent variable was wetness sensation. The mean wetness sensation at each 5 min time interval was established across participants in both the loose- and tight-fitting clothing conditions. As wetness sensation was recorded on a Likert scale and constitutes ordinal data, a non-parametric method was used to assess if there were significant differences between the two conditions as a function of time. A graphic representation is in Fig. 6.

Wetness Perception in Humans

195

Fig. 5 Body maps of wetness perceptions in males (n = 10) and females (n = 10) resulting from the application of the cold wet, neutral wet and warm wet stimuli at rest [34]

Fig. 6 Means ± SD (n = 10) for wetness perception across time throughout the tight- (filled symbols) and loose-fitting (open symbols) trials. Although during both tight-fit and loose-fit trials the level of physical skin wetness did not differ at any time point, the overall perception of skin wetness was significantly reduced during the tight-fit trial as opposed to the loose-fit trial. This main effect significantly interacted with time, 20 min after the exercise protocol was initiated (*P < 0.05) [35]

196

4

Charlotte Merrick et al.

Notes

4.1 Instructing Participants

It is important that participants fully understand the nature of the task and what is required of them, and this should be explained clearly, concisely, and hopefully with some enthusiasm. The use of consistent commands such as “lift,” “touch,” or “dry” will also act as signposting points which help to guide participants through stimuli interactions until they become almost rhythmic. Depending on the methods, it may be useful for specific interactions to be demonstrated or for participants to practice them to improve technique and consistency as part of a familiarization process. However, the sufficient learning of protocols needs to be balanced with limiting potential for learning effects or response bias. The need for clarity also applies to the use of rating scales and calibration protocols, such as giving examples of anchor points, encouraging participants to rate stimuli independently, or simply reassuring them to trust their own judgment.

4.2 Experimental Duration

There should be an established timescale that participants are aware of. This aims to ensure that they complete the tasks and corresponding perceptual recordings effectively, without either lingering or rushing. Additionally, an optimum length of a session needs to be established. The more research data that needs to be collected, the longer an experimental session will take, and may require splitting into smaller sessions. If a single session is too long, participants may lose motivation, become distracted, rush, or otherwise disengage from the task. Further, if there is long-term exposure to wetness, changes in physical skin characteristics such as increased hydration levels or dermal plasticity should also be considered and the study adjusted accordingly.

4.3 Experimental Design

A common aspect is the repetition of stimuli, which needs to be established based on the desired resolution, accuracy, and precision of the resulting data, as well as accounting for time constraints (see Note 4.2). This is often established in a small pilot study and/or taking into account effects from previous similar studies. The order in which stimuli are presented also needs to be decided, both within and between participants, such as presenting stimuli in a counterbalanced as opposed to a random order. Another aspect to consider in the experimental design is the potential of heat transfer, be this between stimuli, participants, or the ambient environment. For example, a solution is prepared at 35 °C and applied to a fabric sample. By the time interaction occurs, heat energy may have been lost to the ambient environment such that the actual temperature is much lower than anticipated. This can be accounted for by assessing the thermal decay across the range of samples and adjusting temperatures accordingly.

Wetness Perception in Humans

197

4.4 BetweenParticipant Variation

Additionally, variation between participants that is not being regulated as part of inclusion criteria needs to be considered. For example, when creating an L-shaped obscuring screen to limit visual cues for participants, both the varying diameter of participant arms, their seated eye level, and hand dominance needs to be accommodated. While this may seem trivial, it may interfere with interactions and subsequent perceptions.

4.5 The Psychophysical Task

In the determination of sensory thresholds, both a two-alternative forced choice method (2AFC) and a yes/no task (Y/N) can be used (see also Chapter 1, this volume). The 2AFC allows participants to choose which of the two stimuli corresponds best to a single descriptor, whereas Y/N involves only a single stimulus to which either a positive or negative response must be assigned [35, 36]. The latter is often modified such that the single stimulus is assigned to one of two opposing descriptors.

4.6 Temperature Manipulation

While approaches to manipulating temperature vary, many are applicable with appropriate justification. For example, when stimuli and the skin are in contact, the stimulus could either have been manipulated in relation to physical skin properties (e.g., Tsk - 10 °C), or the skin itself can be maintained at a specific baseline condition (e.g., 33 °C) to ensure the same transient magnitude of change between participants. Typically, the former is used in passive touch and the latter in active touch studies.

References 1. Vargas NT, Chapman CL, Johnson BD et al (2019) Exercise intensity independently modulates thermal behavior during exercise recovery but not during exercise. J Appl Physiol 126: 1150–1159. https://doi.org/10.1152/ japplphysiol.00992.2018 2. Gray M, Black JM, Baharestani MM et al (2011) Moisture-associated skin damage overview and pathophysiology. J Wound Ostomy Cont Nurs 38:233–241. https://doi.org/10. 1097/WON.0b013e318215f798 3. Filingeri D, Ackerley R (2017) The biology of skin wetness perception and its implications in manual function and for reproducing complex somatosensory signals in neuroprosthetics. J Neurophysiol 117:1761–1775. https://doi. org/10.1152/jn.00883.2016 4. Kim SM, Wang JW (2016) Hygrosensation: feeling wet and cold. Curr Biol 26:R408– R410. https://doi.org/10.1016/j.cub.2016. 04.040 5. Ackerley R, Kavounoudias A (2015) The role of tactile afference in shaping motor behaviour and implications for prosthetic innovation. Neuropsychologia 79:192–205. https://doi.

org/10.1016/j.neuropsychologia.2015. 06.024 6. Ackerley R, Watkins RH (2018) Microneurography as a tool to study the function of individual C-fiber afferents in humans: responses from nociceptors, thermoreceptors, and mechanoreceptors. J Neurophysiol 120:2834– 2846 7. Filingeri D, Havenith G (2017) Peripheral and central determinants of skin wetness sensing in humans. Temperature 2:86–104. https://doi. org/10.1080/23328940.2015.1008878 8. Bentley I (1900) The synthetic experiment. Am J Phys 11:405–425. https://doi.org/10. 1525/tph.2001.23.2.29 9. Sullivan A (1923) The perceptions of liquidity, semi-liquidity and solidity. Am J Psychol 34: 531–541 10. Zigler MJ (1923) An experimental study of the perception of stickiness. Am J Psychol 34:73– 84. https://doi.org/10.4992/jjpsy.11.279 11. Zigler MJ (1923) An experimental study of the perception of clamminess. Am J Psychol 34: 550–561

198

Charlotte Merrick et al.

12. Cobbey LW, Sullivan AH (1922) An experimental study of the perception of oiliness. Am J Psychol 33:121. https://doi.org/10.2307/ 1413756 13. Demartine ML, Cussler EL (1975) Predicting subjective spreadability, viscosity, and stickiness. J Pharm Sci 64:976–982. https://doi. org/10.1002/jps.2600640618 14. Gagge AP, Stolwijk JAJ, Hardy JD (1967) Comfort and thermal sensations and associated physiological responses at various ambient temperatures. Environ Res 1:1–20. https:// doi.org/10.1016/0013-9351(67)90002-3 15. Sweeney MM, Branson DH (1990) Sensorial comfort: part I: a psychophysical method for assessing moisture sensation in clothing. Text Res J 60:371–377. https://doi.org/10.1177/ 004051759006000701 16. Sweeney MM, Branson DH (1990) Sensorial comfort part II: a magnitude estimation approach for assessing moisture sensation. Text Res J. https://doi.org/10.1177/ 004051759006000803 17. Merrick C, Rosati R, Filingeri D (2021) Skin wetness detection thresholds and wetness magnitude estimations of the human index fingerpad and their modulation by moisture temperature. J Neurophysiol 125:1987–1999. https://doi.org/10.1152/jn.00538.2020 18. Filingeri D, Fournet D, Hodder S, Havenith G (2014) Why wet feels wet? A neurophysiological model of human cutaneous wetness sensitivity. J Neurophysiol 112:1457–1469. https://doi.org/10.1152/jn.00120.2014 19. Farage MA, Meyer S, Walter D (2004) Evaluation of modifications of the traditional patch test in assessing the chemical irritation potential of feminine hygiene products. Skin Res Technol 10:73–84. https://doi.org/10. 1111/j.1600-0846.2004.00054.x 20. Humbert P, Ferial F, Agache P, Maibach HI (2017) Agache’s measuring the skin, 2nd edn. Springer International Publishing AG, Cham 21. Filingeri D, Fournet D, Hodder S, Havenith G (2014) Body mapping of cutaneous wetness perception across the human torso during thermo-neutral and warm environmental exposures. J Appl Physiol 117:887–897. https:// doi.org/10.1152/japplphysiol.00535.2014 22. Vargas NT, Chapman CL, Johnson BD et al (2019) Thermal behavior alleviates thermal discomfort during steady-state exercise without affecting whole body heat loss. J Appl Physiol 127:984–994. https://doi.org/10. 1152/japplphysiol.00379.2019 23. Ehrenstein W, Ehrenstein A (1999) Psychophysical methods. In: Windhorst U, Johansson

H (eds) Modern techniques in neuroscience research. Springer, Berlin, pp 128–157 24. Hatzfeld C, Ku¨hner M, So¨llner S et al (2017) Human perception measures for product design and development—a tutorial to measurement methods and analysis. Multimodal Technol Interact 1:1–23. https://doi.org/10. 3390/mti1040028 25. Treutwein B (1995) Adaptive psychophysical procedures. Vis Res 35:2503–2522. https:// doi.org/10.1016/0042-6989(95)00016-X 26. Breedlove SM, Rosenzweig MR, Watson NV (2007) Biological psychology : an introduction to behavioral, cognitive, and clinical neuroscience, 5th edn. Sinauer Associates, Sunderland 27. Peters A (2007) The effects of normal aging on nerve fibers and neuroglia in the central nervous system. In: Riddle D (ed) Brain aging: models, methods, and mechanisms. CRC Press/Taylor & Francis, Boca Raton 28. Stevens JC, Choo KK (1996) Spatial acuity of the body surface over the life span. Somatosens Mot Res 13:153–166 29. Stevens J, Choo K (1998) Temperature sensitivity of the body surface over the life span. Somatosens Mot Res 15:13–28 30. Ortiz A, Grando SA (2012) Smoking and the skin. Int J Dermatol 51:250–262 31. Yadav RL, Sharma D, Yadav PK et al (2016) Somatic neural alterations in non-diabetic obesity: a cross-sectional study. BMC Obes 3:50. https://doi.org/10.1186/s40608-0160131-3 32. Schrieks I (2016) Influence of moderate alcohol consumption on emotional and physical well-being. Wageningen University 33. Valenza A, Bianco A, Filingeri D (2019) Thermosensory mapping of skin wetness sensitivity across the body of young males and females at rest and following maximal incremental running. J Physiol 597:3315–3332. https://doi. org/10.1113/JP277928 34. Filingeri D, Fournet D, Hodder S, Havenith G (2015) Tactile cues significantly modulate the perception of sweat-induced skin wetness independently of the level of physical skin wetness. J Neurophysiol 113:3462. https://doi.org/10. 1152/jn.00141.2015 35. Green DM (1964) General prediction relating yes/no and forced/choice results. J Acoust Soc Am 36:1042–1042. https://doi.org/10. 1121/1.2143339 36. Schulman AJ, Mitchell RR (1966) Operating characteristics from yes/no and forced/choice procedures. J Acoust Soc Am 40:473–477. https://doi.org/10.1121/1.1910098

Chapter 10 Skin-Mediated Interoception: The Perception of Affective Touch and Cutaneous Pain Laura Crucianelli and India Morrison Abstract The last decades have seen an increasing focus on skin-mediated interoceptive modalities, such as tactile pleasantness and pain. For touch, this interest has been partially motivated by the discovery of a specialized group of skin afferents, C-tactile (CT) afferents that are found mainly in the hairy skin of the body and have been proposed as a supporting system for the detection of slow, caress-like touch, often referred to as affective touch. Similarly, there is growing interest in the affective and motivational dimensions of pain. Although painful and pleasant interoceptive sensations are associated predominantly with the skeletomuscular system, vascular system, and inner organs, the neural processing of affective touch and cutaneous pain starts in the skin. These modalities are homeostatically relevant since they provide information about physiological safety or threat. In this chapter, we first offer an overview of neuroanatomical, physiological, and functional evidence supporting the interoceptive nature of certain types of tactile stimuli, namely, tactile pleasantness and pain. We then describe the materials, methods, and procedure of experimental tasks focused on these modalities, which offer a promising avenue for the development of somatosensory methods to measure interoception. Key words Affective touch, Cutaneous pain, Interoception, Body awareness, Visceral signals, Insula

1

Introduction Both touch and pain are crucial for our survival and are accompanied by affective feelings. The view outlined in this chapter is that the affective feelings of pleasure and aversion inspired by touch and pain mediate the relationship between the external environment (i.e., exteroception) and the status of the body (i.e., interoception) [1]. On this view, pain and pleasure are strong drivers of adaptive behaviors and can work in concert to maintain homeostasis, the often-anticipatory adjustment of physiological systems toward stable points within a range of possible states. If, on the one hand, pain represents a disruption in homeostatic balance, on the other hand, pleasure can signal the restoration of such balance [2, 3]. Whether

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_10, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

199

200

Laura Crucianelli and India Morrison

Fig. 1 The interoceptive systems, which include the skin. (Image credit Wikipedia (CC BY-SA 4.0))

they are independent processes or two aspects of the same process is still debated. When it comes to perception of the bodily senses, a traditional classification relies on a distinction between exteroception (perception of signals originating from outside the body), interoception (perception of signals originating from inside the body), and proprioception (perception of signals about the position of our limbs and body parts, see Chapter 3, this volume) [4, 5]. According to classic views, exteroception included the thermal senses of heat and cold, as well as nociception (see Chapters 7 and 9, this volume). However, given the evidence that thermoreception and nociception are important controllers of behavior and homeostatic mechanisms needed to maintain thermoneutrality and survival, these senses have been redefined as interoceptive [2, 6, 7] (Fig. 1). Although both affective touch and cutaneous pain (see Chapters 6 and 7, this volume) are usually the results of stimulation originating outside the body, these modalities also provide crucial information about the internal, physiological status of the body. This functional feature is key to the physiologically centered view of interoception presented in this chapter. In this view, tactile pleasantness and pain have been re-defined as interoceptive modalities [6, 8] since they signal physiological safety or threat for the survival of the organism [2, 9]. As such, these modalities also carry a strong motivational component that prompts the individual to take

An Interoceptive take on Touch

201

Table 1 Main characteristics of the slow thinly myelinated Aδ and unmyelinated C skin receptors Receptor type

Fiber group

Modality

Cutaneous and subcutaneous mechanoreceptors

Touch

C mechanoreceptors

Stroking, affective touch

C

Thermal receptors

Temperature

Cool receptors

Aδ

Skin cooling (35 °C)

Cold nociceptors

Aδ/C

Cold temperature (45 °C)

Nociceptors

Pain

Mechanical

Aδ

Sharp, pricking pain

Thermal mechanical (heat)

Aδ/C

Burning pain

Thermal mechanical (cold)

Aδ/C

Freezing pain

Polymodal

C

Slow, burning pain

Table adapted from Gardener and Johnson [12]

voluntary action to achieve, maintain, or restore homeostatic balance [2, 8], a process that is defined as allostatis [10]. They thus contribute to our ability to use tactile signals to gather information about the external world, with the ultimate purpose of homeostatic regulation and survival (e.g., perception of temperature, pain, pleasure from touch, and itch—Chapter 8, this volume) [2]. Cutaneous pleasure and pain are the result of a stimulation on the skin, and the related response of the peripheral nerves and central neural system. The peripheral receptors in the skin can mainly be classified based on their dimension and conduction (Table 1), namely, myelinated fibers (i.e., Aβ) are usually large and provide a fast response to stimulation. In contrast, small fibers provide a relatively slower response to stimulation, and they can be unmyelinated (i.e., C) or thinly myelinated (i.e., Aδ) fibers. The small, slower fibers are responsible for nociception, thermoception, and some types of touch. More specifically, some forms of pleasant, affective touch may be underpinned by a relatively recently discovered type of C afferent nerve fiber in the skin, C-tactile or CT afferents ([11], Chapters 6 and 15, this volume). This is striking because CTs belong to the same class of unmyelinated afferents associated with interoceptive signaling as in pain, as well as temperature and visceral perception (the slow, unmyelinated CT system, and the thinly myelinated Aδ fibers). Although tactile sensation has long been considered an “exteroceptive” sense, the neurophysiology of CT-mediated

202

Laura Crucianelli and India Morrison

affective touch has more in common with these “interoceptive” senses, both regarding the fiber types and the spinal pathways they follow, suggesting a shared functional link. In terms of pain sensation, the activation of the different skin receptors underpins two different pain components. A first, fast, and acute pinpricking sensation is mediated by Aδ, as well as Aβ fibers, which provide spatial information about the localization of the stimulus on the body, in order for the organism to react promptly. A second, slower response is mediated by C fibers and provides a diffuse burning sensation [13]. Most nociceptors are free nerve endings of primary sensory neurons (Table 1). The information about pain sensation is conveyed by thinly myelinated Aδ and unmyelinated C afferents. These pain signals reach the primary somatosensory cortex via the spinothalamic tract, which is the most prominent ascending nociceptive pathway in the spinal cord, as well as the cingulate, insula, and amygdala, where the affective and motivational components of the stimulation are constructed [14]. Nevertheless, this ascending physiological pathway and neural mechanisms underpinning nociception are not the only contributors to the onset of the sensation of pain. The discovery of the descending pain modulation pathway [15] provided further insight into the mind-body interaction in the experience of pain. The ability of the human brain to modulate the perception of pain challenges the measurement of the subjective pain experience [16]. Paying attention to the subjective experience of pain has changed the way we measure this experience. The person feeling the pain is the only one to have access to the actual experience, and the experimenter’s task is to observe the effect of such unpleasant feelings on actions or behaviors, or let the person rate the experience using a scale or words. To better clarify how affective touch gives arise to the pleasant percept, it is useful to explain first what is hypothesized about the ascending pathway of the tactile and/or affective touch stimulation from the periphery (skin) to the brain (Table 1 and Fig. 2), and then how these signals are thought to be modulated by and integrated with top-down information [17] (see Notes 4.1 and 4.2). The affective dimension of touch can be investigated by means of a low-pressure, slow, caress-like tactile stimulation delivered at velocities between 1 and 10 cm/s [18], see Chapter 6, present volume). Studies conducted applying a neurophysiological method called microneurography (see Chapter 15, present volume), which allows recording of the activity of single peripheral nerves on the skin, showed an activation of CT afferents when touch was presented with the aforementioned characteristics [11]. These fibers are present mainly on the hairy skin of the body [11] and, when activated, individuals report a pleasant percept. Indeed, Lo¨ken and colleagues showed that there is a linear correlation

An Interoceptive take on Touch

203

Fig. 2 Pathways from the skin to the brain

between the activation of the CT fibers and the subjective report of pleasantness [18]. Moreover, CT afferents seem to respond more vigorously to touch stimuli that are close to the typical skin temperature compared to colder or warmer stimuli, and also this activation correlates to subjective pleasantness ratings [19]. Slow, CT optimal touch not only optimally activates the CT system, but it has also been linked to the reward system [20, 21]. After receiving touch at CT-optimal velocities, participants show a preference for such stroking speeds as compared to slower or faster speeds, suggesting a potential activation of the reward system and rewardseeking behavior. Importantly, the pattern of preference for the speeds resembles that of subjective pleasantness, as well as the optimal activation of the CT system, which is an inverted-U shape or negative quadratic curve [18, 20, 22]. Neuroimaging studies showed that CT afferents, via the spinothalamic tract [23], seem to take a specific ascending pathway from the periphery to the posterior insular cortex [24, 25]. This brain area is strongly interconnected with the amygdala, hypothalamus, and orbital frontal cortex, and is understood to support an early convergence of sensory and affective signals about the body that are then represented in the mid and anterior insula, the proposed sites of interoceptive awareness [8, 26]. Both in the case of

204

Laura Crucianelli and India Morrison

tactile pleasantness and pain perception, the posterior-to-anterior insula processing seems to contribute to affective evaluation in terms of salience [27, 28] and/or danger [29–31].

2

Materials

2.1 Affective Touch Task

CT-optimal stimulation is usually delivered with a soft brush which moves at specific velocities and controlled pressure over the hairy skin. A major distinction in studies investigating the perception of affective touch is between the use of manual or handheld stimulation and of a rotatory tactile stimulator (RTS, see Chapter 6, present volume for further details on affective touch studies). Although both methods seem to target the CT system to a similar extent and elicit comparable levels of perceived pleasantness [32], the implication of such methodologies should be discussed. While using an RTS to deliver the touch allows high precision and specificity in terms of velocity of touch and pressure, it can also lack ecological validity. One could argue that manually delivering the touch presents the limitation of lower precision and consistency in the velocity used to deliver the touch, however, manual touch also affords the opportunity to investigate actual interpersonal touch occurring between two individuals; as such, it can be argued that it has higher ecological validity (Table 2). In the context of affective touch delivered manually, another important distinction is the one between brush- and handdelivered touch. Most studies investigating the relationship between velocity of touch and subjective perception of pleasantness

Table 2 Comparison between manual and mechanical stroking Manual stroking

RTS stroking

Consistency

Somewhat high—requires training

Very high

Ecological validity

Very high

Low

Range of velocity

Limited

Extensive

Noise

Very low

Somewhat high

Reproducibility

Somewhat high

Very high

Items used to deliver touch

Brush, other material, hand

Brush or other material

Characteristics

Picture of RTS adapted from McGlone et al. [33]

An Interoceptive take on Touch

205

or intensity have used a soft brush to deliver the touch, both manually and mechanically. This also allows control over the pressure of the touch delivered. When delivering (affective) touch with hands, additional signals, such as the temperature of the hands and sweating, can act as confounds. Thus, when designing experiments with touch delivered by hands, it would be advisable that the experimenter wears a pair of gloves. Another factor to take into account is whether participants should wear a blindfold or not during the experimental procedure. In some studies, participants are asked to wear a blindfold to focus as much as possible only on the tactile stimulation, and to control for any potential influence of vision on touch. However, other studies prefer not to use a blindfold to avoid any anxiety-related influence on tactile perception, and to increase the ecological validity of the study, given the fact that in real social interaction we usually can see both the touch and the toucher. 2.2 Cutaneous Pain Task

Experimental studies involving the perception of pain usually take into consideration two important aspects, the standardized activation of the peripheral nociceptive system and the measurements of the related evoked response (see Chapter 7, this volume). Specifically, it would be helpful to have a way to directly measure the amount of painful input and the related activation of the nervous system [15]. However, obtaining an objective outcome measure of people’s perception of pain rather than self-report measure (i.e., asking people how much pain they are experiencing) is still quite challenging and intrinsically related to the subjective percept [15]. After all, pain is an emotional subjective experience by definition [34]. Different techniques have been used to induce and assess pain in humans, and most of these focus on nociceptive stimuli applied to the skin. The main experimental techniques that target the activation of the Aδ and C afferents [35] are:

2.2.1

Thermal Pain

Thermal pain includes contact heat and cold stimuli (e.g., thermode), no contact heat (e.g., radiant sources, such as CO2 laser stimulation), and cold (e.g., dry ice) as well as cold pressor methods, in which pain is induced by immersion of a limb in very cold water. This type of stimulation yields reproducible responses, but the stimulus location must be changed regularly to reduce peripheral habituation. Thermal pain allows one to test both cold and warm pathways (see Chapter 9, this volume).

2.2.2

Mechanical Pain

Mechanical methods produce a wide range of pain intensities and durations, and they can excite low- or high-threshold mechanoreceptors depending on both the probe surface and the applied force. These pain sensations can be distinguished as either fast responses, evoked by deformation of the skin via von Frey hairs

206

Laura Crucianelli and India Morrison

Fig. 3 The Somedic MSA Thermal Stimulator and the thermode attached to a thermal stimulator during an experimental session on both forearm (hairy skin) and palm (non-hairy skin). The participants are holding the response button on the non-stimulated hand to complete the method of limit procedure

and needles (e.g., pinprick), or slow responses (e.g., tourniquet pain). The latter, by arresting blood flow in an arm, can produce ischemic pain, a severe, continuous, and increasing pain that can generally be tolerated for 20 min [35]. 2.2.3

Electrical Pain

2.2.4 Experimental Design Factors

Electrical pain allows excitation of nerve fibers directly in the epidermis. When it comes to design cutaneous pain tasks, there is the issue of the specificity of the nociceptive stimulus, given the fact that peripheral nerves contain all the types of fibers, Aδ, Aβ, and C. An important limitation of mechanical and electrical peripheral stimulation is that it is difficult to disentangle the activation of these fibers, given the inevitable co-excitation of the tactile Aβ fibers, responsible for the perception of discriminatory touch [36]. Given the focus on the Aδ and C systems in this chapter, here we will describe only thermal pain, and in particular contact heat and cold stimuli delivered by means of a thermode (Fig. 3). Most thermal stimulators usually come with bespoke software which allows manipulation of the temperature of the thermode, and creation of standardized experimental procedures to identify individual detection and threshold levels, such as the methods of limits [37]. However, software like PsychoPy or Matlab can often be used to run more complex tasks using the thermal stimulator. Many modern contact thermodes use the Peltier principle, in which a direct current through a semiconductor substrate results in an increase in temperature on one side and a decrease in temperature on the other. Other contact stimulators use electrical heaters which may be cooled by circulating fluid. Thermal pain can also be achieved by simple immersion in hot or cold water, or by infusion of hot or cold water into muscle [35].

An Interoceptive take on Touch

3

Methods

3.1 Methodological Considerations in Affective Touch Studies 3.1.1

207

Stimulus Properties

Speed/Velocity Affective touch is frequently dynamic and delivered at velocities between 1 and 10 cm/s [18]. The RTS can be programmed to deliver the stimulation at specific velocities. When the touch is delivered manually, the experimenter should be trained to deliver the touch at different velocities, with the help of a metronome or other sound cues. The inclusion of a control condition is essential to specifically target CT activation in the context of affective touch. It is good practice to include velocities that are outside the optimal window of activation of the CT system—either slower than 1 cm/s or faster than 10 cm/s. These speeds would yield lower levels of tactile pleasantness as well as a reduction in the CT-firing pattern (according to the inverted-U shape pattern of responses [18]). Pressure It is crucial that the touch is light, and that the pressure is constant across velocities in order to optimally activate the CT system (0.2–0.4 N [18, 38]). Similar to the control of velocity, the RTS can be calibrated to deliver brush strokes at a constant pressure. In the case of manual stroking, it is crucial to mark the stroking area with a washable marker to control the spread of the brush laterally, as a way to control the pressure of the brush stroking (i.e., an increase or decrease of pressure would lead to a change in the spread of the brush) [39–41]. Recently, deep pressure stimulation using an oscillating compression sleeve has been identified as a form of pleasant touch with comparable subjective and neural effects as stroking [42]. Temperature The CT system is maximally activated at neutral temperatures, typical of the skin (e.g., 32 °C) [19]. Thus, it is important to monitor the temperature of the material used to deliver the touch (see Note 4.3). Stroking Area The CT system is known to be easily fatigued, which means that it responds more vigorously at the beginning of stimulation, and it gradually reduces its response the longer it is stimulated [43]. CT activity is highly dependent on previous stimulation, showing a decrease in response to several identical stimuli [11]. It is important to identify at least two stroking areas that can then be alternated allowing the fibers to recover from stimulation [41]. Furthermore, to target the interoceptive specificity of the CT system, several studies include a glabrous skin site such as the palm, for which evidence of CT innervation is less robust [11] (although a recent study has detected CT-like units at or near the boundary between hairy and glabrous skin [44]).

208 3.1.2

Laura Crucianelli and India Morrison Rating Scales

3.1.3 Participant Inclusion and Exclusion Criteria

A crucial issue to consider in affective touch research is the outcome measure of the experiment, in other words which aspect(s) of the tactile experience we are interested in measuring. The experimental factors mentioned so far are under the control of the experimenter; however, by definition, the perception of affective touch refers to a personal and subjective experience. Most studies to date have focused on the experience of pleasantness associated with the activation of the CT system, since there is a linear correlation between the peripheral activation of such system and the subjective pleasantness percept reported by the participants [18]. A good example of a pleasantness rating scale usually adopted in affective touch research is a VAS ranging from 0, “not at all pleasant” (or unpleasant, depending on the experimental design and hypothesis) to 100, “extremely pleasant.” If the participants are blindfolded as part of the experimental procedure, then they can report the answer verbally after each tactile trial by saying a number between 0 and 100. A few studies have focused on other aspects of the tactile experience such as intensity, softness/hardness, smoothness/ roughness, or stickiness, as well as different aspects related to the material used to deliver the touch [45, 46, 47]. A touch perception task (TPT) was developed to address this issue and establish a standardized tactile lexicon [46]. Participants were asked to judge active and passive tactile stimulation according to 26 sensory attributes (e.g., roughness, slipperiness, firmness) and 14 emotional attributes (e.g., pleasurability, comfort, arousal). Interestingly, performance on the TPT showed different results in hairy compared to non-hairy skin, with touch on the hairy skin being better described by emotional attributes, and showing preferences for other versus self-touch as compared to non-hairy skin [48]. Along this line, a few studies have focused on trying to understand the way in which we use touch to communicate emotions in interpersonal encounters, thus focusing on more complex and varied aspects related to the tactile perception than pure pleasantness, such as intentions [49–51]. Recent metacognitive accounts of subjective perception have also focused on the confidence ratings when participants provided a subjective estimate of perception (i.e., pleasantness). Thus, depending on the aim of the study, it could be relevant to add an item such as “how confident are you with your answer?” following each pleasantness rating [52] (see Note 4.4). These criteria refer only to healthy participants (for assessment of somatosensation in clinical populations, see Chapter 24, this volume).

An Interoceptive take on Touch

209

Age It has been suggested that age can affect positive and negative emotions associated with the perception of touch [53] as well as interoception more generally [54]. Thus, it might be relevant to consider a group of participants within the same or at least similar age range (e.g., age ranges 21–30 and 31–40 years-old show the most uniform performance in tactile and interoceptive tasks, although it is not clear if this is because this age range is overrepresented in the experimental tasks). Gender When it comes to the gender of participants, a recent meta-analysis showed that females, as compared to males, generally perceive touch, both affective and discriminative, as more pleasant [55]. The gender of the experimenter could affect the perception of touch too. It has been shown that the recipient’s belief about the person giving the touch modulated the perceived pleasantness of slow touch [56, 57]. Male participants rated slow, CT-optimal touch as more pleasant when they believed that they were touched by a female experimenter, but unpleasant when they believed that they were touched by a male experimenter. In all the conditions the experimenter was, in fact, always the same [57]. A recent study also showed interactions between reported tactile pleasantness and the perceived attractiveness of the faces participants were asked to look at while receiving touch [58]. In particular, attractive opposite-sex face stimuli enhanced subjective pleasantness and heart rate variability in heterosexual participants, with greater subjective effects in males. Thus, social top-down factors (such as the gender of the toucher and perceived attractiveness) can modulate the meaning and desirability of the touch, and also the positive experience of it, in terms of ratings (see also [59] for a recent study showing preferences for body size and sex of the person delivering virtual touch based on sexual preferences). Health The general criteria for healthy participants apply, such as not having any history of dermatological conditions nor of psychiatric, neurological, or neurodevelopmental conditions, as these have been found to affect the perception of (affective) touch (e.g., anorexia nervosa [40, 60], stroke [61], autism [62, 63]). To avoid any health-related issue associated with being under- or overweight, it is suggested to test participants having healthy body mass index (BMI, range 18.5–24.9) by asking participants to provide their weight and height (see Note 4.5). Skin There should be no scars (which may also suggest a potential history of self-harm), tattoos, or medical conditions that result in skin condition (e.g., psoriasis) on the stimulated area, and no other sensory or skin disorders (e.g., eczema). These factors are important because some participants might feel uncomfortable with being touched in parts of the skin where there are marks, damage, or excoriations.

210

Laura Crucianelli and India Morrison

3.1.4 Experimental Design

Experimental design in affective touch research usually involves the repetition of tactile stimulation delivered at CT-optimal velocities in hairy skin (see Chapter 6, present volume). The inclusion of control conditions such as non-CT-optimal velocities (i.e., slower than 1 cm/s and/or faster than 10 cm/s) [18] and skin sites where CT activation has only partially been recorded to date (i.e., glabrous skin, such as the palm of the hand [11, 44]) is crucial, since it enhances the validity of the study and allows control for the involvement of the CT system in tactile pleasantness. Repeating this range of optimal and non-optimal velocities in a randomized order is important because it counteracts habituation factors and allows for the gradual reduction in optimal firing observed in the activity of the CT system, which is highly dependent on previous stimulation, showing decreased responses as several identical stimuli are presented [11]. The number of repetitions for each trial depends very much on the type of analysis that will be performed on the data. Generally, it is good practice to find a balance between having enough repetitions to be able to detect the effects of interest, but not too many to create habituation or a plateau in response. Another experimental choice to make in affective touch research is between a blocked design where different skin locations are tested in separate blocks (e.g., all the trials on the palm are completed before starting all the trials on the forearm), and a fully randomized design (e.g., a touch trial on the palm can be followed by a touch trial on the forearm). Experiments that manipulate the velocity of touch usually keep fixed either the number of strokes or the duration of the trial across velocities. To control the velocity of touch over the same traveled distance, we can manipulate either the number of strokes (and have the same duration of each trial) or the duration of one stroke (and have different duration of each trial). Nevertheless, it has been shown that these two aspects of experimental design do not yield differences in the results, suggesting that stimulation speed is the most important parameter when it comes both to perceived pleasantness and cerebral activation [25].

3.1.5 Behavioral Data Analysis

Given the repetition of multiple trials per velocity, many studies proceed with averaging the response to each velocity to obtain an average (pleasantness) rating score for each velocity, also to account for any observed difference in ratings due to habituation or fluctuation in attention during the experimental procedure. In the context of interoceptive studies, the focus should be on the ability to distinguish the pleasantness obtained from CT-optimal and CT non-optimal stimulation. Thus, we usually calculate an average of pleasantness for CT-optimal velocities (any velocity between 1 and 10 cm/s) and CT non-optimal velocities (any velocity < 1 cm/ s and > 10 cm/s).

An Interoceptive take on Touch

211

We can then calculate an affective touch sensitivity score, by subtracting the average pleasantness scores obtained from nonCT-optimal speeds from the average pleasantness scores obtained from CT-optimal speeds [41, 61]. As such, we obtain a sensitivity score which resembles the interoceptive awareness that is usually measured with classic heartbeat counting task methods [64], and that is meaningful in quantifying how accurate people are in detecting interoceptive pleasantness. Furthermore, this procedure allows you to identify participants who show a reverse pattern of pleasantness by focusing on the negative affective touch sensitivity scores, that is, participants who perceive non-CT-optimal velocities as more pleasant than CT-optimal velocities. It is then up to the discretion of the experimenter to decide whether this should be a criterion for the exclusion of participants or not (see [41] for a similar approach). A slightly different approach calculates the affective touch awareness score, which takes into account also the total pleasantness reported [65]. Alternatively, more recent approaches have focused on modeling individual pleasantness responses to different velocities to highlight the individual differences in the pattern of responses as compared to the group-level approach [66]. A continuous modelfitting approach has the benefit to be more powerful in detecting the pattern of response based on the different velocities [67]. 3.1.6 Neuroimaging Methods

Several neuroimaging studies provide support to the idea that CT-targeted stimulation activates the contralateral posterior insular cortex (reviewed in [68, 69], see also Chapters 17 and 18, this volume), placing the CT system within an interoceptive network. Important insight came from functional magnetic resonance imaging (fMRI) studies in two unique patients lacking Aβ afferents due to neuropathy, which left the CT system intact [24, 70]. These studies showed a vigorous activation of the insular cortex in response to slow, caress-like touch, but no activation of the primary and secondary somatosensory cortices, suggesting independent central processing of CT and Aβ signals [24, 70]. Along the same line, Morrison and colleagues have shown a reduced activation of the insular cortex in response to CT-optimal touch in patients with congenital C-fiber denervation, suggesting a linear relationship between the density of CT afferents in the skin and insular activation in response to affective touch [71]. A recent lesion study including 59 patients with right-hemisphere stroke involving the insular cortex, showed a reduced behavioral (pleasantness) response to CT-optimal touch, suggesting that the insula is necessary for affective processing of touch [61]. At the level of the spinothalamic tract, it has been suggested that the interoceptive CT signal is inextricably bound to the nociceptive input [72], in line with the idea of a distinct coding channel

212

Laura Crucianelli and India Morrison

projecting primarily to emotional or interoceptive rather than somatosensory regions [6, 73]. However, more recent evidence showed a dissociation between the affective touch and pain pathways. Marshall and colleagues investigated the effect of disruption of the spinothalamic tract by means of antero-lateral cordotomy in the perception of skin-mediated interoceptive signals, such as pain, temperature, itch, and affective touch [74]. The anterolateral cordotomy led to a disruption in the perception of temperature, itch, and pain; however, contrary to their predictions, ratings of CT-optimal touch pleasantness were not affected. These results suggest that more work is needed to characterize the functional neuroanatomy of affective touch. Recent advances in spinal cord neuroimaging techniques [75] might be able to provide new insight in this direction. Finally, additional insight into the functional and central processing of affective touch can be provided by magnetoencephalography (MEG, see also Chapter 19, this volume). Hagberg and colleagues have used MEG to investigate the temporal patterns of brain activation associated with slow, caress-like touch and fast touch delivered on hairy skin [76]. Affective touch rapidly activated somatosensory, motor, and cingulate regions, suggesting that the very first response to slow touch is driven by fast-conducting mechanoreceptive afferents. A slower, temporally separate activation in the posterior insular cortex was also observed, and this may modulate the emotional processing of gentle touch on hairy skin [76]. Thus, studies conducted with MEG provide important information on the temporal and spatial brain activity associated with affective touch. 3.2 Methodological Considerations in Cutaneous Pain Studies 3.2.1

Stimulus Properties

Temperature One aspect common to all the studies investigating cutaneous pain is that an external stimulus must be applied to elicit the experience of pain [35]. Firstly, in pain studies, it is good practice to stress that the participant is in control of the painful sensation (i.e., temperature) and that the machine has been calibrated not to cause any injury. To this end, it can be useful to demonstrate the pain task on the experimenters or a confederate before applying the stimulus to the participant. To choose the range of temperature to use in a thermal pain study, guidelines on the activation of thermal and nociceptive pathways should be applied (cold to hot perception elicited by temperature below 15 °C or above 45 °C [77, 78]; Note 4.3). Pain Thresholds The choice of stimuli will also depend on the paradigm that the experiment is using (e.g., a forced-choice task [78]). Usually, the first step in most pain studies is to identify two important single-point measures at the subjective level: pain detection, defined as the minimum amount of stimulation that reliably evokes a report of pain, and pain threshold, that is the maximum

An Interoceptive take on Touch

213

amount of stimulation evoking pain that can be tolerated. These responses are usually expressed in physical units of stimulus intensity or time; however, they are poor psychophysical measures (see Chapter 7, this volume). Methods of Limits One of the most common methods to quantify pain thresholds is the method of limits, where participants are presented with trials involving ascending and descending stimulus intensities. For example, in the well-established Martsock methods of limits [36, 79], the experimenter holds the thermode on the area of interest (e.g., left forearm or palm) without applying any additional pressure. Participants are asked to hold a response button in the non-stimulated hand, and to press it as soon as they perceive the thermal stimulation becoming uncomfortable or painful [79] (Fig. 3). When providing the experimental instruction, the experimenter should clarify whether the task is to press the button as soon as the sensation of discomfort or pain is starting (i.e., detection) or when the pain is unbearable (i.e., threshold). The baseline starting temperature is neutral (32 °C), the maximum probe temperature is usually set to 50 °C, and the minimum to 10 °C, for safety reasons (in the case of the SOMEDIC and MEDOC thermal stimulators, but this depends on the machine used). As soon as the button is pressed, the temperature automatically changes in the opposite direction and returns to the baseline temperature of 32 °C, where it stays for 5 s before moving to the next trial. The temperature changes at a rate of 2 °C/s, and returns to baseline at a speed of 4 ° C/s. This method is widely used to detect neuropathy in clinical settings, and it includes a total of five warm and five cold trials, presented in two blocks (warm and cold blocks). 3.2.2

Rating Scales

As in the case of affective touch, the measurement of pain is essential in experimental and clinical settings, yet it presents challenges related to measuring a subjective and personal experience. The choice of the appropriate scale, as well as the most optimal labels, to capture the perception of pain have been the topics of various studies (e.g., [34]). Usually, after each nociceptive stimulus, participants rate the intensity of the stimulus on an 11-point VAS ranging from 0 (no pain) to 10 (worst pain ever). The scale can be presented on a computer screen and participants can silently enter their ratings using a numeric keypad or by selecting a point on the visual straight line. A mean pain rating is then calculated for each condition by averaging the pain ratings for the experimental trials in that condition. A disadvantage of pain rating scales is that they capture only a unidimensional experience of pain, whereas pain cannot be reduced to a single sensation. The development of rating scales and questionnaires that measure pain affect or unpleasantness provided a more accurate description of the experience of pain [80].

214

Laura Crucianelli and India Morrison

In terms of participants’ subjective self-report, the most valid measure of the experience of pain is the McGill Pain Questionnaire (MPQ [79, 81]), which was developed to capture the multidimensional experience of pain and has been demonstrated to be a reliable measurement tool, extensively used in clinical settings. The MPQ comprises four main subscales, addressing the sensory (items 1–10), affective (items 11–15), evaluative (item 16), and miscellaneous (items 17–20) experience of pain. A composite score is obtained, and it provides a pain rating index (PRI). The MPQ is to date the most valid tool to obtain a qualitative and quantitative measure of the subjective experience of pain. Behavioral and physiological methods have also been developed to substitute self-report measurements; however, they fail to capture the subjective experience in the same way that self-reports do [80] (see Note 4.4). 3.2.3 Participant Inclusion and Exclusion Criteria

These criteria refer only to healthy participants. Age Age seems to influence pain perception, as shown by studies pointing to an increased pain sensitivity in clinical conditions and in different experimental settings [82, 83]. For example, as people age, they show increased sensitivity to painful heat but no effect on heat thresholds. In contrast, pressure pain thresholds decrease with age [84]. Gender Much research points to the effect of gender on heat and cold pain, as well as in pressure, chemical, and electrically induced pain [84]. Generally, women seem to be more sensitive to experimental and clinical pain than men (e.g., in the sense of lower pain threshold, higher pain detection and intensity), but there are various exceptions that should be taken into account when selecting the experimental sample (review in [85]). Ethnicity and Race Ethnic differences in pain perception seem to play an important role. In particular, increased pain sensitivity (e.g., in the sense of lower pain threshold, higher pain detection and intensity) has been reported in minority groups [35, 86]. Health and Skin The same criteria as affective touch task apply (see Subheading 3.1.3).

3.2.4 Experimental Design

Experimental design in cutaneous pain research involving contact thermal pain usually involves the repetition of moderate to high nociceptive (warm or cold) stimuli applied in hairy or non-hairy skin (see also Chapters 7 and 9, present volume). In the case of discrete stimuli, the inclusion of control conditions such as thermal stimuli in the innocuous range (i.e., stimuli between 15 and 45 °C

An Interoceptive take on Touch

215

[77]) and skin sites where CT activation has only partially been recorded to date (i.e., glabrous skin, such as the palm of the hand [11, 44]) can be very important to understand the involvement of the CT system in tactile pain perception. Furthermore, pain perception can vary quite consistently between skin sites [87] so it is advisable to carefully identify the stimulation sites and keep them consistent across participants and within the experimental session with the same participant. In the case of continuous stimulation, most studies use the standardized method of limits (see Subheading 3.2.1). Next, there should be a repetition of this range of more or less painful and non-painful thermal stimuli in a randomized order, which controls for habituation factors. Some studies also include sham or distractor trials to counteract top-down factors influencing the perception of pain [88]. In terms of number of repetitions and blocked or randomized design, the same principles as affective touch research apply (see Subheading 3.1.4). To capture the subjective experience of pain perception, participants are asked to report when the stimulus starts to be painful (detection) or when the stimulus becomes intolerable (threshold). Participants can provide this information verbally, by using a VAS, by pressing a button, or by withdrawing the hand if the setup allows it. In addition, a measure of confidence in pain perception can be collected to investigate whether metacognitive processes contribute to how distinctive and salient the pain experience is, in line with recent metacognitive accounts [89]. Finally, another issue to consider is whether participants should be blindfolded during the nociceptive stimulation, and whether this could affect the subjective pain perception. This debate is motivated by studies reporting visually induced analgesia when participants can look at the stimulated hand as compared to when they are instructed to look at another object [90]. 3.2.5 Behavioral Data Analysis

An important issue is also to identify the outcome measure of interest, which is usually the subjective pain percept. Thus, the mean pain rating of experimental trials for a stimulus delivered at the same temperature and in the same location is usually calculated and used in subsequent analysis. In the method of limits, the pre-set program is usually able to register the temperature at which the participants press the button to signal pain detection. An average is usually obtained for all the values of the different trials. In pain research in the context of interoception, it might be interesting to focus also on the consistency of the pain perception across trials, by focusing, for example, on the standard deviations (see Note 4.4). Participants who show higher consistency (e.g., for ratio-scale data, using the coefficient of variation, SD/mean) in reporting the same amount of pain in

216

Laura Crucianelli and India Morrison

response to the same external stimuli can be described as having higher interoceptive abilities (see [91] for a similar approach). 3.2.6 Neuroimaging Methods

The main neuroimaging technique used with pain procedure is electroencephalography (EEG), where the usual guidelines apply, and they will not be covered here (see Chapters 7 and 19, this volume). When it comes to contact thermal pain, the most validated methods to objectively assess nociceptive pathway function in humans are contact heat-evoked potentials (cHEP) and contact cold-evoked potentials (cCEP) [36]. These signals can be analyzed by focusing on two main parameters of the evoked potentials: amplitude and latency. To generate a cortical response that can be separated from the background noise in the EEG, the stimulus has to be brief (on the order of ms) and intense, so that the elicited nerve response consists of a highly synchronized afferent signal. Heat or cold stimuli can be applied to the skin with a thermode (see Subheading 2.2.1), and cHEP and cCEP responses can be correlated with fiber activation, as deducted from conduction velocity estimations [92–94]. As compared to laser-evoked potentials, it has been suggested that contact heat might stimulate different C-fiber sub-populations, based on conduction velocities. Functional near-infrared spectroscopy (fNIRS) has recently been suggested for monitoring cortical hemodynamic responses to experimental and clinical acute pain [94]. Functional magnetic resonance imaging (fMRI) allows one to measure differences in brain activation in response to the individual experience of pain [95], by providing an indirect measure of neural activity associated with nociception (see also Chapter 18, this volume). These studies showed that insular cortex and cingulate gyrus are the most active areas during pain evaluation. These areas are closely related to the limbic system, suggesting that such activity is related to the processing of emotional states associated with pain. Indeed, patients with a lesion in the insula present asymbolia for pain, that is they fail to show an emotionally appropriate response to the experience of pain [13]. The insula has long been related to the experience of pain, together with primary and secondary somatosensory cortices, anterior cingulate cortex, prefrontal cortices, and thalamus (the so-called “pain matrix” [96]). Similarly, spinal cord imaging could provide a promising tool to investigate the skin-to-brain pain pathways in vivo (e.g., [97]). Finally, it is possible to gain insight into pain perception by means of brain stimulation (see also Chapter 20, this volume). The stimulation of specific areas of the brain can induce pain sensations or analgesia. Specifically, it has been suggested that electric stimulation of the insula and of the thalamus can uniquely evoke the experience of pain [98]. Injury to the spinothalamic tract and its thalamic targets causes a severe form of pain termed central pain, perceived as originating from diverse regions of the body. Similarly,

An Interoceptive take on Touch

217

in chronic pain conditions, there is a fundamental change in thalamic and cortical circuitry [13]. However, more recent studies suggest that the activation of such a network is not specific for the experience of pain [99], but rather to the experience of the salience of any sensory input, including non-painful stimuli [100, 101]. The stimulation-produced analgesia obtained following stimulation of the periaqueductal gray region uniquely reduces the experience of pain while the perception of touch, pressure, and temperature remain intact. 3.3

4

Conclusion

This chapter highlights the interoceptive aspects of somatosensation. In particular, we focused on the perception of affective (pleasant) touch and cutaneous pain, two sub-modalities of interoception of great homeostatic importance, signaling safety or threat to the organism [8, 9]. We reviewed physiological, functional, and neuroanatomical evidence in support of such interoceptive take on touch. We provided methodological and experimental guidelines, as well as suggestions and tips on how to replicate the described procedures. We believe that the scientific study of tactile pleasantness and pain represents an important window into interoception and body awareness. In particular, a scientific approach to the study of the interoceptive facets of somatosensation can have important translational potential for developing diagnostic tools and novel treatments in clinical conditions characterized by an over- or under-activation of tactile afferent systems, such as chronic pain, autism spectrum disorders, and anorexia nervosa (see Chapter 11, this volume, for the assessment of somatosensation in atypical populations).

Notes In addition to the experimental guidelines reported above, here we provide complementary suggestions and potential issues to consider when conducting affective touch and cutaneous pain studies.

4.1 Top-Down Factors and SelfReport Measures

It should always be taken into account that the pleasant and painful experience associated with touch is not only related to the physical characteristics of tactile stimuli and peripheral activation, but also to top-down mechanisms that modulate and give rise to the actual subjective experience [17]. Some of these top-down, modulatory factors are out of the experimenter’s control (e.g., participants’ mood or trust in the experimenter), while others are easily controllable and measurable during the experimental design and procedure (e.g., previous experiences of tactile pleasantness and pain). It might be of interest to include questionnaires and self-report measures informative about the way in which participants experience tactile stimulation and social contact in general. For example,

218

Laura Crucianelli and India Morrison

some affective touch studies used the Social Touch Questionnaire [102], which provides information on whether participants like or dislike being touched generally. Some studies also include a measure of depression and anxiety (e.g., Depression Anxiety Stress Scale [103]; State-Trait Anxiety Inventory [104]), to control for the presence of anhedonia, which is related to depressive symptomatology, and anxiety symptoms, which might contribute to making the tactile experience uncomfortable. 4.2 The Role of Attachment Style

Studies have shown that the perception of pain can be modulated by providing affective touch as well as by participants’ patterns of relating (i.e., attachment style [88]) and the presence of the participant’s partner [105, 106]. For example, the Responses and Attitudes to Support during Pain questionnaire, RASP [107], allows you to have a measure of individual differences in receiving support while perceiving pain. Similarly, the perception of affective touch seems to be modulated by attachment style [108, 109]. The most common attachment-style questionnaires include the Experience in Close Relationships Revised (ECR-R [110]) and the Adult Attachment Scale (AAS [111]). Thus, one could consider taking these factors into account when running affective touch studies and avoiding any additional tactile input when investigating pain perception (e.g., avoid using an experimental setup that stimulates a large area of the skin to hold the limb) as well as control the number of experimenters in the room and keep it constant.

4.3 Monitoring Skin Temperature

We stress the importance of monitoring the skin temperature in both affective touch and cutaneous pain studies [80, 112]. In the case of affective touch, the CT afferent system is temperaturesensitive, showing an optimal activation at neutral temperature, typical of human skin (32 °C) [19]. In the case of cutaneous pain, this issue is particularly relevant in thermal pain procedures [113, 114]. However, monitoring skin temperature is a challenge, because current methods and techniques traditionally used to monitor skin temperature are sub-optimal. For example, most studies have used an infrared thermometer, the reliability of which closely depends on fluctuations in blood circulation, and it varies consistently across skin sites. Some studies use psychophysical thermosensors (e.g., produced by Biopac), which, however, must be attached to the skin site at all times, providing an additional tactile input on the skin. One of the most optimal methods to date involves thermal cameras, which seem to provide a contactless and reliable option to monitor skin temperature. Asking participants to wash the stimulated part before starting the experiment is also good practice, as it allows removal of any product (such as cream of lotions) that can

An Interoceptive take on Touch

219

influence the accuracy of the tools used to measure skin temperature. Furthermore, it is of great importance to conduct this type of study in testing rooms kept at a constant temperature and humidity, and to give participants the time to get acclimatized in the testing room before starting the experimental procedure. This is to avoid the possibility that any thermoregulatory processes might interfere with and influence the effects of interest, namely, cutaneous pleasantness and pain (see Chapter 9, this volume). 4.4 Three Dimensions of Interoception

In the context of interoceptive studies, it is important to take into account the three dimensions of interoception [115]. These are as follows: (1) interoceptive accuracy, the objective performance at an interoceptive task, in our case the perception of affective touch and cutaneous pain; (2) awareness, the subjective confidence about the objective interoceptive performance; and (3) sensibility, the subjective beliefs about the perception of bodily signals, measured by self-reported questionnaires [116]. Interoception sensibility can be assessed by means of questionnaires, such as the Body Awareness Questionnaire [117]; the Body Perception Questionnaire [118], or the Multidimensional Assessment of Interoceptive Awareness [119].

4.5

When the experiment involves hand or arm stimulation, it is very important to consider the handedness of the participants. It is good practice to keep the stimulated hand or arm constant throughout the study, and usually the non-dominant limb should be tested in passive touch paradigm as it is more specialized in receiving rather than giving touch (as compared to the dominant hand), thus showing enhanced tactile specialization.

Handedness

References 1. Damasio A, Carvalho GB (2013) The nature of feelings: evolutionary and neurobiological origins. Nat Rev Neurosci 14(2):143–152 2. Craig AD (2003) A new view of pain as a homeostatic emotion. Trends Neurosci 26(6):303–307 3. Leknes S, Tracey I (2008) A common neurobiology for pain and pleasure. Nat Rev Neurosci 9(4):314–320 4. Sherrington CS (1906) The integrative action of the nervous system. Yale University Press, New Haven 5. Ceunen E, Vlaeyen JW, Van Diest I (2016) On the origin of interoception. Front Psychol 7:743

6. Craig AD (2002) How do you feel? Interoception: the sense of the physiological condition of the body. Nat Rev Neurosci 3(8): 655–666 7. Crucianelli L, Ehrsson HH (2022) The role of the skin in interoception: a neglected organ? Perspectives on Psychological Science, 1–15 8. Craig AD (2009) How do you feel--now? The anterior insula and human awareness. Nat Rev Neurosci 10(1):59–70 9. Von Mohr M, Fotopoulou A (2018) The cutaneous borders of interoception: active and social inference of pain and pleasure on the skin. In: Tsakiris M, De Preester H (eds) The interoceptive mind: from homeostasis to

220

Laura Crucianelli and India Morrison

awareness. Oxford University Press, Oxford, p 102 10. Morrison I (2016) Keep calm and cuddle on: social touch as a stress buffer. Adapt Hum Behav Physiol 2:344–362 11. Vallbo ÅB, Olausson H, Wessberg J (1999) Unmyelinated afferents constitute a second system coding tactile stimuli of the human hairy skin. J Neurophysiol 81(6):2753–2763 12. Gardener EP, Johnson KO (2013) The somatosensory system: receptors and central pathways. In: Kandel ER, Schwartz JH, Jessell TM, Siegelbaum SA, Hudspeth AJ (eds) Principles of neural science, vol 5. McGraw-Hill, New York, pp 475–496 13. Basbaum AI, Jessell TM (2013) Pain. In: Kandel ER, Schwartz JH, Jessell TM (eds) Principles of neural science. McGraw-Hill, International Edition, New York 14. Basbaum AI, Bautista DM, Scherrer G, Julius D (2009) Cellular and molecular mechanisms of pain. Cell 139(2):267–284 15. Melzack R, Wall PD (1965) Pain mechanisms: a new theory. Science 150(3699):971–979 16. Farrar JT (2010) Advances in clinical research methodology for pain clinical trials. Nat Med 16(11):1284 17. Ellingsen DM, Leknes S, Løseth G, Wessberg J, Olausson H (2016) The neurobiology shaping affective touch: expectation, motivation, and meaning in the multisensory context. Front Psychol 6:1986 18. Lo¨ken LS, Wessberg J, McGlone F, Olausson H (2009) Coding of pleasant touch by unmyelinated afferents in humans. Nat Neurosci 12(5):547 19. Ackerley R, Wasling HB, Liljencrantz J, Olausson H, Johnson RD, Wessberg J (2014) Human C-tactile afferents are tuned to the temperature of a skin-stroking caress. J Neurosci 34(8):2879–2883 20. Perini I, Olausson H, Morrison I (2015) Seeking pleasant touch: neural correlates of behavioral preferences for skin stroking. Front Behav Neurosci 9:8 21. Pawling R, Cannon PR, McGlone FP, Walker SC (2017) C-tactile afferent stimulating touch carries a positive affective value. PLoS One 12(3):e0173457 22. Lo¨ken LS, Evert M, Wessberg J (2011) Pleasantness of touch in human glabrous and hairy skin: order effects on affective ratings. Brain Res 1417:9–15 23. Morrison I (2012) CT afferents. Curr Biol 22(3):R77

24. Olausson H, Lamarre Y, Backlund H, Morin C, Wallin BG, Starck G, Ekholm S, Strigo I, Worsley K, Vallbo AB, Bushnell MC (2002) Unmyelinated tactile afferents signal touch and project to insular cortex. Nat Neurosci 5(9):900–904 25. Morrison I, Bjo¨rnsdotter M, Olausson H (2011) Vicarious responses to social touch in posterior insular cortex are tuned to pleasant caressing speeds. J Neurosci 31(26): 9554–9562 26. Critchley HD, Wiens S, Rotshtein P, ¨ hman A, Dolan RJ (2004) Neural systems O supporting interoceptive awareness. Nat Neurosci 7(2):189–195 27. Menon V, Uddin LQ (2010) Saliency, switching, attention and control: a network model of insula function. Brain Struct Funct 214(5–6):655–667 28. Pessoa L (2014) Attention, motivation, and emotion. In: Nobre K, Nobre AC, Kastner S (eds) The Oxford handbook of attention. Oxford University Press, Oxford, pp 725–752 29. Mouraux A, Diukova A, Lee MC, Wise RG, Iannetti GD (2011) A multisensory investigation of the functional significance of the “pain matrix”. NeuroImage 54(3):2237–2249 30. Morrison I, Perini I, Dunham J (2013) Facets and mechanisms of adaptive pain behavior: predictive regulation and action. Front Hum Neurosci 7:755. https://doi.org/10.3389/ fnhum.2013.00755 31. Perini I, Bergstrand S, Morrison I (2013) Where pain meets action in the human brain. J Neurosci 33(40):15930–15939 32. Triscoli C, Olausson H, Sailer U, Ignell H, Croy I (2013) CT-optimized skin stroking delivered by hand or robot is comparable. Front Behav Neurosci 7:208 33. McGlone F, Vallbo AB, Olausson H, Lo¨ken L, Wessberg J (2007) Discriminative touch and emotional touch. Can J Exp Psychol 61(3):173 34. Raja SN, Carr DB, Cohen M, Finnerup NB, Flor H, Gibson S, Keefe FJ, Mogil JS, Ringkamp M, Sluka KA, Song X-J, Stevens B, Sullivan MD, Tutelman PR, Ushida T, Vader K (2020) The revised International Association for the Study of Pain definition of pain: concepts, challenges, and compromises. Pain 161(9):1976–1982 35. Gracely RH (2013) Studies of pain in human subjects. In: McMahon SB, Koltzenburg M, Tracey I, Turk DC (eds) Wall and Melzack’s textbook of pain, 6th edn. Elsevier, Philadelphia, pp 283–300

An Interoceptive take on Touch 36. Baumg€artner U, Greffrath W, Treede RD (2012) Contact heat and cold, mechanical, electrical and chemical stimuli to elicit small fiber-evoked potentials: merits and limitations for basic science and clinical use. Neurophysiol Clin 42(5):267–280 37. Fruhstorfer H, Lindblom U, Schmidt WC (1976) Method for quantitative estimation of thermal thresholds in patients. J Neurol Neurosurg Psychiatry 39(11):1071–1075 38. Vallbo ÅB, Olausson H, Wessberg J, Norrsell U (1993) A system of unmyelinated afferents for innocuous mechanoreception in the human skin. Brain Res 628:301–304 39. Crucianelli L, Metcalf NK, Fotopoulou AK, Jenkinson PM (2013) Bodily pleasure matters: velocity of touch modulates body ownership during the rubber hand illusion. Front Psychol 4:703 40. Crucianelli L, Cardi V, Treasure J, Jenkinson PM, Fotopoulou A (2016) The perception of affective touch in anorexia nervosa. Psychiatry Res 239:72–78 41. Crucianelli L, Krahe´ C, Jenkinson PM, Fotopoulou AK (2018) Interoceptive ingredients of body ownership: affective touch and cardiac awareness in the rubber hand illusion. Cortex 104:180–192 42. Case LK, Liljencrantz J, McCall MV, Bradson M, Necaise A, Tubbs J, Olausson H, Wang B, Bushnell MC (2021) Pleasant deep pressure: expanding the social touch hypothesis. Neuroscience 464:3–11 43. McGlone F, Olausson H, Boyle JA, JonesGotman M, Dancer C, Guest S, Essick G (2012) Touching and feeling: differences in pleasant touch processing between glabrous and hairy skin in humans. Eur J Neurosci 35(11):1782–1788 44. Watkins RH, Dione M, Ackerley R, Backlund Wasling H, Wessberg J, Lo¨ken LS (2021) Evidence for sparse C-tactile afferent innervation of glabrous human hand skin. J Neurophysiol 125(1):232–237 45. Guest S, Essick G, Dessirier JM, Blot K, Lopetcharat K, McGlone F (2009) Sensory and affective judgments of skin during interand intrapersonal touch. Acta Psychol 130(2): 115–126 46. Essick GK, McGlone F, Dancer C, Fabricant D, Ragin Y, Phillips N, Jones T, Guest S (2010) Quantitative assessment of pleasant touch. Neurosci Biobehav Rev 34(2):192–203 47. Guest S, Dessirier JM, Mehrabyan A, McGlone F, Essick G, Gescheider G, Fontana A, Xiong R, Ackerley R, Blot K

221

(2011) The development and validation of sensory and emotional scales of touch perception. Atten Percept Psychophys 73(2): 531–550 48. Ackerley R, Saar K, McGlone F, Backlund Wasling H (2014) Quantifying the sensory and emotional perception of touch: differences between glabrous and hairy skin. Front Behav Neurosci 8:34 49. Hertenstein MJ, Keltner D, App B, Bulleit BA, Jaskolka AR (2006) Touch communicates distinct emotions. Emotion 6(3):528 50. Hertenstein MJ, Holmes R, McCullough M, Keltner D (2009) The communication of emotion via touch. Emotion 9(4):566 51. Kirsch LP, Krahe´ C, Blom N, Crucianelli L, Moro V, Jenkinson PM, Fotopoulou A (2018) Reading the mind in the touch: neurophysiological specificity in the communication of emotions by touch. Neuropsychologia 116:136–149 52. von Mohr M, Kirsch LP, Loh JK, Fotopoulou A (2019) Affective touch dimensions: from sensitivity to metacognition. bioRxiv:669259 53. Riva F, Tschernegg M, Chiesa PA, Wagner IC, Kronbichler M, Lamm C, Silani G (2018) Age-related differences in the neural correlates of empathy for pleasant and unpleasant touch in a female sample. Neurobiol Aging 65:7–17 54. Khalsa SS, Rudrauf D, Tranel D (2009) Interoceptive awareness declines with age. Psychophysiology 46(6):1130–1136 55. Russo V, Ottaviani C, Spitoni GF (2020) Affective touch: a meta-analysis on sex differences. Neurosci Biobehav Rev 108:445–452 56. Gazzola V, Spezio ML, Etzel JA, Castelli F, Adolphs R, Keysers C (2012) Primary somatosensory cortex discriminates affective significance in social touch. Proc Natl Acad Sci 109(25):E1657–E1666 57. Scheele D, Kendrick KM, Khouri C, Kretzer E, Schl€apfer TE, Stoffel-Wagner B, Gu¨ntu¨rku¨n O, Maier W, Hurlemann R (2014) An oxytocin-induced facilitation of neural and emotional responses to social touch correlates inversely with autism traits. Neuropsychopharmacology 39(9): 2078–2085 58. Novembre G, Etzi R, Morrison I (2020) Hedonic responses to touch are modulated by the perceived attractiveness of the caresser. Neuroscience 464:79 59. Fusaro M, Lisi MP, Tieri G, Aglioti SM (2021) Heterosexual, gay, and lesbian people’s reactivity to virtual caresses on their

222

Laura Crucianelli and India Morrison

embodied avatars’ taboo zones. Sci Rep 11(1):1–12 60. Crucianelli L, Demartini B, Goeta D, Nistico` V, Saramandi A, Bertelli S, Todisco P, Gambini O, Fotopoulou A (2021) The anticipation and perception of affective touch in women with and recovered from anorexia nervosa. Neuroscience 464: 143–155 61. Kirsch LP, Besharati S, Papadaki C, Crucianelli L, Bertagnoli S, Ward N, Moro V, Jenkinson PM, Fotopoulou A (2020) Damage to the right insula disrupts the perception of affective touch. elife 9: e47895 62. Voos AC, Pelphrey KA, Kaiser MD (2013) Autistic traits are associated with diminished neural response to affective touch. Soc Cogn Affect Neurosci 8(4):378–386 63. Haggarty CJ, Malinowski P, McGlone FP, Walker SC (2020) Autistic traits modulate cortical responses to affective but not discriminative touch. Eur J Neurosci 51(8): 1844–1855 64. Schandry R (1981) Heart beat perception and emotional experience. Psychophysiology 18(4):483–488 65. Croy I, Geide H, Paulus M, Weidner K, Olausson H (2016) Affective touch awareness in mental health and disease relates to autistic traits–an explorative neurophysiological investigation. Psychiatry Res 245:491–496 66. Croy I, Bierling A, Sailer U, Ackerley R (2021) Individual variability of pleasantness ratings to stroking touch over different velocities. Neuroscience 464:33–43 67. Crucianelli L, Chancel M, Ehrsson HH (2022) Modeling affective touch pleasantness across skin types at the individual level reveals a reliable and stable basic function. J Neurophysiol 128(6):1435–1452 68. Bjo¨rnsdotter M (2016) Brain processing of CT-targeted stimulation. In: Olausson H, Wessberg J, McGlone F (eds) Affective touch and the neurophysiology of CT afferents. Springer, New York, pp 187–194 69. Morrison I (2016) ALE meta-analysis reveals dissociable networks for affective and discriminative aspects of touch. Hum Brain Mapp 37(4):1308–1320 70. Olausson HW, Cole J, Vallbo Å, McGlone F, Elam M, Kr€amer HH, Rylander K, Wessberg J, Bushnell MC (2008) Unmyelinated tactile afferents have opposite effects on insular and somatosensory cortical processing. Neurosci Lett 436(2):128–132

71. Morrison I, Lo¨ken LS, Minde J, Wessberg J, Perini I, Nennesmo I, Olausson H (2011) Reduced C-afferent fibre density affects perceived pleasantness and empathy for touch. Brain 134(4):1116–1126 72. Andrew D, Craig AD (2016) Processing of C-tactile information in the spinal cord. In: Olausson H, Wessberg J, McGlone F (eds) Affective touch and the neurophysiology of CT afferents. Springer, New York, pp 159–173 73. Morrison I (2016) CT afferent-mediated affective touch: brain networks and functional hypotheses. In: Olausson H, Wessberg J, McGlone F (eds) Affective touch and the neurophysiology of CT afferents. Springer, New York, pp 195–208 74. Marshall AG, Sharma ML, Marley K, Olausson H, McGlone FP (2019) Spinal signalling of C-fiber mediated pleasant touch in humans. elife 8:e51642 75. Tsiakaka O, Feruglio S (2019) Toward the monitoring of the spinal cord: a feasibility study. Microelectron J 88:145–153 76. Hagberg EE, Ackerley R, Lundqvist D, Schneiderman J, Jousm€aki V, Wessberg J (2019) Spatio-temporal profile of brain activity during gentle touch investigated with magnetoencephalography. NeuroImage 201: 116024 77. J€anig W (2018) Peripheral thermoreceptors in innocuous temperature detection. In: Handbook of clinical neurology, vol 156. Elsevier, London, pp 47–56 78. Perini I, Ceko M, Cerliani L, van EttingerVeenstra H, Minde J, Morrison I (2020) Mutation carriers with reduced C-afferent density reveal cortical dynamics of pain–action relationship during acute pain. Cereb Cortex 30:4858–4870 79. Heldestad V, Linder J, Sellersjo¨ L, Nordh E (2010) Reproducibility and influence of test modality order on thermal perception and thermal pain thresholds in quantitative sensory testing. Clin Neurophysiol 121(11): 1878–1885 80. Melzack R, Katz J (2013) Pain measurement in adult patients. In: McMahon SB, Koltzenburg M, Tracey I, Turk DC (eds) Wall & Melzack’s textbook of pain. Elsevier, Philidelphia, pp 301–314 81. Melzack R (1975) The McGill Pain Questionnaire: major properties and scoring methods. Pain 1(3):277–299 82. Gibson SJ, Helme RD (2001) Age-related differences in pain perception and report. Clin Geriatr Med 17(3):433–456

An Interoceptive take on Touch 83. Lautenbacher S, Kunz M, Strate P, Nielsen J, Arendt-Nielsen L (2005) Age effects on pain thresholds, temporal summation and spatial summation of heat and pressure pain. Pain 115(3):410–418 84. Fillingim RB, King CD, Ribeiro-Dasilva MC, Rahim-Williams B, Riley JL III (2009) Sex, gender, and pain: a review of recent clinical and experimental findings. J Pain 10(5): 447–485 85. Greenspan JD, Traub RJ (2013) Gender differences in pain and its relief. Wall and Melzack’s textbook of pain, 6th edn. Elsevier Saunders, Philadelphia, pp 221–231 86. Tashani OA, Alabas OA, Johnson MI (2010) Cold pressor pain responses in healthy Libyans: effect of sex/gender, anxiety, and body size. Gend Med 7(4):309–319 87. Mancini F, Bauleo A, Cole J, Lui F, Porro CA, Haggard P, Iannetti GD (2014) Whole-body mapping of spatial acuity for pain and touch. Ann Neurol 75(6):917–924 88. Krahe´ C, Drabek MM, Paloyelis Y, Fotopoulou A (2016) Affective touch and attachment style modulate pain: a laser-evoked potentials study. Philos Trans R Soc Lond B Biol Sci 371(1708):20160009 ˜ a-Vivas V, Fleming S, Haggard P 89. Beck B, Pen (2019) Metacognition across sensory modalities: vision, warmth, and nociceptive pain. Cognition 186:32–41 90. Longo MR, Iannetti GD, Mancini F, Driver J, Haggard P (2012) Linking pain and the body: neural correlates of visually induced analgesia. J Neurosci 32(8):2601–2607 91. Crucianelli L, Enmalm A, Ehrsson HH (2021) Thermosensation as a novel method to probe interoception and its relationship with other interoceptive modalities. bioRxiv. 92. Granovsky Y, Matre D, Sokolik A, Lorenz J, Casey KL (2005) Thermoreceptive innervation of human glabrous and hairy skin: a contact heat evoked potential analysis. Pain 115(3):238–247 93. Wydenkeller S, Wirz R, Halder P (2008) Spinothalamic tract conduction velocity estimated using contact heat evoked potentials: what needs to be considered. Clin Neurophysiol 119(4):812–821 94. Barati Z, Zakeri I, Pourrezaei K (2017) Functional near-infrared spectroscopy study on tonic pain activation by cold pressor test. Neurophotonics 4(1):015004 95. Tracey I (2008) Imaging pain. Br J Anaesth 101(1):32–39

223

96. Brooks J, Tracey I (2005) From nociception to pain perception: imaging the spinal and supraspinal pathways. J Anat 207(1):19–33 97. Tinnermann A, Buechel C, Cohen-Adad J (2021) Cortico-spinal imaging to study pain. NeuroImage 224:117439 98. Ostrowsky K, Magnin M, Ryvlin P, Isnard J, Guenot M, Mauguiere F (2002) Representation of pain and somatic sensation in the human insula: a study of responses to direct electrical cortical stimulation. Cereb Cortex 12(4):376–385 99. Iannetti GD, Mouraux A (2010) From the neuromatrix to the pain matrix (and back). Exp Brain Res 205(1):1–12 100. Mouraux A, Iannetti GD, Colon E, Nozaradan S, Legrain V, Plaghki L (2011) Nociceptive steady-state evoked potentials elicited by rapid periodic thermal stimulation of cutaneous nociceptors. J Neurosci 31(16): 6079–6087 101. Liberati G, Klo¨cker A, Safronova MM, Ferrao Santos S, Ribeiro Vaz JG, Raftopoulos C, Mouraux A (2016) Nociceptive local field potentials recorded from the human insula are not specific for nociception. PLoS Biol 14(1):e1002345 102. Wilhelm FH, Kochar AS, Roth WT, Gross JJ (2001) Social anxiety and response to touch: incongruence between self-evaluative and physiological reactions. Biol Psychol 58(3): 181–202 103. Lovibond PF, Lovibond SH (1995) The structure of negative emotional states: comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behav Res Ther 33(3): 335–343 104. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA (1983) Manual for the state-trait anxiety scale. Consulting Psychologists Press, Palo Alto 105. Krahe´ C, Paloyelis Y, Condon H, Jenkinson PM, Williams SC, Fotopoulou A (2015) Attachment style moderates partner presence effects on pain: a laser-evoked potentials study. Soc Cogn Affect Neurosci 10(8): 1030–1037 106. von Mohr M, Krahe´ C, Beck B, Fotopoulou A (2018) The social buffering of pain by affective touch: a laser-evoked potential study in romantic couples. Soc Cogn Affect Neurosci 13(11):1121–1130 107. Krahe´ C, Paloyelis Y, Sambo CF, Fotopoulou A (2014) I like it when my partner holds my hand: development of the Responses and

224

Laura Crucianelli and India Morrison

Attitudes to Support during Pain questionnaire (RASP). Front Psychol 5:1027 108. Krahe´ C, von Mohr M, Gentsch A, Guy L, Vari C, Nolte T, Fotopoulou A (2018) Sensitivity to CT-optimal, affective touch depends on adult attachment style. Sci Rep 8(1):1–10 109. Spitoni GF, Zingaretti P, Giovanardi G, Antonucci G, Galati G, Lingiardi V, Cruciani G, Titone G, Boccia M (2020) Disorganized attachment pattern affects the perception of affective touch. Sci Rep 10(1): 1–10 110. Fraley RC, Waller NG, Brennan KA (2000) An item response theory analysis of self-report measures of adult attachment. J Pers Soc Psychol 78(2):350 111. Hazen RL, Shaver S (1987) Romantic love conceptualized as an attachment process. J Pers Soc Psychol 52:511–524 112. Ronga I, Valentini E, Mouraux A, Iannetti GD (2013) Novelty is not enough: laserevoked potentials are determined by stimulus saliency, not absolute novelty. J Neurophysiol 109(3):692–701 113. Iannetti GD, Leandri M, Truini A, Zambreanu L, Cruccu G, Tracey I (2004) Aδ nociceptor response to laser stimuli: selective effect of stimulus duration on skin

temperature, brain potentials and pain perception. Clin Neurophysiol 115(11): 2629–2637 114. Leandri M, Saturno M, Spadavecchia L, Iannetti GD, Cruccu G, Truini A (2006) Measurement of skin temperature after infrared laser stimulation. Neurophysiol Clin 36(4): 207–218 115. Crucianelli L, Enmalm A, Ehrsson HH (2022) Interoception as independent cardiac, thermosensory, nociceptive, and affective touch perceptual submodalities. Biol Psychol 108355 116. Garfinkel SN, Seth AK, Barrett AB, Suzuki K, Critchley HD (2015) Knowing your own heart: distinguishing interoceptive accuracy from interoceptive awareness. Biol Psychol 104:65–74 117. Shields SA, Mallory ME, Simon A (1989) The body awareness questionnaire: reliability and validity. J Pers Assess 53(4):802–815 118. Porges S (1993) Body perception questionnaire. Laboratory of Developmental Assessment, University of Maryland 119. Mehling WE, Price C, Daubenmier JJ, Acree M, Bartmess E, Stewart A (2012) The multidimensional assessment of interoceptive awareness (MAIA). PLoS One 7(11):e48230

Part III Individual Differences, Development, and Illusions

Chapter 11 Atypical Development of Tactile Processing Nicolaas A. J. Puts and Carissa J. Cascio Abstract Tactile processing is of tremendous importance during development. Touch is one of the first senses to develop, and is even active prenatally. Touch allows children not only to explore the physical world, but also to form social bonds. Both atypical touch perception and the absence of touch in early development have been linked to poor social development. In particular, difficulties in tactile perception and processing have been established in neurodevelopmental conditions such as autism spectrum disorder and attention deficit hyperactive disorder, and recent studies show an impact on the core symptoms of those conditions. Tactile difficulties have been examined in neurodevelopment using a variety of approaches including psychophysics, questionnaires, and brain imaging, examining touch at different conceptual levels. While the link between these levels is increasingly being studied, each informs a different aspect of atypical touch perception. Here, we summarize the different approaches used to examine touch in developmental conditions, including observational, dimensional, psychophysical, and neuroimaging work, and discuss the limitations of these approaches. Key words Atypical touch processing, disorders

1 1.1

Psychophysics,

Questionnaires,

Neurodevelopmental

Introduction Background

Tactile processing plays an extremely important role in development. Prenatally, touch is one of the first senses to develop [1–3] shown to start as early as 8 weeks into gestation. Better developmental outcomes such as weight gain, healthy sleep/wake cycles, and better motor development, have been associated with more tactile and kinesthetic stimulation from caregivers [4]. Furthermore, touch plays an important role in social development, such as in the formation of dyadic groups [5]. The role of touch in mother-infant attachment has been well documented [6]. Studies have shown that in early childhood, tactile perception influences children’s ability to engage in play with peers [7].

227

228

Nicolaas A. J. Puts and Carissa J. Cascio

Fig. 1 Correlations between sensory scores. (Adapted from https://psyarxiv.com/ay73t/). Sensory scores were obtained from a large cohort of children who were neurotypical and neurodiverse. Sensory hyper- and hyporesponsivity as well as sensory seeking scores were obtained from the Sensory Experiences Questionnaire. Scores were shown to correlate strongly and positively

Differences in tactile perception have become of increased interest with respect to neurodevelopmental conditions. Of those, sensory processing differences have been particularly well-described in autism spectrum disorder (ASD), where sensory differences are common, reported in 60–90% of people with ASD [8, 9], although presentation is extremely heterogeneous [10]. As of the fifth version of the American Psychiatric Association’s diagnostic and statistical manual, DSM-5, sensory difficulties have been added as a core symptom of autism. Tactile difficulties are among the mostreported sensory alterations. These alterations are often reported in terms of clinicallydefined patterns in observable responses to sensory stimuli (e.g., hyper- and hypo-responsivity, as well as ‘sensory seeking’ in which individuals overly seek out sensory stimulation [11]. Although hyper- and hypo-responsivity are seen as different constructs, recent work shows a strong positive correlation, such that individuals who are hyper-responsive tend to be hypo-responsive as well [12] (Fig. 1). These findings suggest that people tend to have global difficulties in interpreting sensory input rather than being a particular subtype defined by one stable pattern of difficulty within an individual. An increasing number of studies have reported altered perception of touch, including differences in tactile detection and discrimination [12–17] although with varied outcomes [18, 19], and potentially important roles of non-sensory factors such as response criterion [20] or response variability [12, 21]. People on the autism spectrum also appear to differ in their experience of affective touch [22] and its association with neural responses [23]. Difficulties in sensory processing have also been described in non-idiopathic autism, but little work has focused on the tactile domain.

Touch Processing in Atypical Development

229

In addition to autism, tactile differences have been described in other neurodevelopmental disorders such as attention deficit hyperactivity disorder (ADHD) and Tourette Syndrome (TS). In ADHD, sensory difficulties are often associated with difficulties in attention [24, 25]. Hyper- and hypo-responsivity are commonly reported for ADHD, although these are also often described as difficulties in habituation or paying attention to sensory information [24, 25]. TS typically emerges in childhood or adolescence, and is characterized by involuntary movements and vocalizations (tics). Individuals with TS often report tactile difficulties [26, 27]. The premonitory urge to tic is often reported as tactile sensations, and sensory over-responsivity or alterations in sensory habituation have also been reported. These are positively associated with clinical symptoms [28, 29]. Finally, although not always considered a stand-alone condition, sensory processing disorder (SPD) is defined by difficulties in processing sensory information, including both hyper- and hypo-responsivity [30, 31], which often leads to clumsiness and feelings of disembodiment. Because there are conflicting opinions about whether SPD should be recognized as a separate condition, it has been studied less than the more established clinical diagnostic groups, for which sensory processing is a part, but not the whole, of the clinical presentation. 1.2 The Neurodevelopment of Tactile Function Studied Using Multiple Methods

The examination of tactile differences in neurodevelopmental disorders is unique in that the vast majority of work has used clinical approaches to study difficulties in sensory function, with varying tactile components as part of those examinations. These studies have predominantly been performed in ASD, where tactile differences are most recognized. Most of the work to date has used structured but subjective caregiver-, teacher-, or self-report to examine the extent of sensory differences in neurodevelopment. While informative [32] and high in ecological validity based on sampling from multiple contexts in daily life, these approaches are limited in their ability to address the specificity of sensory and, more specifically, tactile difficulties. Many questionnaires differentiate between sensory domains (e.g., sight, touch, and hearing), although often with limited items for each sensory domain. Furthermore, it remains unclear to what extent questionnaire scores are driven by perceptual, rather than emotional reactions to sensory information, since questionnaire items often include both sensory sensitivity and reactivity or responsivity. The difference lies in whether someone’s perception of a stimulus is atypical, versus whether someone’s response to sensory stimuli is atypical. Indeed, many of the embedded questions assess an individual’s subjective behavioral reaction to sensory stimuli (e.g., avoidance of crowded places), rather than provide an objective measurement of actual alterations in sensory processing. Moreover, it remains unclear whether and how they relate to the core symptoms of each of these developmental disorders.

230

Nicolaas A. J. Puts and Carissa J. Cascio

Recent years have seen an increase in the use of psychophysical approaches to examine perceptual and affective differences in touch perception in neurodevelopmental disorders. However, as will be described in this chapter, these are often limited by difficulties in applying these methods to clinical developmental populations. Psychophysical approaches limit application in the most affected populations (as individuals cannot adequately perform the tasks), and the interpretation suffers from heterogeneity in the type of approaches used, heterogeneity in participant populations, and small participant cohorts [18]. Despite these limitations, difficulties in tactile processing at the perceptual level have increasingly been described in neurodevelopmental conditions. Particularly in autism, there has been an emergence of work showing worse tactile detection [12–15, 33], although it should be noted that there are inconsistencies between “flutter” and “vibration” senses. There is evidence for poorer tactile discrimination of both vibrotactile amplitude and frequency, more nuanced aspects of vibrotactile timing [13, 34, 35], textures [36], and poor tactile adaptation [13, 37]. Perhaps more relevant clinically, several studies have reported that these difficulties in low-level perception are associated with the core clinical symptoms of autism; poor discrimination correlates with communication difficulties and poor detection with heightened sensory reactivity. Difficulties in tactile detection are associated with clinical markers of hyperreactivity [12, 14], core social symptoms of autism in both neurotypical [38] and autistic [14] populations, and attentional symptoms in autism samples [39]. Recent work also found an impact of non-sensory factors such as response criterion [20] and response variability [12, 21]. It remains unclear what drives these differences, but some evidence suggests a role of cortical hyperexcitability in driving these perceptual and clinical differences [40, 41]. Beyond questionnaires and the psychophysical approaches of detection or discrimination of low-level sensory features, the study of social touch has emerged as a novel link between the assessment of perception and its social impacts. Social touch is often operationalized as touch targeted to the optimal response of C-Tactile or CT-afferents, small unmyelinated fibers that project to the posterior insula, and respond preferentially to gentle, stroking touch [42–44] (Chapters 6, 15, this volume). However, both physiological and psychological factors determine the affective or social nature of touch [45], and the physiological definition should not be limited only to CT-afferents, as other types of C fibers such as those for temperature also have affective and social relevance [46, 47]. Furthermore, social touch is often assessed by establishing the valence of sensory stimulations. Affective touch between infants and caregivers is thought to be an important component of establishing attachment [48, 49], upon which subsequent social developmental milestones are built

Touch Processing in Atypical Development

231

[50]. The role of social or affective touch in early development has been studied in the context of autism by Kadlaskar and colleagues [51], who noted that 12-month-olds at elevated genetic likelihood for autism tended not to orient as consistently to caregiver touch, and this reduced orienting predicts autism severity at two years of age. This was despite evidence suggesting that parents of children with neurodevelopmental disabilities tend to use touch more than parents of typically-developing children, though the nature of touch (i.e., playful vs. affectionate) differentially influences children’s social attention across groups [52]. Children and adults with autism have altered behavioral [53, 54] and neural [22, 45] responses to affective touch (e.g., textures normatively rated as highly pleasant or unpleasant), suggesting that altered interpretation of the emotional and social aspects of touch is a lifelong phenomenon for people on the autism spectrum. Finally, imaging approaches allow the neurobiological correlates of altered tactile function to be determined. People with autism show diminished blood oxygenation level-dependent (BOLD) response to pleasant [22, 45], and an exaggerated response to unpleasant tactile stimulation [45, 55] in both sensory and limbic regions of the brain. The posterior temporal sulcus has been implicated in altered affective touch processing in autism, showing reduced relation to perceived pleasantness of affective touch [23]. Connectivity between sensory, limbic, and semantic regions during pleasant and unpleasant touch is also affected by autism [56]. Green and colleagues [57] demonstrated a pattern of decreased habituation in the amygdala for repeated mildly aversive tactile stimuli, in young people with autism who self-reported high levels of sensory over-reactivity. Some imaging studies have indicated that this sensory over-reactivity is also related to differences in thalamocortical gating [40, 58, 59] in autism. Looking specifically at tactile reactivity patterns in autism, electroencephalographic (EEG, see Chapter 19, this volume) evidence suggests that tactile hyper-reactivity is related to cortical processing of touch at an earlier processing stage than hypo-reactivity, implicating potentially dissociable mechanisms of perception and attention for the two patterns [60]. A recent EEG study in infants suggested that altered responses to paired-pulse stimulation at 11 months are predictive of their neurodevelopmental outcome at 24 months [61]. Finally, studies of brain structure in autism also suggest that tactile reactivity patterns relate to white matter connections within the insula [62] and inferior longitudinal fasciculus [63]. The neurodevelopmental process of creating multiple somatotopic representations of the body surface, and the variety of inputs to it, is one that supports preverbal social development, and has profound implications for the study of disordered development including, but not limited to, autism [64].

232

1.3

2 2.1

Nicolaas A. J. Puts and Carissa J. Cascio

Summary

Tactile differences have been well established in a variety of neurodevelopmental conditions, and increasing evidence links the different levels of investigation, whereby changes in neurophysiology lead to altered perception. This may in turn exacerbate, or even causally contribute to, some of the core features of ASD. Given that these differences are so well-established, their underlying physiology is a relatively easy target to study. Given that there is a potential impact on core clinical symptoms, studying the atypical development of touch has a huge potential for the development of biological markers of neurodevelopmental conditions (see Note 4.1). There are a number of different approaches to examine these tactile differences. In this chapter, we set out the variety of methods that have been used, the requirements, and their costs and benefits. Finally, we discuss limitations and practical considerations of these approaches.

Materials Questionnaires

2.2 ObservationBased Approaches

Many of the questionnaire and observation-based approaches to measurement of atypical sensation and perception are not specific to the tactile domain. Most include tactile components, but often consider sensory processing as a whole. Questionnaire-based approaches (see Subheading 3 for examples) are fairly material-light, often need training and normalization procedures, and these measures need to be validated. Often the materials are copyrighted and thus need to be purchased. Questionnaire methods are often performed with pen and paper, although increasingly are available digitally. Often, these questionnaire approaches require manual transcribing to databases. Users should take note of how and for whom questionnaires have been validated. Many measures were developed and normed on populations with typical development or using mixed samples, and factor structures derived in this way do not always generalize to populations with specific neurodevelopmental disabilities [65]. Observation-based approaches entail clinician or experimenter observation of directed engagement with sensory stimuli. For the tactile domain, these allow for engagement with tactile toys. In terms of materials, these tactile toys need to be acquired (often through a distributor) as they consist of standardized toys with specific structures. Testing usually requires dedicated rooms, often child-proof with limited distraction, yet with sufficient appeal to pediatric cohorts (not ‘too clinical’). Recently, researchers have developed measures that combine questionnaire and observational approaches into a single instrument, balancing the strengths and weaknesses of each approach used in isolation [66].

Touch Processing in Atypical Development

233

2.3 Quantitative Sensory Approaches

The domain of lab-based or more objective testing in neurodevelopmental conditions starts with quantitative sensory stimulation approaches, including quantitative sensory testing (QST) and the Evaluation in Ayres Sensory Integration (EASI, see Subheading 3). These approaches provide structured engagement of the participant in a range of sensory (and often motor) tasks. For the tactile domain, these refer to, for example, digit indication and detection threshold (e.g., using von Frey hairs). These also require dedicated testing rooms.

2.4 Psychophysical Approaches

Psychophysical approaches are less common and often require dedicated equipment. Psychophysical approaches in neurodevelopmental cohorts are generally consistent with other psychophysical approaches, and we thus refer to those chapters for more detail (Chapters 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10, this volume). Psychophysical approaches may use von Frey hair, vibrotactile stimulators (either with electrical or vibratory stimulation), or graded structures. For psychophysical approaches in pediatric or clinical cohorts, we also recommend taking into account room noise, room temperature, and standardized setups with a dedicated computer, seating situation, seating distance, and response options. Any experimental software may be used, and studies have reported psychophysical data from neurodevelopmental cohorts through a number of different approaches, including commercial software (e.g., Cortical Metrics), Presentation, Matlab, PsychoPy, and others.

2.5 Social and Affective Touch

Social and affective touch (see Chapter 6, this volume) approaches typically use objective structured paradigms, where stimulation is provided by a robot or the experimenter [67]. Stimuli may consist of specific structures with different textures. Experiments may use items commonly found in shops (e.g., plastic) or may purchase custom-made structures from vendors which are more consistent (e.g., gratings with precise spacing between ridges). Social touch typically requires an experimenter, and some experiments may require visual presentation of stimuli. Experimenters may present stimuli manually, using soft materials such as cotton swabs or brushes, or even with just the fingertip, which has higher ecological validity than mechanically delivered touch, but at the expense of precise control of speed and force. This is important because social and affective touch is often experimentally defined within a range of speeds and forces based on the responses of peripheral afferent nerves. Some studies have also begun to examine active social touch (i.e., giving of touch rather than passive receipt of touch, [68]), which has potential applications for neurodevelopmental cohorts. Relatively few studies have performed social and affective touch experiments in children.

234

Nicolaas A. J. Puts and Carissa J. Cascio

2.6 Imaging Approaches

3

Imaging approaches in neurodevelopment are no different from those in neurotypical populations. Imaging equipment is expensive, and may require the use of scanning facilities (e.g., magnetic resonance imaging, MRI or magnetoencephalography, MEG facilities) or dedicated equipment (e.g., EEG caps, see Chapters 18 and 19, this volume). For neurodevelopmental conditions, it is especially pertinent that this equipment is suitable for a clinical pediatric population, and caution should be taken with any difficulties wearing EEG caps (children with tactile difficulties may not tolerate these), compliance, and movement. Training staff to work with special populations can increase compliance and data quality. Near-infrared spectroscopy (NIRS) may also be a useful non-invasive method which allows for optical measurement of tissue oxygenation in the cortex. In terms of tactile stimulation, equipment needs to be suitable for imaging purposes; for MRI these need to be MRI-compatible (i.e., non-ferromagnetic) which often limits their use and increases the costs. For EEG these need to not affect the EEG recording itself and need to have a means of sending a trigger to mark the continuous EEG trace to indicate precisely when stimulation occurred. For infants, one may not be able to stimulate the fingertips; one study chose to stimulate the feet instead [61].

Methods In this section, we briefly discuss the main methods used for tactile (sensory) assessment in clinical neurodevelopmental conditions.

3.1

Questionnaires

A large number of different inventories or questionnaires exist, and here we discuss only the major ones. Please note that several of these measures continuously undergo revision or amendment, so several versions of the questionnaires may exist (see Note 4.2).

3.1.1 Short Sensory Profile (SSP)

The SSP [69] comprises a 38-item parent or caregiver-report questionnaire that assesses sensory processing differences across different modalities: visual, auditory, tactile, taste, movement, and proprioception, as well as sensory under-responsiveness. Item responses are given on a five-point Likert rating scale from 1 (always occurs) to 5 (never occurs). As well as a total score, the measure results in seven subscales: Tactile Sensitivity, Taste/Smell Sensitivity, Movement Sensitivity, Under-responsive/Seeks Sensation, Auditory Filtering, Low Energy/Weak, and Visual/Auditory Sensitivity.

3.1.2 Adult/Adolescent Sensory Profile (AASP)

Developed by the authors of the SP, the AASP [70], https://www. pearsonclinical.co.uk/store/ukassessments/en/Store/Profes sional-Assessments/Motor-Sensory/Adolescent-Adult-SensoryProfile/p/P100009054.html?tab=resources) is a 60-item self-

Touch Processing in Atypical Development

235

report questionnaire that mirrors the modalities and generates quadrant scores in two dimensions: high versus low registration (whether a sensory stimulus is registered or oriented to), and reactivity (seeking or avoiding sensory input). It is normed for ages 11 and over, and uses a five-point Likert rating scale across six domains (taste/smell, movement, visual, touch, activity level, and auditory processing). 3.1.3 Sensory Perception Quotient (SPQ)

The SPQ [71], is a 92-item self-report measure with one-dimensional scale (0 = hyper-sensitivity, 276 = hypo-sensitivity) across five sensory domains, with each item scored on a fourpoint Likert scale. It was originally amended from the AASP, addressing the concern that the AASP merges measures of reactivity and sensitivity, whereas the SPQ focuses on the perception of sensory stimuli rather than responses to sensory stimuli. The revised version [72] is scored in both hyper- and hypo-sensitivity (rather than one or the other), with higher scores leading to more atypical sensitivity.

3.1.4 Sensory Experiences Questionnaire (SEQ3)

The SEQ3 [73–75] is a caregiver report used to characterize sensory features in children ages 2–12 years with ASD and/or other developmental disabilities, and contains both social and non-social items. It can also be administered in a structured interview format if necessary. The latest version, the SEQ Version 3.0, contains 105 items that assess the frequencies of sensory behaviors across patterns of response (i.e., hypo-responsiveness, hyper-responsiveness, sensory seeking, and enhanced perception) and in sensory modalities (i.e., auditory, visual, tactile, gustatory, and vestibular).

3.1.5 Sensory Processing Measure (SPM)

The SPM [76] is a scale that assesses sensory processing, as well as praxis (motor control) and social function in school-aged children of 5 to 12 years, and contains three sections, a home (75 items), classroom (62 items), and school environment section (10–15 items). For the home and classroom sections, scores are obtained for 5 sensory domains: vision, hearing, touch, body awareness, and balance or motion, as well as measures for social participation and praxis. A total score is also computed.

3.1.6 Glasgow Sensory Questionnaire (GSQ)

The GSQ [77] is an adult scale with 42 items across seven modalities, which assesses both hyper- and hypo-sensitivities (visual, auditory, gustatory, olfactory, tactile, vestibular, and proprioceptive). Three questions explicitly assess hyper-responsivity and hyporesponsivity. Items reflect frequency of experiences on a 5-point Likert scale, with the possible total scores ranging from 0 to 168, or separated for each modality.

236

Nicolaas A. J. Puts and Carissa J. Cascio

3.2 ObservationBased Approaches

Observation-based approaches go beyond mere rating scales or questionnaires, and examine sensory responsivity through structured observation in a clinical or laboratory setting (see Note 4.3).

3.2.1 Sensory Assessment for Neurodevelopmental Differences (SAND)

The SAND [66] measures sensory reactivity in a standardized observational assessment and an associated caregiver interview. In the direct observation, children are presented with a range of sensory toys with tactile (e.g., textured toy), auditory (e.g., musical toy), and visual (e.g., flashing toy) components. Children are scored for behavioral responses consistent with sensory hyperreactivity, hypo-reactivity, and seeking. The administration is conducted by a trained examiner, and behaviors are scored dichotomously (1 = present, 0 = not present) across 36 items. If a behavior is present there is the additional coding of a severity score (1 = mild, 2 = moderate to severe). The subsequent interview component requires the caregiver to indicate whether their child presents with the various sensory reactivity differences in visual, tactile, and auditory domains (1 = present, 0 = not present), and to rate the severity if present (1 = mild, 2 = moderate to severe). The items of this interview map onto the scoring for the observation. Observed and reported scores can then be combined to produce a total score from 0 to 90, and subscales from 0 to 30 for each of sensory hyper-reactivity, hypo-reactivity, and sensory seeking. Higher scores reflect greater sensory reactivity differences.

3.2.2 Sensory OverResponsivity (SensOR) Scales

As with the SAND, the sensOR [78] includes an examiner-led evaluation and a caregiver rating scale (the inventory). Both scales measure sensory over-responsivity in seven sensory domains (tactile, auditory, visual, olfactory, taste, vestibular, and proprioceptive). The scale is a directed experiment where responses to sensory stimuli in all domains are registered. The inventory is a 76-item questionnaire with eight categories (Tactile-Textures, Tactile-Activities of Daily Living, Auditory Settings, Auditory-Specific, Visual, Olfactory, Movement-Proprioceptive, and Food-Textures/ Eating).

3.2.3 The Sensory Processing 3-Dimensions Scale (SP-3D)

The SP-3D [31, 79] is a performance-based measure for assessing sensory processing abilities and challenges, including sensory modulation difficulties, difficulties in sensory discrimination, and sensory-based motor conditions. This scale includes both a performance-based scale and a caretaker questionnaire (the ‘inventory’) for evaluating sensory processing in children. The SP-3D has seven domains, each consisting of between 4 and 8 subsections. Sensory modulation and discrimination abilities are measured within five sensory domains (visual, tactile, proprioceptive, vestibular, and auditory). Sensory modulation is organized by three scales:

Touch Processing in Atypical Development

237

sensory over-responsivity (SOR), sensory under-responsivity (SUR), and sensory craving (SC). Sensory-based motor abilities are measured by subtests within posture and praxis. 3.3 Quantitative Sensory Approaches

Several approaches to so-called quantitative sensory testing are available (see Note 4.4).

3.3.1

NIH Toolbox

Structured sensory testing containing a tactile component can be performed as part of the NIH toolbox for the assessment of neurological and behavioral function. The NIH Toolbox® is a comprehensive set of neurobehavioral measurements that allow for the assessment of cognitive, emotional, sensory, and motor functions [80]. For tactile perception [81], tactile discrimination tests are applied as a 3AFC-forced choice design for texture gratings. Participants must manually explore three gratings and determine which differs from the other two. Tactile detection is measured on the bottom of the foot, using Semmes-Weinstein filaments (identical to Von Frey hairs) applying standard protocols. A form perception test is used to match an unseen tactile stimulus to an image or to a previously felt shape.

3.3.2 Sensory Integration and Praxis Test (SIPT)

The SIPT [82] is a standardized battery designed to evaluate various aspects of sensory processing in children. It has been well validated through factor and cluster-based analyses [83] and has good reliability. With respect to tactile function, the SIPT has five tactile subtests and includes finger identification, graphesthesia, manual form perception, kinesthesia, and localization of tactile stimuli.

3.3.3 Evaluation in Ayres Sensory Integration (EASI)

A more recent approach, currently undergoing normalization, is the EASI [83–85], which is suitable for children 3–12 years of age. The EASI aims to provide a valid and reliable assessment of sensory integration function. Beyond tactile function, it measures perception and posture, as well as motor integration and reactivity. The tactile tests include tactile localization, tactile designs (in which a form needs to be recognized), shapes, and oral discrimination. These also allow for testing tactile hyper and hypo-responsivity.

3.4 Psychophysical Approaches

The application of well-established psychophysical approaches to tactile perception (see Chapter 1, this volume) to clinical and developmental samples represents a promise to add objectivity and reproducibility to the research landscape. Often, these studies are grounded in a deeper understanding of somatosensory cortical processing, allowing for linking findings more directly to biological mechanisms. Recent work has seen increased use of psychophysical approaches in the study of neurodevelopment, with tailored pediatric approaches or approaches for those with intellectual difficulties.

238

Nicolaas A. J. Puts and Carissa J. Cascio

Psychophysical approaches in pediatric or neurodevelopmental cohorts do not differ tremendously from those used in general, as discussed elsewhere in this volume. We will therefore not describe specific methods. All typical psychophysical approaches have been used in neurodevelopment, including flutter, vibration, monofilament, and electrical stimuli; method-of-constant stimuli, staircasebased approaches, Yes/No, and 2- or 3-AFC designs. Studies of tactile pleasantness have often used psychophysical approaches, including visual analog scale (VAS) ratings or similar ways of quantifying the subjective feeling of pleasantness, roughness, or other attributes of the stimuli. With respect to the domains studied, neurodevelopmental research has examined various aspects of somatosensory perception, including detection, discrimination, temporal order judgments, as well as modulation of tactile perception through adaptation (Subheading 2). The literature is mostly inconsistent with respect to these findings but, as discussed elsewhere [18], this is largely due to small samples and heterogeneous approaches. Differences between healthy and neurodevelopmental populations have been found in a variety of tactile perceptual metrics. In brief, difficulties in sensory adaptation seem to be prevalent in ADHD and Tourette Syndrome, whereas those with SPD and ASD appear to be affected across several tactile domains. The strongest consistency comes from several studies that suggest difficulties in low-level detection and discrimination in autism, although the role of attention and response difficulties in these needs further study. One often-seen difference between neurodevelopmental and “conventional” psychophysics is that studies often suffer from reduced trial numbers. This also somewhat limits the use of method of constant stimuli (MCS) approaches due to their duration, and related lapses in participants’ attention are more common in pediatric and neurodevelopmental conditions (see Notes 4.5 and 4.6). 3.5 Social and Affective Touch

Social and affective touch approaches typically use objective structured methods, where stimulation is provided by a robot or by the experimenter, and stimulus parameters are changed on a structured basis [67] (see Chapter 6, this volume). Participants may be asked to rate the stimulus or rate their preference (or aversion) for a certain stimulus. For non-verbal participants, implicit measures of affective response may be recorded, including heart rate, skin conductance, and changes in facial expression. Social touch typically requires an experimenter, while some experiments may require visual presentation of stimuli (see Note 4.7).

3.6 Imaging Approaches

Methods for brain imaging in atypical development do not differ substantially from those in neurotypical cohorts. Stimulation is often-performed using MRI or EEG-compliant stimulators, or by individual experimenters. The approach depends on the type of

Touch Processing in Atypical Development

239

stimulation (vibrations, structures, or even visual scenes of tactile situations). While there are no dedicated approaches purely for neurodevelopmental purposes, successful data acquisition critically depends on adequately preparing participants for the experience of EEG or MRI. This may include systematic desensitization with staff trained to work with individuals with atypical neurodevelopment, in the context of “mock” scanning or recording sessions and acclimation to the sensory aspects of the environment such as loud noises or damp EEG nets. See Chapters 18 and 19 for more on these methods (see Note 4.8). 3.7 Interpretation of Altered Tactile Processing in Neurodevelopment

Although sensory tactile difficulties have been clinically recognized in neurodevelopmental conditions for several decades, only recently has there been an emergence of studies focusing on the nature and the impact of these difficulties. Autistic individuals often report that their sensory difficulties are extremely debilitating and affect their ability to participate in society. From a developmental perspective, it is well-known that an absence or reduction of tactile stimulation in early life has an impact on social development, yet this argument is only now gaining traction with respect to autism. Recent theories suggest that early difficulties in low-level tactile experiences, be these perceptual or social, may lead, to or at least exacerbate, the social core symptoms of ASD. This early emergence may lead to touch being a potential marker and treatment target in neurodevelopmental conditions, with potential cascading effects on social processing. The nature of these difficulties is also increasingly being understood. Several lines of evidence, both in humans and in animal models, point to an imbalance of cortical excitation and inhibition [92]. Such an imbalance could lead to poor perception, co-occurring seizures, clinical hyper- and hypo-reactivity, and the core social symptoms of autism [40, 41, 58, 93, 94]. While it currently remains unclear how this imbalance leads to altered perception, theories include that of poorly defined priors [95, 96], inefficient sensory coding [97], poor thalamocortical sensory gating [40, 61, 98, 99], reduced habituation [13, 37, 57, 100], or poor internalization. Similarly, in ADHD, tactile difficulties have been associated with altered inhibitory function [24, 25, 101, 102] which is core to ADHD at a behavioral and cortical level. Poor or inconsistent sensory coding or sensory gating may in fact lead to difficulty with attentional control.

3.8

One strong benefit of studying touch is that, especially compared to social function, attention, or hyper-activity, it is well-delineated and can be studied both quantitatively and qualitatively at a number of levels of investigation, using a wide range of approaches. This facilitates linking the neurobiology (mostly focused on excitation-

Future Work

240

Nicolaas A. J. Puts and Carissa J. Cascio

inhibition balance) with the perceptual and clinical domains such that direct inferences can be made as to how difficulties in brain processing cause the core real-life features of these conditions. As such, tactile processing has strong potential as a biomarker for altered neurodevelopment, with implications for stratification, diagnosis, and treatment. With optimized methods and larger samples, differences in tactile function may more readily capture the heterogeneity both across and within neurodevelopmental conditions.

4

Notes Performing sensory testing in neurodevelopment comes with several important caveats. Often these are driven by the nature of the populations studied and should be taken into consideration when interpreting the findings, rather than trying to address these.

4.1

Populations

The nature of clinical neurodevelopmental populations needs to be taken into account. Particularly among psychophysical, social, or affective and MRI approaches, there exist selection biases favoring those participants who can partake in cognitive experiments, understand instructions, and have sufficient attention to sit through such experiments. While questionnaire approaches are limited in providing a detailed understanding of somatosensory function itself, they at least allow for an inventory of children who are perhaps more affected by their developmental condition, such as those with autism and intellectual disability. It is important to recognize that these heavily-affected populations are widely understudied, despite often reporting the most severe difficulties in sensory processing. This leads to the next consideration, which is cohort heterogeneity; in autism alone, it is thought that 78% of children with autism have at least one co-occurring psychiatric condition [86], with ADHD and anxiety being the most common. Given that sensory difficulties are core to both ADHD and anxiety itself, teasing out the specificity and generalizability of these sensory difficulties is difficult [10]. This also leads to smaller study samples. Recruitment, particularly of people with rarer conditions, is difficult and time-consuming, and thus studies may often have few participants to report on. Recent advice, also per the National Institutes of Health and the Research Domain Criteria initiative, strongly suggests studying these difficulties trans-diagnostically, and to examine, for example, sensory difficulties separate from the primary diagnosis. Many of these populations also suffer from a strong sex bias; for example, autism, ADHD, and Tourette Syndrome are more commonly diagnosed in boys than girls, and thus there has been limited study of sex effects [72, 87, 88]. Finally, many children

Touch Processing in Atypical Development

241

are under medication. Some studies remove children from medication prior to testing, whereas others keep them on medication. The impact of medication should be taken into account when possible. 4.2

Questionnaires

In addition to the subjective nature of questionnaire approaches, there are several caveats to using them. From a practical perspective, they often require purchase (although free ones exist as well), and many require normalization and standardization, especially when used in cultures or languages different from the normalization sample. Therefore, these questionnaire approaches may need substantial validation prior to their clinical use. A large variety of approaches exist, and some continue to be updated and revised, so when using them, researchers should check which version is being used, and whether there are any revisions. Please also ensure you check whether you need to contact the developers or whether the materials require purchase. Manuscripts have been rejected or even retracted due to the unlawful use of questionnaires [89], although this example was ultimately resolved. Given that caregivers or teachers are often asked to answer many questionnaires or scales on behalf of children as part of their participation, these approaches can often seem burdensome. The benefit of standardized questionnaires, however, is that clinical scores can be obtained which can be compared easily across different research sites (due to their standardization). Many of these can be applied to a very young age and for children severely affected by a particular condition. Theoretically, one limitation of these approaches is that there is often a ‘similarity’ or ‘memory’ component to these questionnaires. Caregivers who rate one item as severe are more likely to rate other items as severe, and in addition, their answers may be based on a historical assessment of sensory difficulties. Conversely, caregiver responses can also show a recency effect, such that completing the questionnaire on a particularly “bad day,” can inflate scores. Another limitation of questionnaires is that they often do not differentiate between what could be considered ‘perception’ and reactivity (unless explicitly so, e.g., the SPQ). Many questions involve a reactive or responsive component such as “I generally avoid social situations in which I can be touched”. The perceptual equivalent would be “I generally feel it when someone touches me or not”. This is further muddled by generalizing across hyperreactivity and hyper-sensitivity in the literature; one could argue that reactivity refers to the emotional response to sensory information, whereas sensitivity should refer to perceptual components such as detection or discrimination [70]. Questionnaires often generalize across social and non-social items. In our opinion, reactivity and sensitivity refer to different constructs, both of interest, but with different underlying mechanisms. This is further muddled by a variety of approaches to encode hyper- and hypo-reactivity.

242

Nicolaas A. J. Puts and Carissa J. Cascio

Certain questionnaires (e.g., the AASP) code hyper- and hyporeactivity as a single continuum based on a combination of items, whereas in recent work we showed that hyper- and hypo-reactivity were strongly and positively correlated [12], such that the idea of a ‘hyper-reactive’ phenotype seems not to exist and the condition should be examined more as a ‘lack of modulation’ of sensory input. 4.3 ObservationBased Approaches

The limitations of questionnaire and dimensional rating approaches apply also to observational approaches. Furthermore, these approaches require even more standardization. Because these approaches are often validated and normalized, it is pertinent that the setting, script, and assessment of these approaches is as consistent as possible. It often requires multiple ratings and substantial training. While it is perhaps less applicable to more clinically affected populations than questionnaires and dimensional scales, these approaches can be used in young and affected cohorts as well. They also have the benefit of being a direct observation of a child’s behavior, which is then coded, rather than the subjective historical perspective of a caretaker or teacher. However, the necessity to ‘visit’ significantly impacts those less willing to travel or access the services.

4.4 Quantitative Sensory Approaches

While more robust and less subjective than observational and questionnaire approaches, QST approaches also suffer from the same limitations as detailed above, and need validation and purchase of equipment. QST relies on small numbers of trials and the quantification of the measures is not as structured as psychophysical data, is based on normative data, and often has components beyond sensory, tactile processing alone, for example, there is often a communicative component to them.

4.5 Psychophysical Approaches

Although psychophysics likely allows the most objective measure of ‘pure’ tactile perception, psychophysical approaches come with several caveats. Beyond the different types of approaches such as in flutter or vibration perception, different types of stimulators (which require purchase), and different methods (MCS, 2AFC) which are discussed elsewhere, caution should be taken interpreting these approaches in neurodevelopmental populations. While historically, psychophysical approaches have used numerous trials, and often multiple blocks, these are often not feasible in pediatric cohorts, and especially not in those with clinical conditions. Attention in particular is a source of bias in these populations [39] and reaction time variability plays an important role too [21]. Staircase approaches are often the fastest, with Bayesian tracking approaches such as QUEST allowing for the estimation of threshold in a small number of trials. However, performance on these tasks is highly dependent on previous trials, and thus lapses in

Touch Processing in Atypical Development

243

Fig. 2 Example staircase tracking. Despite a small number of trials, Participant 1 reaches a robust threshold (also, see [18]). Participant 2 however, starts the same but lost attention making estimation of threshold problematic. Blue dots indicate trials where a correct answer was made, red dots indicate trials where a mistake was made

attention may substantially affect performance. We previously showed that, in healthy adults, there is no significant effect of block length (20 vs. 40 trials) on tactile psychophysical tasks using a 2-down-1-up staircase approach [90]. In children, however, we often see the effect shown in Fig. 2, where children stop paying attention after several trials, and performance gets worse. Visual inspection of the data is key, as is estimation of attention, response times, and differences in the response criteria [20]. Increased numbers of trials may therefore negatively impact the estimation of threshold in children. That said, in vision, Bayesian staircase approaches show that these may be primarily impacted by attention [91]. Thus, it is important to estimate trial-by-trial variability, reaction time, and if possible, to provide additional details on performance and attention. While MCS approaches do not suffer as much from this trialby-trial dependence, they often require a large number of trials and therefore are also not well-suited. When studying conditions affected by attention, ensuring the data aren’t driven by simple lapses in attention is important. This is especially the case when trials become difficult (e.g., in detection tasks) and participants may get bored, frustrated, or even aggressive. It is often more useful to give participants breaks and use shorter tasks than to continue with substantial testing. 4.6 Facilitating Pediatric Testing

Recent studies have started to incorporate story-driven components where the tactile tasks are set within stories for the participants to follow. This often gives participants motivation (which in itself may differ between clinical conditions) and while these may contain sensory components beyond that of touch, their benefits outweigh their downsides.

244

Nicolaas A. J. Puts and Carissa J. Cascio

There has been limited attention to noise, temperature, and humidity control for the vast majority of studies (or this is not reported, see also Chapter 9, this volume). Dedicated testing rooms are useful, but there is once more a trade-off between dedicated testing rooms without distraction and those which are child appropriate. Sometimes, children may want a known adult in the room which adds an additional component. Another key aspect to keep in mind is that, given that these populations often suffer from sensory hyper-reactivities, the tactile stimulation itself may lead to emotional or adverse responses. While many children these days are used to perceiving vibrations (e.g., through phones or video games) these stimuli often need a careful introduction. Finally, the lack of ecological validity of psychophysics is a limitation. We are rarely asked to precisely assess the properties of single modality stimuli in the real world. Converging methods may help to link everyday perception to both its neural underpinnings and clinical impact. In summary psychophysical testing in pediatric and neurodevelopmental conditions is possible, but requires an important trade-off between task length and task engagement. Several caveats should be considered, including attention and hyper-sensitivities. It is important to consider, but not necessarily to minimize these, since attentional and sensory difficulties are often core to the conditions studied. 4.7 Social and Affective Touch

A major consideration for investigating social touch is in how to operationalize it as a construct. While a growing literature has operationalized it physiologically and neuroanatomically by focusing on CT-afferents and gentle stroking touch, other studies have taken a broader view that includes hand holding, tapping, squeezing, and other forms of touch that can carry social importance. While there is no single “correct” way to operationalize the construct, it is important that researchers are clear in how they are defining it. Another consideration for researching social touch is the tradeoff between stimulus control and ecological validity in the mode of stimulus delivery. One potential bias is also whether differences in lower- and higher-level perception (this can be either the discriminatory capacity or the ability to apply valence to a stimulus) of touch in autism are due, for example, to difficulties in interoception or imagination and thus affect an individual’s ability to reliably assess their response. Some studies delivered CT-targeted touch using automated stimulators that precisely control speed and pressure, but may not be truly “social.” Others have opted for manually-delivered stimuli that are less controlled in speed and/or pressure, but introduce another person into the experience and thus are more social in the conventional sense of the word. Establishing a clear difference between ‘affective’ touch and social touch is important moving forward.

Touch Processing in Atypical Development

4.8 Imaging Approaches

245

Similar to psychophysics and social touch, participant selection bias is problematic with imaging approaches. Particularly in task-based approaches, selection of participants is often tailored to those who ‘scan well’ and are of typical to high intelligence. Even within these cohorts, it is important to take scan quality into account. Those with attentional issues are likely to move around more, and it is well-known that movement impairs scan quality and can affect outcomes (see Chapter 18, this volume). Passive, or less-invasive approaches, such as tactile EEG, may be better suited for these more difficult populations, and have been used successfully in children with severe intellectual disability as well as in infants (Chapter 19, this volume).

References 1. Reissland N, Francis B, Aydin E, Mason J, Schaal B (2014) The development of anticipation in the fetus: a longitudinal account of human fetal mouth movements in reaction to and anticipation of touch. Dev Psychobiol 56(5):955–963. https://doi.org/10.1002/ dev.21172 2. Maurer D, Maurer C (1988) The world of the Newborn. Basic Books, New York 3. Montagu A (1978) Touching : the human significance of the skin. Harper & Row, New York 4. White JL, Labarba RC (1976) The effects of tactile and kinesthetic stimulation on neonatal development in the premature infant. Dev Psychobiol 9(6):569–577. https://doi.org/ 10.1002/dev.420090610 5. Cascio CJ, Moore D, McGlone F (2019) Social touch and human development. Dev Cogn Neurosci 35:5–11. https://doi.org/ 10.1016/j.dcn.2018.04.009 6. Gliga T, Farroni T, Cascio CJ (2019) Social touch: A new vista for developmental cognitive neuroscience? Dev Cogn Neurosci 35:1– 4. https://doi.org/10.1016/j.dcn.2018. 05.006 7. Pellegrini AD, Smith PK (1998) Physical activity play: the nature and function of a neglected aspect of play. Child Dev 69(3): 577–598. https://doi.org/10.1111/j. 1467-8624.1998.tb06226.x 8. Little LM, Dean E, Tomchek S, Dunn W (2018) Sensory processing patterns in autism, attention deficit hyperactivity disorder, and typical development. Phys Occup Ther Pediatr 38(3):243–254. https://doi.org/10. 1080/01942638.2017.1390809

9. Dunn W (1997) The impact of sensory processing abilities on the daily lives of young children and their families: a conceptual model. Infants Young Child 9(4):23–35 10. Uljarevic´ M, Baranek G, Vivanti G, Hedley D, Hudry K, Lane A (2017) Heterogeneity of sensory features in autism spectrum disorder: challenges and perspectives for future research. Autism Res 10(5):703–710. https://doi.org/10.1002/aur.1747 11. Dunn W, Griffith JW, Sabata D, Morrison MT, MacDermid JC, Darragh A et al (2015) Measuring change in somatosensation across the lifespan. Am J Occup Ther 69(3): 6903290020. https://doi.org/10.5014/ ajot.2015.014845 12. He JL, Wodka E, Tommerdahl M, Edden RAE, Mikkelsen M, Mostofsky SH et al (2021) Disorder-specific alterations of tactile sensitivity in neurodevelopmental disorders. Commun Biol 4(1):97. https://doi.org/10. 1038/s42003-020-01592-y 13. Puts NAJ, Wodka EL, Tommerdahl M, Mostofsky SH, Edden RAE (2014) Impaired tactile processing in children with autism spectrum disorder. J Neurophysiol 111(9): 1803–1811. https://doi.org/10.1152/jn. 00890.2013 14. Tavassoli T, Bellesheim K, Tommerdahl M, Holden JM, Kolevzon A, Buxbaum JD Altered tactile processing in children with autism spectrum disorder. Autism Res 9(6): 616–620. https://doi.org/10.1002/aur. 1563 15. Blakemore SJ, Tavassoli T, Calo S, Thomas RM, Catmur C, Frith U et al (2006) Tactile sensitivity in Asperger syndrome. Brain Cogn

246

Nicolaas A. J. Puts and Carissa J. Cascio

61(1):5–13. https://doi.org/10.1016/j. bandc.2005.12.013 16. Duerden EG, Taylor MJ, Lee M, McGrath PA, Davis KD, Roberts SW Decreased sensitivity to thermal stimuli in adolescents with autism spectrum disorder: relation to symptomatology and cognitive ability. J Pain 16(5): 463–471. https://doi.org/10.1016/J. JPAIN.2015.02.001 17. Riquelme I, Hatem SM, Montoya P (2016) Abnormal pressure pain, touch sensitivity, proprioception, and manual dexterity in children with autism spectrum disorders. Neural Plast:1723401. https://doi.org/10.1155/ 2016/1723401 18. Mikkelsen M, Wodka EL, Mostofsky SH, Puts NAJ (2018) Autism spectrum disorder in the scope of tactile processing. Dev Cogn Neurosci 29:140–150. https://doi.org/10. 1016/j.dcn.2016.12.005 19. Fru¨ndt O, Grashorn W, Scho¨ttle D, Peiker I, David N, Engel AK et al (2017) Quantitative sensory testing in adults with autism spectrum disorders. J Autism Dev Disord 47(4): 1183–1192. https://doi.org/10.1016/j. dcn.2016.12.005 20. Quinde-Zlibut JM, Okitondo CD, Williams ZJ, Weitlauf A, Mash LE, Heflin BH et al (2020) Elevated thresholds for light touch in children with autism reflect more conservative perceptual decision-making rather than a sensory deficit. Front Hum Neurosci 14:122. https://doi.org/10.3389/FNHUM.2020. 00122 21. Williams ZJ, Failla MD, Davis SL, Heflin BH, Okitondo CD, Moore DJ et al (2019) Thermal perceptual thresholds are typical in autism spectrum disorder but strongly related to intra-individual response variability. Sci Rep 9(1):12595. https://doi.org/10.1038/ S41598-019-49103-2 22. Kaiser MD, Yang DYJ, Voos AC, Bennett RH, Gordon I, Pretzsch C et al (2016) Brain mechanisms for processing affective (and nonaffective) touch are atypical in autism. Cereb Cortex 26(6):2705–2714. https://doi.org/ 10.1093/CERCOR/BHV125 23. Perini I, Gustafsson PA, Igelstro¨m K, Jasiunaite-Jokubaviciene B, K€ampe R, Mayo LM et al (2021) Altered relationship between subjective perception and central representation of touch hedonics in adolescents with autism-spectrum disorder. Transl Psychiatry 11(1). https://doi.org/10.1038/S41398021-01341-7 24. Cascio CJ (2010) Somatosensory processing in neurodevelopmental disorders. J Neurodev

Disord 2(2):62–69. https://doi.org/10. 1007/s11689-010-9046-3 25. Ghanizadeh A (2008) Tactile sensory dysfunction in children with ADHD. Behav Neurol 20(3):107–112. https://doi.org/10. 3233/BEN-2008-0221 26. Puts NAJ, Harris AD, Crocetti D, Nettles C, Singer HS, Tommerdahl M et al (2015) Reduced GABAergic inhibition and abnormal sensory symptoms in children with Tourette syndrome. J Neurophysiol 114(2):808–817. https://doi.org/10.1152/jn.00060.2015 27. Tinaz S, Belluscio BA, Malone P, van der Veen JW, Hallett M, Horovitz SG (2014) Role of the sensorimotor cortex in Tourette syndrome using multimodal imaging. Hum Brain Mapp 35(12):5834–5846. https://doi. org/10.1002/hbm.22588 28. Isaacs D, Key AP, Cascio CJ, Conley AC, Walker HC, Wallace MT et al (2020) Sensory hypersensitivity severity and association with obsessive-compulsive symptoms in adults with tic disorder. Neuropsychiatr Dis Treat 16:2591–2601. https://doi.org/10.2147/ NDT.S274165 29. Isaacs D, Key AP, Cascio CJ, Conley AC, Riordan H, Walker HC et al (2022) Crossdisorder comparison of sensory overresponsivity in chronic tic disorders and obsessive-compulsive disorder. Compr Psychiatry 113:152291. https://doi.org/10. 1016/j.comppsych.2021.152291 30. Chang YS, Owen JP, Desai SS, Hill SS, Arnett AB, Harris J et al (2014) Autism and sensory processing disorders: shared white matter disruption in sensory pathways but divergent connectivity in social-emotional pathways. PLoS One 9(7):e103038. https://doi.org/ 10.1371/journal.pone.0103038 31. Tavassoli T, Brandes-Aitken A, Chu R, Porter L, Schoen S, Miller LJ et al (2019) Sensory over-responsivity: parent report, direct assessment measures, and neural architecture. Mol Autism 10(1):4. https://doi. org/10.1186/S13229-019-0255-7 32. Lane AE, Young RL, Baker AEZ, Angley MT (2010) Sensory processing subtypes in autism: association with adaptive behavior. J Autism Dev Disord 40(1):112–122. https:// doi.org/10.1007/S10803-009-0840-2 33. Cascio C, McGlone F, Folger S, Tannan V, Baranek G, Pelphrey KA et al (2008) Tactile perception in adults with autism: a multidimensional psychophysical study. J Autism Dev Disord 38(1):127–137. https://doi.org/10. 1007/s10803-007-0370-8

Touch Processing in Atypical Development 34. Tommerdahl M, Tannan V, Holden JK, Baranek GT (2008) Absence of stimulus-driven synchronization effects on sensory perception in autism: evidence for local underconnectivity? Behav Brain Funct 4:19. https://doi.org/ 10.1186/1744-9081-4-19 35. Tommerdahl M, Tannan V, Cascio CJ, Baranek GT, Whitsel BL (2007) Vibrotactile adaptation fails to enhance spatial localization in adults with autism. Brain Res 1154:116–123. https://doi.org/10.1016/j.brainres.2007. 04.032 36. Haigh SM, Minshew N, Heeger DJ, Dinstein I, Behrmann M (2016) Overresponsiveness and greater variability in roughness perception in autism. Autism Res 9(3):393–402. https://doi.org/10.1002/ aur.1505s 37. McKernan EP, Wu Y, Russo N (2020) Sensory overresponsivity as a predictor of amplitude discrimination performance in youth with ASD. J Autism Dev Disord 50(9): 3140–3148. https://doi.org/10.1007/ S10803-019-04013-0 38. Bryant LK, Woynaroski TG, Wallace MT, Cascio CJ (2019) Self-reported sensory hypersensitivity moderates association between tactile psychophysical performance and autism-related traits in neurotypical adults. J Autism Dev Disord 49(8): 3159–3172. https://doi.org/10.1007/ s10803-019-04043-8 39. Wodka EL, Puts NAJ, Mahone EM, Edden RAE, Tommerdahl M, Mostofsky SH (2016) The role of attention in somatosensory processing: a multi-trait, multi-method analysis. J Autism Dev Disord 46(10):3232–3241. https://doi.org/10.1007/s10803-0162866-6 40. He JL, Oeltzschner G, Mikkelsen M, Deronda A, Harris AD, Crocetti D et al (2021) Region-specific elevations of glutamate + glutamine correlate with the sensory symptoms of autism spectrum disorders. Transl Psychiatry 11(1):411. https://doi. org/10.1038/S41398-021-01525-1 41. Sapey-Triomphe L-A, Lamberton F, Sonie´ S, Mattout J, Schmitz C (2019) Tactile hypersensitivity and GABA concentration in the sensorimotor cortex of adults with autism. Autism Res 12(4):562–575. https://doi. org/10.1002/aur.2073 42. Morrison I, Lo¨ken LS, Olausson H (2010) The skin as a social organ. Exp Brain Res 204(3):305–314. https://doi.org/10.1007/ S00221-009-2007-Y 43. Olausson H, Lamarre Y, Backlund H, Morin C, Wallin BG, Starck G et al (2002)

247

Unmyelinated tactile afferents signal touch and project to insular cortex. Nat Neurosci 5(9):900–904. https://doi.org/10.1038/ NN896 44. Morrison I (2012) CT afferents. Curr Biol 22(3):R77–R78. https://doi.org/10.1016/ j.cub.2011.11.032 45. Cascio CJ, Moana-Filho EJ, Guest S, Nebel MB, Weisner J, Baranek GT et al (2012) Perceptual and neural response to affective tactile texture stimulation in adults with autism spectrum disorders. Autism Res 5(4):231–244. https://doi.org/10.1002/aur.1224 46. Bjo¨rnsdotter M, Morrison I, Olausson H (2010) Feeling good: on the role of C fiber mediated touch in interoception. Exp Brain Res 207(3–4):149–155. https://doi.org/10. 1007/S00221-010-2408-Y 47. Cabibihan JJ, Joshi D, Srinivasa YM, Chan MA, Muruganantham A (2015) Illusory sense of human touch from a warm and soft artificial hand. IEEE Trans Neural Syst Rehabil Eng 23(3):517–527. https://doi.org/10. 1109/TNSRE.2014.2360533 48. Duhn L (2010) The importance of touch in the development of attachment. Adv Neonatal Care 10(6):294–300. https://doi.org/10. 1097/ANC.0B013E3181FD2263 49. Carozza S, Leong V (2021) The role of affectionate Caregiver touch in early neurodevelopment and parent-infant interactional synchrony. Front Neurosci 14:613378. https://doi.org/10.3389/FNINS.2020. 613378 50. Malik F, Marwaha R Developmental stages of social emotional development in children. StatPearls 2021 Jun 10 [cited 2021 Dec 6]; Available from: https://www.ncbi.nlm.nih. gov/books/NBK534819/ 51. Kadlaskar G, Seidl A, Tager-Flusberg H, Nelson CA, Keehn B (2020) Caregiver touchspeech communication and infant responses in 12-month-olds at high risk for autism spectrum disorder. J Autism Dev Disord 50(3): 1064–1072. https://doi.org/10.1007/ S10803-019-04310-8 52. Provenzi L, Rosa E, Visintin E, Mascheroni E, Guida E, Cavallini A et al (2020) Understanding the role and function of maternal touch in children with neurodevelopmental disabilities. Infant Behav Dev 58:101420. https:// doi.org/10.1016/J.INFBEH.2020.101420 53. Baranek GT, Berkson G (1994) Tactile defensiveness in children with developmental disabilities: responsiveness and habituation. J Autism Dev Disord 24(4):457–471. https:// doi.org/10.1007/BF02172128

248

Nicolaas A. J. Puts and Carissa J. Cascio

54. Cascio CJ, Lorenzi J, Baranek GT (2016) Self-reported pleasantness ratings and examiner-coded defensiveness in response to touch in children with ASD: effects of stimulus material and bodily location. J Autism Dev Disord 46(5):1528–1537. https://doi.org/ 10.1007/s10803-013-1961-1 55. Green SA, Hernandez L, Tottenham N, Krasileva K, Bookheimer SY, Dapretto M (2015) Neurobiology of sensory overresponsivity in youth with autism spectrum disorders. JAMA Psychiat 72(8):778–786. h t t p s : // d o i . o r g / 1 0 . 1 0 0 1 / JAMAPSYCHIATRY.2015.0737 56. Lee Masson H, Op de Beeck H, Boets B (2020) Reduced task-dependent modulation of functional network architecture for positive versus negative affective touch processing in autism spectrum disorders. NeuroImage 219: 117009. https://doi.org/10.1016/J. NEUROIMAGE.2020.117009 57. Green SA, Hernandez L, Lawrence KE, Liu J, Tsang T, Yeargin J et al (2019) Distinct patterns of neural habituation and generalization in children and adolescents with autism with low and high sensory overresponsivity. Am J Psychiatry 176(12):1010–1020. https://doi. org/10.1176/APPI.AJP.2019.18121333 58. Wood ET, Cummings KK, Jung J, Patterson G, Okada N, Guo J et al (2021) Sensory over-responsivity is related to GABAergic inhibition in thalamocortical circuits. Transl Psychiatry 11(1):39. https://doi. org/10.1038/s41398-020-01154-0 59. Green SA, Hernandez L, Bookheimer SY, Dapretto M (2017) Reduced modulation of thalamocortical connectivity during exposure to sensory stimuli in ASD. Autism Res 10(5): 801–809. https://doi.org/10.1002/AUR. 1726 60. Cascio CJ, Gu C, Schauder KB, Key AP, Yoder P (2015) Somatosensory event-related potentials and association with tactile behavioral responsiveness patterns in children with ASD. Brain Topogr 28(6):895–903. https:// doi.org/10.1007/S10548-015-0439-1 61. Piccardi ES, Begum Ali J, Jones EJH, Mason L, Charman T, Johnson MH et al (2021) Behavioural and neural markers of tactile sensory processing in infants at elevated likelihood of autism spectrum disorder and/or attention deficit hyperactivity disorder. J Neurodev Disord 13(1):1. https://doi. org/10.1186/s11689-020-09334-1 62. Failla MD, Peters BR, Karbasforoushan H, Foss-Feig JH, Schauder KB, Heflin BH et al (2017) Intrainsular connectivity and somatosensory responsiveness in young children with

ASD. Mol Autism 8(1). https://doi.org/10. 1186/s13229-017-0143-y 63. Pryweller JR, Schauder KB, Anderson AW, Heacock JL, Foss-Feig JH, Newsom CR et al (2014) White matter correlates of sensory processing in autism spectrum disorders. Neuroimage Clin 6:379–387. https://doi. org/10.1016/j.nicl.2014.09.018 64. Marshall PJ, Meltzoff AN (2020) Body maps in the infant brain: implications for neurodevelopmental disabilities. Dev Med Child Neurol 62(7):778–783. https://doi.org/10. 1111/DMCN.14540 65. Williams ZJ, Failla MD, Gotham KO, Woynaroski TG, Cascio C (2018) Psychometric evaluation of the short sensory profile in youth with autism spectrum disorder. J Autism Dev Disord 48(12):4231. https:// doi.org/10.1007/S10803-018-3678-7 66. Siper PM, Kolevzon A, Wang AT, Buxbaum JD, Tavassoli T (2017) A clinicianadministered observation and corresponding caregiver interview capturing DSM-5 sensory reactivity symptoms in children with ASD. Autism Res 10(6):1133–1140. https://doi. org/10.1002/aur.1750 67. Taneja P, Olausson H, Trulsson M, Svensson P, Baad-Hansen L (2021) Defining pleasant touch stimuli: a systematic review and meta-analysis. Psychol Res 85(1):20–35. https://doi.org/10.1007/S00426-01901253-8 68. Chou TS, Bucci LD, Krichmar JL (2015) Learning touch preferences with a tactile robot using dopamine modulated STDP in a model of insular cortex. Front Neurorobot 9: 6. https://doi.org/10.3389/FNBOT.2015. 00006 69. Tomchek SD, Dunn W (2007) Sensory processing in children with and without autism: a comparative study using the short sensory profile. Am J Occup Ther 61(2):190–200. https://doi.org/10.5014/AJOT.61.2.190 70. Brown C, Tollefson N, Dunn W, Cromwell R, Filion D (2001) The Adult Sensory Profile: measuring patterns of sensory processing. Am J Occup Ther 55(1):75–82. https://doi.org/ 10.5014/AJOT.55.1.75 71. Tavassoli T, Hoekstra RA, Baron-Cohen S (2014) The Sensory Perception Quotient (SPQ): development and validation of a new sensory questionnaire for adults with and without autism. Mol Autism 5:29. https:// doi.org/10.1186/2040-2392-5-29 72. Taylor E, Holt R, Tavassoli T, Ashwin C, Baron-Cohen S (2020) Revised scored Sensory Perception Quotient reveals sensory

Touch Processing in Atypical Development hypersensitivity in women with autism. Mol Autism 11(1). https://doi.org/10.1186/ S13229-019-0289-X. Tinaz S, Belluscio BA 73. Ausderau KK, Baranek GT (2013) Sensory experiences questionnaire. Encycl Autism Spectr Disord:2770–2774. https://doi.org/ 10.1007/978-1-4419-1698-3_1192 74. Ausderau KK, Baranek GT (1999) Sensory Experiences Questionnaire (SEQ). University of North Carolina at Chapel Hill 75. Baranek GT, David FJ, Poe MD, Stone WL, Watson LR (2006) Sensory experiences questionnaire: discriminating sensory features in young children with autism, developmental delays, and typical development. J Child Psychol Psychiatry 47(6):591–601. https:// doi.org/10.1111/j.1469-7610.2005. 01546.x 76. Parham LD, Ecker C, Miller Kuhaneck H, Henry DA, Glennon TJ (2007) Sensory Processing Measure (SPM): manual. Western Psychological Services, Los Angeles 77. Robertson AE, David RSR (2015) The sensory experiences of adults with autism spectrum disorder: a qualitative analysis. Perception 44(5):569–586. https://doi.org/ 10.1068/p7833 78. Schoen SA, Miller LJ, Green KE (2008) Pilot study of the Sensory Over-Responsivity Scales: assessment and inventory. Am J Occup Ther 62(4):393–406. https://doi. org/10.5014/AJOT.62.4.393 79. Schoen SA, Miller LJ, Sullivan J (2016) The development and psychometric properties of the Sensory Processing Scale Inventory: a report measure of sensory modulation. J Intellect Develop Disabil 42(1):12–21. https://doi.org/10.3109/13668250.2016. 1195490 80. Weintraub S, Bauer PJ, Zelazo PD, WallnerAllen K, Dikmen SS, Heaton RK et al (2013) I. NIH toolbox cognition battery (CB): introduction and pediatric data. Monogr Soc Res Child Dev 78(4):1–15. https://doi.org/10. 1111/MONO.12031 81. Dunn W, Griffith JW, Morrison MT, Tanquary J, Sabata D, Victorson D et al (2013) Somatosensation assessment using the NIH toolbox. Neurology 80(11 Supplement 3):S41–S44. https://doi. org/10.1212/WNL.0B013E3182872C54 82. Ayres AJ (1987) The sensory integration and praxis tests. Western Psychological Services, Los Angeles 83. Mailloux Z (1990) An overview of sensory integration and praxis tests. Am J Occup

249

Ther 44(7):589–594. https://doi.org/10. 5014/ajot.44.7.589 84. Mailloux Z, Grady-Dominguez P, Petersen J, Parham LD, Roley SS, Bundy A et al (2021) Evaluation in Ayres Sensory Integration® (EASI) vestibular and proprioceptive tests: construct validity and internal reliability. Am J Occup Ther 75(6):7506205070. https:// doi.org/10.5014/ajot.2021.043166 85. Mailloux Z, Parham LD, Roley SS, Ruzzano L, Schaaf RC (2018) Introduction to the evaluation in Ayres Sensory Integration ® (EASI). Am J Occup Ther 72(1). https:// doi.org/10.5014/AJOT.2018.028241 86. Kerns CM, Rast JE, Shattuck PT (2020) Prevalence and correlates of Caregiver-reported mental health conditions in youth with autism spectrum disorder in the United States. J Clin Psychiatry 82(1). https://doi.org/10.4088/ JCP.20m13242 87. Loomes R, Hull L, Mandy WPL What is the male-to-female ratio in autism spectrum disorder? A systematic review and meta-analysis. J Am Acad Child Adolesc Psychiatr 56(6): 466–474. https://doi.org/10.1016/j.jaac. 2017.03.013 88. Roy M, Prox-Vagedes V, Ohlmeier MD, Dillo W (2015) Beyond childhood: psychiatric comorbidities and social background of adults with Asperger syndrome. Psychiatr Danubina 27(1):50–59 89. Williams ZJ, Gotham KO (2021) RETRACTED ARTICLE: improving the measurement of alexithymia in autistic adults: a psychometric investigation and refinement of the twenty-item Toronto Alexithymia Scale. Mol Autism 12(1):1–23. https://doi. org/10.1186/S13229-021-00427-9 90. Mikkelsen M, He J, Tommerdahl M, Edden RAE, Mostofsky SH, Puts NAJ (2020) Reproducibility of flutter-range vibrotactile detection and discrimination thresholds. Sci Rep 10(1):6528. https://doi.org/10.1038/ s41598-020-63208-z 91. Manning C, Jones PR, Dekker TM, Pellicano E (2018) Psychophysics with children: investigating the effects of attentional lapses on threshold estimates. Atten Percept Psychophys 80(5):1311–1324. https://doi.org/10. 3758/s13414-018-1510-2 92. Rubenstein JL, Merzenich MM (2003) Model of autism: increased ratio of excitation/inhibition in key neural systems. Genes Brain Behav 2(5):255–267. https://doi.org/ 10.1034/j.1601-183x.2003.00037.x

250

Nicolaas A. J. Puts and Carissa J. Cascio

93. Orefice LL, Zimmerman AL, Chirila AM, Sleboda SJ, Head JP, Ginty DD (2016) Peripheral mechanosensory neuron dysfunction underlies tactile and behavioral deficits in Mouse Models of ASDs. Cell 166(2): 299–313. https://doi.org/10.1016/j.cell. 2016.05.033 94. Puts NAJ, Wodka EL, Harris AD, Crocetti D, Tommerdahl M, Mostofsky SH et al (2017) Reduced GABA and altered somatosensory function in children with autism spectrum disorder. Autism Res 10(4):608–619. https://doi.org/10.1002/aur.1691 95. Lawson RP, Mathys C, Rees G (2017) Adults with autism overestimate the volatility of the sensory environment. Nat Neurosci 20(9): 1293–1299. https://doi.org/10.1038/nn. 4615 96. Pellicano E, Burr D (2012) When the world becomes ‘too real’: a Bayesian explanation of autistic perception. Trends Cogn Sci 16(10): 504–510. https://doi.org/10.1016/J.TICS. 2012.08.009 97. Ward J (2019) Individual differences in sensory sensitivity: a synthesizing framework and evidence from normal variation and developmental conditions. Cogn Neurosci 10(3): 139–157. https://doi.org/10.1080/ 17588928.2018.1557131 98. Khan S, Hashmi JA, Mamashli F, Bharadwaj HM, Ganesan S, Michmizos KP et al (2016)

Altered onset response dynamics in somatosensory processing in autism spectrum disorder. Front Neurosci 10(JUN). https://doi. org/10.3389/fnins.2016.00255 99. Khan S, Michmizos K, Tommerdahl M, Ganesan S, Kitzbichler MG, Zetino M et al (2015) Somatosensory cortex functional connectivity abnormalities in autism show opposite trends, depending on direction and spatial scale. Brain 138(5):1394–1409. https://doi. org/10.1093/brain/awv043 100. Rogers SJ, Ozonoff S (2005) Annotation: what do we know about sensory dysfunction in autism? A critical review of the empirical evidence. J Child Psychol Psychiatry Allied Discip 46(12):1255–1268. https://doi.org/ 10.1111/j.1469-7610.2005.01431.x 101. Ben-Sasson A, Soto TW, Heberle AE, Carter AS, Briggs-Gowan MJ (2014) Early and concurrent features of ADHD and sensory overresponsivity symptom clusters. J Atten Disord 21(10):835–845. https://doi.org/10.1177/ 1087054714543495 102. Puts NAJ, Harris AD, Mikkelsen M, Tommerdahl M, Edden RAE, Mostofsky SH (2017) Altered tactile sensitivity in children with Attention-Deficit hyperactivity disorder. J Neurophysiol 118(5):2568–2578. https:// doi.org/10.1152/jn.00087.2017

Chapter 12 Measuring Touch Sensitivity in an Aging Population Aldrin R. Loomes, Roberta Roberts, and Harriet A. Allen Abstract As we age, our touch sensitivity declines due to changes in the skin, receptors, afferent nerve fibers, and central nervous system. Aging also results in changes in the speed and accuracy of movements. Such changes have the potential to impact exploratory strategies used in active touch. Some of the motor changes during exploration may themselves reflect changing somatosensory signals supporting movement execution. At the same time as sensorimotor changes, aging results in cognitive changes spanning perception, attention, memory, and motivation. When studying somatosensation over the lifespan, it is important to design tasks and performance measures that make it possible to determine the underlying sensory basis for age-related changes, while controlling for the cognitive effects of aging. This chapter discusses some of the issues arising when testing older participants. We suggest various practical adjustments to testing somatosensory functions that allow comparisons of performance across different age groups. Key words Sensitivity, Discrimination, Matching, Texture, Softness, Stickiness, Multisensory, Age, Hydration

1

Introduction Aging is associated with declines in sensory, motor, and cognitive abilities [1]. In the sensory domain, it is visual and auditory impairments that are most commonly detected and subsequently corrected using perceptual aids such as glasses and hearing aids. Two-thirds of those over 70 years experience auditory impairment [2–4], while 3.2 million over 40 years experience visual impairment [5, 6]. The percentage of adults experiencing dual sensory impairment increases with age and is reported to be around 11.3% of US adults over 80 years [7]. Sensory declines are often accompanied by, and can be linked to, peripheral changes. For example, aging is associated with hardening of the lens leading to presbyopia [8]. Yellowing and thickening of the lenses lead to decreased visual acuity, color perception, and impaired visual resolution of fine detail [9, 10]. The risk of eye diseases such as glaucoma, cataracts, and macular degeneration are increased with age

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_12, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

251

252

Aldrin R. Loomes et al.

[11]. Peripheral changes affecting hearing are thought to be caused by cochlear degeneration [12] with the main effect being the loss of sensitivity at higher frequencies [13]. While sensitivity loss in vision and hearing are well documented in the aging literature, less is known about age-related changes in touch perception. The absence of routine touch sensitivity screening is likely a contributing factor. Moreover, measurement of touch sensitivity is complicated by the fact that sensory signals can have an important role in control of the movements used during tasks designed to measure touch perception. In our lab, we have found that older adults have shown decreased ability to modulate exploratory contact forces compared with young adults [14]. This is consistent with the known changes in proprioception [15–17] and suggests force control needs to be factored into the measurement of touch perception during active exploration. As in vision and audition, aging effects on somatosensation are thought, to some extent, to be due to peripheral changes. For instance, age-related deterioration in spatial acuity at the fingertips [18, 19] is assumed to be, at least in part, due to peripheral changes observed in aging skin such as mechanoreceptor loss [20–23], and skin changes including thinning of the skin, reduction in hydration and elasticity, thinning of the nail plate, and reduced blood flow [24]. With increased age, there are also changes to the structure of the skin which impact the signals transmitted to the mechanoreceptors. Changes to the skin, such as thinning and reduction in elasticity and hydration [25–28], affect the way the skin interacts with textured surfaces [29, 30]. In addition to age-related mechanoreceptor loss, there is also a demyelination of neurons leading to the disruption of neural signals [31] and loss of peripheral information [32, 33]. One consequence of these differences for researchers is the increased difficulty of designing a single set of stimuli appropriate for testing different age groups. Thus, performance lower than, or deviating from that of the young on any touch task may reflect changes at the periphery, in signal transmission, in early somatosensory processes in the brain, or changes in higher-level cognitive mechanisms. Awareness and careful consideration of the design of stimuli will help to differentiate which effects are due to central and peripheral mechanisms [34]. Cognitive changes related to aging are well documented, and include a decline in working memory, attention, and executive control [35]. When testing an aging population, it is important to design methods where tasks are practiced sufficiently and are suitably straightforward so that a difference in performance reflects a difference in sensitivity, not a more general failure to perform the task. Furthermore, for studies of touch, these generalized differences have other implications. Non-pathological brain atrophy observed in aging has been linked to cognitive changes including impairments in tactile

Measuring Ageing Touch

253

memory, processing speed, and attention [36, 37]. Moreover, aging is also linked to impaired perception associated with brain changes in the somatosensory network beyond the primary somatosensory cortex [38]. In aging, it is sometimes thought that peripheral sensory as well as cognitive changes lead to an increased reliance on multisensory integration [39] to compensate for declining unisensory performance [40, 41]. This change in sensory processing could lead to an increase in cognitive load [1, 42]. Support for the increased use of multisensory cues can be found in studies suggesting that older adults exhibit increased benefit from multisensory cues in perceptual tasks, as compared to younger adults [40, 43]. Furthermore, the ability to ignore conflicting information declines with age [44, 45] giving rise to the inhibitory deficit hypothesis (IDH) [46]. In summary, when studying the effects of aging on the sense of touch, it is important to control the availability of non-touch cues as their effects on behavior are likely to vary with age. In addition, the difficulties experienced by older adults in identifying and selecting relevant information and in ignoring distractors [47, 48] is likely to affect sensory perception, decisionmaking, and the ways in which older adults carry out sensory tasks [49]. There are several open questions about aging and touch perception. First, how does the aging brain cope with the increased cognitive load of multisensory processing? How should this be considered in developing methods? Second, how to tease apart cognitive processing and sensory processing? In this chapter, we will describe the materials and equipment required for typical testing of touch in older people, including environmental factors and skin preparation. We will discuss the recruitment and selection of participants, focusing on important considerations for planning studies. Finally, we will also consider the stimuli that may be used for studies of touch in aging, considering the current knowledge of the tactile sensory systems.

2

Materials Challenges unique to measuring touch perception include the control and/or measurement of both environmental and participant variables such as humidity, room temperature and skin preparation, as well as difficulties producing controlled stimuli (see Chapter 9, this volume). Tactile assessment instruments such as monofilaments can be affected by the relative humidity of the environment [50, 51]. Furthermore, perception itself can be affected by the humidity of the testing environment. Tactile perception has been found to be affected by the hydration [52] and temperature of the skin [53, 54]. Oils and greases on the skin change the friction experienced on contact with surfaces. These issues highlight the

254

Aldrin R. Loomes et al.

Fig. 1 Cutometer® Dual MPA 580 (Courage and Khazaka Electronic GmbH, Cologne, Germany). This device measures the elasticity of the skin using suction and optical measurements. (a) Diagram of the 2 mm probe taking a measurement of elasticity from the participant’s index finger-tip. (b) Diagram of the Cutometer® Dual MPA 580 device with 2 mm and 4 mm probe attachments

importance of controlling or measuring the environment and skin state when assessing tactile perception. Most important, several of these variables are systematically affected by age. Room temperature and humidity can be adjusted by climate control systems. In their absence, hygrometers and air thermometers can be used to keep a record of the environmental conditions. Films on the skin can be removed by preparatory hand washing. Skin hydration, which reduces with age, can be measured directly using a corneometer, and skin elasticity with a cutometer (Fig. 1). Difficulties with monofilaments can be overcome with the use of alternative measures of tactile spatial acuity such as those measuring the ability to discriminate orientation. One such system is JVP domes (Stoelting Instruments; Wood Dale, IL) where square wave gratings cut into convex surfaces are applied onto the surface of the skin (Fig. 2). Volunteers are asked to discriminate the orientation of the applied grating relative to the long axis of the finger. Increased thresholds in older participants require the use of an extended range of grating widths [55]. Figure 3 shows stimuli produced using computer numerical controlled (CNC) cutting. Detailed stimuli on the order of 10 s of microns can be produced using this method. Finely detailed features on surfaces can also be manufactured using laser cutting, but this is a costly and timeconsuming process. Alternatively, etching methods in which a strong acid is used to cut into unprotected parts of metal surfaces

Measuring Ageing Touch

255

Fig. 2 (a) An extended set of 11 JVP domes (Stoelting Co., Wood Dale, IL, USA) with groove widths ranging from 0.35 mm to 6.0 mm. These are placed on the participant’s finger-tip (b) during a Grating Orientation task with the direction of the groove running either horizontally or vertically to the fingertip. The participant must report which way the grating is oriented

Fig. 3 Square-wave polyurethane gratings produced using computer numerical control (CNC) each measuring 30 × 32 mm. (a) represents the fine range of stimuli with (b) representing the course stimuli range

can be used to create small (10 s of microns) features, but the depth of penetration is shallow. 3D printers can be used to create a variety of tactile stimuli. Using this method for stimulus production has advantages over the aforementioned techniques in terms of cost, availability, and ease of use. However, at the time of writing, the spatial precision and resolution of 3-D printers varied more widely than CNC machines and lasers. This is likely to change in the near future. Unlike monofilaments, there is no inbuilt mechanism for reducing or controlling contact force when applying stimuli such as the JVP domes to the skin. The pressing forces (normal to the skin’s

256

Aldrin R. Loomes et al.

surface) used when stimuli contact the skin can be measured by instrumenting the stimuli with force sensors. While the optimal force for applying JVP domes when measuring spatial thresholds in older participants is unknown at present, low contact forces (such as those below 10 g) are likely best avoided. Consistency of force application is desirable, but the extent to which it will affect thresholds in the elderly is unclear [56, 57]. Measurement of contact forces is informative beyond passive touch assessment. Everyday handling and touching is often an active process with exploratory contacts automatically optimized to produce the sensory cues best suited for the evaluation of particular object features [58–60]. Technological developments have made it easier to study these active interactions with objects and surfaces. However, variation in hand dexterity and force application over the lifespan mean that it is important to use robust stimuli and sensors during active exploration, especially those with a higher maximum force capacity (see Note 4.1). Exploratory kinematics during touching can be captured using motion capture systems (Fig. 4). Traditionally, such systems have used optical systems and cameras to capture the positions of reflective markers placed on anatomical landmarks. Drawbacks of these systems include constraints on the number of possible testing environments, cost, and the increased time taken to prepare for

Fig. 4 (a) Experimental set up showing three motion-tracking cameras (Qualisys Motion Capture Systems, Gothenburg, Sweden) focussed on a six-degrees-offreedom force sensory (ATI Nano, NC, USA) housing two grating stimuli. (b) close up of the force sensor and stimuli, including a bridge with tactile guides for blindfolded participants to navigate their way to each texture without triggering the force plate

Measuring Ageing Touch

257

data collection when placing the markers. Preparation time where the volunteer is present is a factor best minimized, particularly with older groups. Markerless motion tracking systems, using algorithms that recognize human poses or skeletons from images, provide an alternative means of motion capture, minimizing preparation time and increasing portability. However, the advantages of these systems have to be weighed against their decreased accuracy and precision relative to marker-based systems. Physical contact with surfaces leads to forces stimulating receptors in the skin and muscles. Signals from these receptors form the basis of our sense of touch. Our haptic system is capable of processing object features across a wide range of spatial scales. These aspects of touch create challenges for making the variety of precisely controlled stimuli required for experimental investigation. Artificial systems such as force-feedback or vibrotactile devices can be used to simulate some aspects of touch; however, both are limited in their recreation of felt objects. Vibrotactors applied to the skin have been used to investigate a range of tactile phenomena from tactile temporal sensitivity [61–64] to some aspects of roughness perception [65, 66] (see Chapters 1, 2, 3, 4, and 5, this volume). Such stimuli are, however, uninformative about active touch conditions. Force-feedback devices are motorized systems which allow users to feel virtual objects by applying reactive forces via handheld devices while the devices are moved. They can be used to simulate dynamic shape, and to some extent texture, though they do not provide a means of stimulating the slowly adapting (SA1) receptors which signal the spatial details of coarsely textured surfaces. In addition to their technical limitations, these devices are unfamiliar to older groups. Furthermore, the greater forces produced by older participants when moving can make testing with force-feedback devices difficult. The majority of stimuli used to investigate touch, particularly active touch, are physical objects. These are increasingly instrumented with force and motion sensors. Commonly available objects or materials can provide useful insights into the somatosensory system. However, when using such stimuli, the experimental design and inferential power is limited by the lack of control over the object’s features. This problem is compounded by age-related decreases in sensitivity, leading to the requirement for a wide variety of stimuli to be able to measure and compare perception in older and younger participants. Difficulties such as these can be addressed through the creation of custom-made experimental objects. Methods for creating stimuli for touch experiments have been described in the preceding text (see Fig. 3 for examples of custom-produced textures).

258

3 3.1

Aldrin R. Loomes et al.

Methods Participants

3.1.1 Age Ranges and Groups

3.1.2 Demographic & Participant Information

The process of defining an age range of interest when studying the sense of touch is often governed by age-related changes in skin properties, which can include thinning of the skin, loss of elasticity, and changes in hydration [25, 33, 67, 68]. Changes in nerveending innovation of the skin [69] and mechanoreceptor loss [23] have also been found in older adults. Although some of these changes have been observed in individuals around 25 years of age [68], researchers agree that tactile aging is reliably detected from the age of 60 years [70]. Therefore, when investigating adult aging, it is commonplace to consider those above 60 years of age to be in the older adult category. It is possible to subdivide people below the age of 60 into different age groups, however, there are often no strong bases for such decisions. For example, Abdouni and colleagues, in a recent study of how skin properties vary with age and sex divided their participants into 4 groups (26 ± 3, 35 ± 3, 45 ± 2 and 58 ± 6 years) [29], while Stevens placed participants in 3 groups (18–33, young adults; 41–63 years, middle-aged adults and 66–91, older adults) [19]. Future research would be considerably strengthened by more standard use of evidence-based motivations for age groupings. In studying human function, it is important to collect accurate and relevant information about participants. In particular, when studying aging touch, it is important to accurately record demographic data as well as patterns of activities such as work, hobbies, and medical conditions affecting the skin, hand, and cognitive function (the latter is covered in more detail in Subheading 3.1.4 below). Well-collected data sets make it possible to more clearly understand the origin of significant changes with age. For example, while females have been shown to have higher tactile spatial acuity than males [57, 71] females also, on average, have a smaller fingertip surface and higher mechanoreceptor density than males [72, 73], accounting for these differences in sensitivity. Similarly, vibrotactile detection thresholds vary with innervation density, being lowest at the most densely innervated sites [74, 75]. While thresholds do not differ between males and females in general [21, 75, 76], sex differences have been noted in older people [75]. As with tactile sensitivity, these differences may be accounted for by mechanoreceptor density or other peripheral differences such as tissue conformation [77]. The thickness of the skin can also vary with activity, as friction over time tends to produce calluses. These examples also highlight the value of collecting information about the physical condition of the skin when evaluating touch perception. For example, hydration and elasticity of the skin on the fingertip are known to differ with age [25, 33, 67, 68], and can be assessed with the use

Measuring Ageing Touch

259

of a cutometer (Fig. 1) to measure elasticity, and a corneometer to measure hydration. 3.1.3 Socioeconomic Background and Life Experience

When thinking about balancing samples, it is very important to be aware of socioeconomic differences in samples of older adult participants. Older adults who have previous experience in science and research are more likely to volunteer to participate in experimental studies, as are those who have other connections to higher education. Specific to studies of touch, differences in employment or skills will have influenced both the skin and their skill and sensitivity. Participants are generally recruited through university advertisements and their details are kept in a database. However, this may bias the sample to a group of older adults who are, or have been, involved in the university in some way, or are members of active local organizations. This may result in a more physically and cognitively active sample than is representative of the aging population as a whole. Older adult participants who have connections to universities and who are members of local organizations are more likely to be of higher socioeconomic status, have had access to higher education, better access to healthcare, and live in less polluted areas. Although research in this area is in its infancy, it is suggested that exposure to air pollution correlates with signs of skin aging [78]. Therefore, older adults who have lived in urban areas or had manual jobs where they are exposed to pollutants could have more age-related skin changes which may influence tactile sensitivity.

3.1.4

During recruitment, it is important to identify any common conditions which could affect the performance of older adults in your study. For example, when researching multisensory perception, it is important to screen for issues that affect touch, audition, and vision. It is common for older adults to require perceptual aids such as hearing aids and glasses due to peripheral changes. With this in mind, it is expected that corrected-to-normal vision and hearing may not necessarily be a problem, unless, for example, the study requires that participants wear headphones or be in a magnetic resonance imaging scanner where these devices cannot be worn. Participants should also be screened for conditions that may affect the sense of touch, including but not limited to, diabetes, multiple sclerosis, Raynaud’s disease, and heart disease. It is unrealistic to expect a representative sample of older adults who do not wear glasses or require hearing aids. If such a sample was created, one might question how representative they are of the wider population. Therefore, some level of perceptual loss must be accounted for in the recruitment and experimental setup. On recruitment forms, for example, researchers should ask whether the participant can wear and use headphones if sound is an important factor in the study. Or, if the participant is only required to hear

Medical Conditions

260

Aldrin R. Loomes et al.

Fig. 5 Purdue Pegboard (Lafayette Instrument Company, IN, USA; Tiffin, 1968) was used to assess participants’ dexterity in both the dominant and non-dominant hands

instructions such as in a scanning study, experimenters should come up with a different way to communicate, for example through written instructions and a button response box. For participants who are hearing impaired, auditory delivery of instructions may increase the difficulty of the task, as they are already working harder to understand the instructions. This may increase cognitive load and result in poorer task performance [79]. Looking further into possible conditions that could confound experimental results, it is important to screen for common conditions associated with aging such as stroke, dementia, Parkinson’s disease, and diabetes. It is important to keep people with these conditions out of the sample when the aim is to study typical aging, as they could all have a potentially confounding effect on the sense of touch and multisensory perception. It may also be important to ensure that participants have adequate hand movement—in aging populations, movement can be affected by arthritis. Participants should be asked if they have a comfortable range of movement, for example at the fingertip, prior to committing to the study. Screening tools such as the Mini Mental State Exam (MMSE), and the Purdue pegboard (Fig. 5) should be used to evaluate cognitive function and manual dexterity respectively. Motion capture equipment could also allow for the detection of abnormal or slowed movements associated with Parkinsonism. 3.2

Preparation

Several experimental and practical considerations are required when working with older adults. Stimuli must be carefully selected with the knowledge that older adults are generally less sensitive than younger adults across all senses [1]. The stimuli must therefore be chosen such that thresholds are likely to be within the range of stimuli for both groups. When investigating texture it is important to keep in mind that fine and coarse textures are modulated through separate channels, as proposed by the duplex theory [80]. The duplex theory indicates that movement has a different role in the perception of coarse and fine textures [81], thus it is important to consider whether participants will be pressing on or

Measuring Ageing Touch

261

moving over the surface, and to carefully control the spatial periods of course and fine textures. Before data collection, it is imperative to involve older people in the experimental design, or run at least two older adult pilot participants through the protocol, as small problems and hurdles that were unforeseen can arise, such as the instructions not being clear enough or simply the task being too difficult. For example, we have found that older adults find it very difficult to judge and control the force they are applying to a surface or texture and, when asked to do so, pay more attention to the applied force than to the task itself. When force control is no longer a requirement, we have found that the applied force is more or less consistent throughout, although it differs for each individual [14]. Staircase methods are a tried and tested way to assess sensory thresholds [82] (see Chapter 1, this volume). They are effective when working with older adults as they reduce the number of trials needed and therefore reduce the testing time for the participant, while establishing a baseline performance level. Much work has been conducted using staircase designs which allow one to manipulate the study design to achieve a threshold at a particular percentage of correct responses. When using such adaptive methods with older adults, it is worth considering working at relatively highperformance levels, unless this conflicts with the research question, for example, where performance is correct 84% of the time. Working close to detection thresholds can be difficult to sustain for long periods in older participants. Aging also leads to a reduction in manual dexterity which, in itself, may affect tactile perception. All young and older adult participants should be assessed on the day of testing to ensure adequate hand movement for the tasks and also to take a general dexterity measure. The Purdue pegboard is a commonly used task to measure fine and gross motor skills [83]. Participants are asked to manipulate small pegs and insert them into a board within a 30 s time frame (Fig. 5). Participants should, at minimum, also complete a short cognitive assessment, such as the MMSE, to ensure there is no cognitive decline beyond what is expected. These tools provide a scoring system which can be used to rule out participants with motor impairment and cognitive decline, respectively. They can also provide a limited insight into the level of cognitive function of participants; however, if the experimenters are interested in individual differences in cognitive function, then this should be measured explicitly using more detailed tools, according to the design of the study. Practical considerations must also be taken into account when working with older adults. Often testing sessions are long and, therefore, it is important to factor in regular breaks during testing such as not to interrupt testing intervals (see Note 4.2). Similarly, it is important to create a relaxed environment, as older adults are

262

Aldrin R. Loomes et al.

unlikely to be familiar with the environment and may have some anxiety regarding the processes. Young adult participants are likely more used to participating in scientific research and are often there as part of a course requirement. Thus, it is necessary to take more time to explain and demonstrate the task with a series of practice trials, where the participant receives feedback on their performance. We have also found it helpful to repeat these practice trials and demonstrations at regular intervals during the course of the session. Ensuring that the participants are comfortable, and are enjoying their experience as much as possible, increases the likelihood of them returning for future studies.

4

Notes

4.1 Measuring Force and Movement

In our laboratory, we are investigating relationships between the forces and kinematics involved in touch exploration and perceptual experiences. Using six degrees of freedom force-torque sensors mounted in metal casings we were able to measure pressing (normal), sliding (tangential), and twisting (torques) forces applied to textured surfaces. Our testing revealed that higher normal, but not tangential, forces were used by older compared with younger volunteers. In our experiment the normal force exerted by some older participants exceeded 7N.

4.2 Duration of Experimental Sessions

Through experience, we have found that older participants prefer to travel to the university for longer studies than many short ones. We have often found it is good to have two experimenters in the room with the participant such that one can be talking through the experiment with the participant while the other sets up—this saves time for what can be quite complicated equipment set-ups, and also makes the participant feel more comfortable as more time can be spent making sure their needs are met.

References 1. Roberts KL, Allen HA (2016) Perception and cognition in the ageing brain: a brief review of the short-and long-term links between perceptual and cognitive decline. Frontiers in aging neuroscience 8:39 2. Davis AC (1989) The prevalence of hearing impairment and reported hearing disability among adults in Great Britain. Int J Epidemiol 18(4):911–917 3. Cruickshanks KJ et al (1998) Prevalence of hearing loss in older adults in Beaver Dam, Wisconsin: the epidemiology of hearing loss study. Am J Epidemiol 148(9):879–886

4. Wilson D et al (1999) The epidemiology of hearing impairment in an Australian adult population. Int J Epidemiol 28(2):247–252 5. Goman AM, Lin FR (2016) Prevalence of hearing loss by severity in the United States. Am J Public Health 106(10):1820–1822 6. Varma R et al (2016) Visual impairment and blindness in adults in the United States: demographic and geographic variations from 2015 to 2050. JAMA Ophthalmol 134(7):802–809 7. Swenor BK et al (2013) The prevalence of concurrent hearing and vision impairment in

Measuring Ageing Touch the United States. JAMA Intern Med 173(4): 312–313 8. Salvi S, Akhtar S, Currie Z (2006) Ageing changes in the eye. Postgrad Med J 82(971): 581–587 9. Said FS, Weale RA (1959) The variation with age of the spectral transmissivity of the living human crystalline lens. Gerontology 3(4): 213–231 10. Ruddock K (1965) The effect of age upon colour vision—II. changes with age in light transmission of the ocular media. Vis Res 5(1–3):47–58 11. Scholl HP, Massof RW, West S (2013) Ophthalmology and the ageing society. Springer 12. Liu X, Yan D (2007) Ageing and hearing loss. J Pathol: A Journal of the Pathological Society of Great Britain and Ireland 211(2):188–197 13. Weinstein B (1994) Age-related hearing loss: how to screen for it, and when to intervene. Geriatrics (Basel, Switzerland) 49(8):40–5; quiz 46 14. Roberts RD, et al Force modulation during tactile perception in an ageing population.. In Preparation 15. Hurley MV, Rees J, Newham DJ (1998) Quadriceps function, proprioceptive acuity and functional performance in healthy young, middleaged and elderly subjects. Age Ageing 27(1): 55–62 16. Landelle C et al (2021) Contribution of muscle proprioception to limb movement perception and proprioceptive decline with ageing. Curr Opin Physio 20:180–185 17. Goble DJ et al (2009) Proprioceptive sensibility in the elderly: degeneration, functional consequences and plastic-adaptive processes. Neurosci Biobehav Rev 33(3):271–278 18. Stevens JC, Patterson MQ (1995) Dimensions of spatial acuity in the touch sense: changes over the life span. Somatosens Mot Res 12(1): 29–47 19. Stevens JC (1992) Aging and spatial acuity of touch. J Gerontol 47(1):P35–P40 20. Bolton CF, Winkelmann R, Dyck PJ (1966) A quantitative study of Meissner’s corpuscles in man. Neurology 16(1):1–1 21. Gescheider GA et al (1994) The effects of aging on information-processing channels in the sense of touch: I. Absolute sensitivity. Somatosens Mot Res 11(4):345–357 22. Cerimele D, Celleno L, Serri F (1990) Physiological changes in ageing skin. Br J Dermatol 122:13–20 23. Garcı´a-Piqueras J et al (2019) Ageing of the somatosensory system at the periphery:

263

age-related changes in cutaneous mechanoreceptors. J Anat 234(6):839–852 24. Fenske NA, Lober CW (1986) Structural and functional changes of normal aging skin. J Am Acad Dermatol 15(4):571–585 25. Gerhardt LC et al (2009) Skin–textile friction and skin elasticity in young and aged persons. Skin Res Technol 15(3):288–298 26. Ryu HS et al (2008) Influence of age and regional differences on skin elasticity as measured by the Cutometer®. Skin Res Technol 14(3):354–358 27. Wilhelm K-P, Cua AB, Maibach HI (1991) Skin aging: effect on transepidermal water loss, stratum corneum hydration, skin surface pH, and casual sebum content. Arch Dermatol 127(12):1806–1809 28. Zhang S, Duan E (2018) Fighting against skin aging: the way from bench to bedside. Cell Transplant 27(5):729–738 29. Abdouni A et al (2017) Biophysical properties of the human finger for touch comprehension: influences of ageing and gender. R Soc Open Sci 4(8):170321 30. Bruce MF (1980) The relation of tactile thresholds to histology in the fingers of elderly people. J Neurol Neurosurg Psychiatry 43(8): 730–734 31. Peters A (2009) The effects of normal aging on myelinated nerve fibers in monkey central nervous system. Front Neuroanat 3:11 32. Decorps J et al (2014) Effect of ageing on tactile transduction processes. Ageing Res Rev 13:90–99 33. McIntyre S et al (2021) The effects of ageing on tactile function in humans. Neuroscience 464:53–58 34. Deflorio D, Di Luca M, Wing AM Skin stiffness and afferent density in the deterioration of tactile spatial acuity with age. J Physiol. Under review 35. Craik FI, Salthouse TA (2011) The handbook of aging and cognition. Psychology Press 36. Cabeza R (2002) Hemispheric asymmetry reduction in older adults: the HAROLD model. Psychol Aging 17(1):85 37. Raz N et al (2005) Regional brain changes in aging healthy adults: general trends, individual differences and modifiers. Cerebral Cortex 15(11):1676–1689 38. Brodoehl S et al (2013) Age-related changes in the somatosensory processing of tactile stimulation—an fMRI study. Behav Brain Res 238:259–264 39. Diaconescu AO, Hasher L, McIntosh AR (2013) Visual dominance and multisensory

264

Aldrin R. Loomes et al.

integration changes with age. NeuroImage 65: 152–166 40. Laurienti PJ et al (2006) Enhanced multisensory integration in older adults. Neurobiol Aging 27(8):1155–1163 41. Peiffer AM et al (2007) Age-related multisensory enhancement in a simple audiovisual detection task. Neuroreport 18(10): 1077–1081 42. Li KZH, Lindenberger U (2002) Relations between aging sensory/sensorimotor and cognitive functions. Neurosci Biobehav Rev 26(7): 777–783 43. Jones SA, Noppeney U (2021) Ageing and multisensory integration: a review of the evidence, and a computational perspective. Cortex 138:1–23 44. Elliott MT, Wing AM, Welchman AE (2011) The effect of ageing on multisensory integration for the control of movement timing. Exp Brain Res 213(2):291 45. Stothart G, Kazanina N (2016) Auditory perception in the aging brain: the role of inhibition and facilitation in early processing. Neurobiol Aging 47:23–34 46. Hasher L, Zacks RT (1988) Working memory, comprehension, and aging: a review and a new view. In: Bower GH (ed) Psychology of learning and motivation. Academic Press, pp 193–225 47. Gazzaley A et al (2008) Age-related top-down suppression deficit in the early stages of cortical visual memory processing. Proc Natl Acad Sci 105(35):13122–13126 48. Vallesi A et al (2009) Age-related differences in processing irrelevant information: evidence from event-related potentials. Neuropsychologia 47(2):577–586 49. Dully J, McGovern DP, O’Connell RG (2018) The impact of natural aging on computational and neural indices of perceptual decision making: a review. Behav Brain Res 355:48–55 50. Haloua MH, Sierevelt I, Theuvenet WJ (2011) Semmes-weinstein monofilaments: influence of temperature, humidity, and age. J Hand Surg Am 36(7):1191–1196 51. Werner MU, Rotbøll-Nielsen P, EllehuusHilmersson C (2011) Humidity affects the performance of von Frey monofilaments. Acta Anaesthesiol Scand 55(5):577–582 52. Skedung L et al (2018) Mechanisms of tactile sensory deterioration amongst the elderly. Sci Rep 8(1):5303 53. Gescheider GA et al (1997) The effects of skin temperature on the detection and discrimination of tactile stimulation. Somatosens Mot Res 14(3):181–188

54. Zhang Z et al (2009) The impact of non-noxious heat on tactile information processing. Brain Res 1302:97–105 55. Libouton X et al (2010) Tactile roughness discrimination threshold is unrelated to tactile spatial acuity. Behav Brain Res 208(2): 473–478 56. Bleyenheuft Y, Thonnard JL (2007) Tactile spatial resolution measured manually: a validation study. Somatosens Mot Res 24(3): 111–114 57. Goldreich D, Kanics IM (2003) Tactile acuity is enhanced in blindness. J Neurosci 23(8): 3439–3445 58. Lederman SJ, Klatzky RL (1987) Hand movements: a window into haptic object recognition. Cogn Psychol 19(3):342–368 59. Klatzky RL, Lederman SJ, Matula DE (1991) Imagined haptic exploration in judgments of object properties. J Exp Psychol Learn Mem Cogn 17(2):314 60. Klatzky RL, Lederman SJ, Balakrishnan J (1991) Task–driven extraction of object contour by human haptics: part 1. Robotica 9(1): 43–51 61. Yamamoto S, Kitazawa S (2001) Reversal of subjective temporal order due to arm crossing. Nat Neurosci 4(7):759–765 62. Shore DI, Spry E, Spence C (2002) Confusing the mind by crossing the hands. Cogn Brain Res 14(1):153–163 63. Roberts RD, Humphreys GW (2008) Task effects on tactile temporal order judgments: when space does and does not matter. J Exp Psychol Hum Percept Perform 34(3):592 64. Badde S, Ro¨der B, Bruns P (2018) Taskirrelevant sounds influence both temporal order and apparent-motion judgments about tactile stimuli applied to crossed and uncrossed hands. Atten Percept Psychophys 80(3): 773–783 65. Hollins SJB, Washburn S, Mark (2001) Vibrotactile adaptation impairs discrimination of fine, but not coarse, textures. Somatosens Mot Res 18(4):253–262 66. Hollins M, Fox A, Bishop C (2000) Imposed vibration influences perceived tactile smoothness. Perception 29(12):1455–1465 67. Montagna W, Carlisle K (1990) Structural changes in ageing skin. Br J Dermatol 122 (s35):61–70 68. Alexander H, Cook T (2006) Variations with age in the mechanical properties of human skin in vivo. J Tissue Viability 16(3):6–11

Measuring Ageing Touch 69. Namer B (2010) Age related changes in human C-fiber function. Neurosci Lett 470(3): 185–187 70. Fromy B et al (2010) Aging-associated sensory neuropathy alters pressure-induced vasodilation in humans. J Investig Dermatol 130(3): 849–855 71. Van Boven RW et al (2000) Tactile spatial resolution in blind Braille readers. Neurology 54(12):2230–2236 72. Peters RM, Hackeman E, Goldreich D (2009) Diminutive digits discern delicate details: fingertip size and the sex difference in tactile spatial acuity. J Neurosci 29(50):15756–15761 73. Dillon YK, Haynes J, Henneberg M (2001) The relationship of the number of Meissner’s corpuscles to dermatoglyphic characters and finger size. J Anatom 199(5):577–584 74. Morioka M, Whitehouse DJ, Griffin MJ (2008) Vibrotactile thresholds at the fingertip, volar forearm, large toe, and heel. Somatosens Mot Res 25(2):101–112 75. Venkatesan L, Barlow SM, Kieweg D (2015) Age-and sex-related changes in vibrotactile sensitivity of hand and face in neurotypical adults. Somatosens Mot Res 32(1):44–50

265

76. Verrillo RT (1979) Comparison of vibrotactile threshold and suprathreshold responses in men and women. Percept Psychophys 26(1):20–24 77. Fruhstorfer H et al (2000) Thickness of the stratum corneum of the volar fingertips. Clin Anat 13(6):429–433 78. Vierko¨tter A et al (2010) Airborne particle exposure and extrinsic skin aging. J Investig Dermatol 130(12):2719–2726 79. Meister H et al (2016) Effects of hearing loss and cognitive load on speech recognition with competing talkers. Front Psychol 7:301 80. Hollins M, Risner SR (2000) Evidence for the duplex theory of tactile texture perception. Percept Psychophys 62(4):695–705 81. Katz D (1989) The world of touch (LE Krueger, Trans.). Rrlbaum.(Original work published 1925), Mahwah 82. Cornsweet TN (1962) The staircase-method in psychophysics. Am J Psychol 75(3):485–491 83. Desrosiers J et al (1995) The Purdue Pegboard Test: normative data for people aged 60 and over. Disabil Rehabil 17(5): 217–224

Chapter 13 Somatosensory Illusions Tatjana Seizova-Cajic´, Regine Zopf, Martin Riemer, and Xaver Fuchs Abstract Illusions arouse interest in the layperson and researcher alike. The layperson learns that perception is fallible, and the researcher wants to better understand mechanisms of perception and implement illusions in various applications. We are accustomed to visual illusions, but less so to their somatosensory sisters, some of which summon fresh amazement bordering on disbelief. What can possibly make us feel that our tongue is split in two, or that a brush touching an artificial hand touches us? How do we study such experiences? This chapter directs the reader to sources describing a variety of somatosensory illusions. We also outline methodological issues in studying them, and describe methods used to study the well-known rubber hand illusion, and the recently described whose hand illusion. Key words Haptic illusions, Tactile illusions, Illusions of touch, Somatosensory illusions, Bodily illusions, Rubber hand illusion, Whose hand illusion, Proprioceptive illusions, Kinesthetic illusions

1

Introduction

1.1 Why Study Illusions?

“Epistemologically, illusions are nonveridical percepts that reveal the processes by which veridical perception mediates our representation of the visual world.” (Spillmann [1], p. 1509)

Naı¨ve realism, the view that our senses give us direct access to reality, is intuitive. Nevertheless, some games children play – described as ‘folk illusions’ [2] – suggest that normal cognitive development includes a realization that our senses are not always trustworthy. In one game, a child lies on the ground, her legs held up, while another child pretends to dig a hole underneath (Fig. 1a). As the legs are then slowly lowered back down, the ground no longer seems to be where it had been – the hole has been dug! Adapting to an unusual leg posture results in an illusory aftereffect. The hole-digging narrative draws the child’s attention to the relevant aspect of the experience and frames it in recognizable terms: it feels as if the hole has been dug (for related illusions and narratives, see pp. 222–3 in [2]). Children learn through such games that, if conditions are right, one can reliably create non-veridical and often

267

268

Tatjana Seizova-Cajic´ et al.

Fig. 1 Example somatosensory illusions. (a) Postural aftereffect: Holding limbs at unusual positions for a minute or so results in postural aftereffects. Top panels illustrate children’s pretend-play of hole ‘digging’ that provides a narrative and an ‘explanation’ for a variant of this illusion. (b) Crossed fingers, split tongue illusion (a version of Aristotle’s illusion, see text) Left panel: Tongue moves sideways or up-and-down between crossed fingers of one hand; Right panel: The observer feels her tongue flanking the fingers from the outside, both sides at once (described in [2], p. 197). (c) Velvet Hand illusion: Observer clasps hands across some hard wires; back-and-forth motion across the wires with palms pressed against each other results in the feeling that skin is extraordinarily soft, or of a soft surface interposed between the palms.

fascinating experiences [2], as the scholarly literature also abundantly documents (a recent compilation of visual illusions [3] has over 100 chapters; for compilations of somatosensory illusions, see below). It is a common understanding that illusions are inaccurate, misleading, or deceptive percepts, but there is no universally accepted formal definition. The concept has “a central core and fuzzy boundaries” ([4], p. 1129) and has been questioned on various grounds, from the idea that the notion of error does not apply to perception, to questioning the idea of one true reality (see [4] for review and discussion of these objections). Illusions can often be described – if not strictly defined – as variations in the perception of the same, unchanged feature with circumstances, where at least one of the variants is in error. Hayward [5] thus proposes an ‘operational’ definition, according to which illusions arise from stimuli combining two separable components: the fixed feature

Somatosensory Illusions

269

the observer attends to, and a variable component that influences the perception of that fixed feature. The variable component can be an external context or internal state. Geometric visual illusions present a classic example of the effect of external context [4], while our earlier example illustrates the effect of an internal state such as a state of adaptation; legs held at 90 degrees prior to extension will feel more extended than they are. It may also happen that different features are perceived as equal [4] (see Note 3.1). Todorovic [4] also proposed that a possibility of veridical perception of a given feature is required for the error in perception to be considered an illusion (I am able to accurately perceive my leg posture). Searching scholarly databases for somatosensory illusions – tactile, haptic, kinesthetic, or proprioceptive (see Note 3.2) – produces thousands of hits. Scholars in philosophy, psychology, neuroscience, medicine, and engineering take a keen interest in illusions both as tools for basic research, and due to their potential applications. Illusions help us understand how the senses function, as two well-known examples illustrate: vibration-induced illusions of limb movement (see Chapter 3, this volume) have settled a long-running controversy about the role of muscle spindles in conscious perception [6, 7], and the rubber hand illusion [8] has allowed body ownership to be studied experimentally. Illusions are also examined and finessed for application in art, communication, virtual reality, sensory augmentation, and substitution, as well as in other areas. One example is the use of tactile motion illusions to increase the capacity for information transmission in simple stimulator arrays [9, 10]; in another example, varying patterns of foot vibration result in illusions of walking on surfaces of different kinds, used in simulated environments [11]. 1.2 Variety of Somatosensory Illusions

One of the best-known tactile illusions is attributed to Aristotle, suggesting that interest in them is more than 2000 years old. In Aristotle’s illusion, also known as diplesthesia or tactile diplopia (the latter term is paradoxical given that diplopia means seeing double images in binocular vision), a single object or edge is perceived as two when touched by the inner sides of crossed fingers (see [12], p. 282 for details, and [13], p. 742 for the ‘reverse Aristotle’ illusion). Its little-known variant, crossed fingers, split tongue illusion [2], is illustrated in Fig. 1b. Feeling that one’s tongue is split in two reveals the malleability of our bodily perception, as does the illusion that our arm moves beyond the anatomical limits of full extension induced by biceps vibration [14], or that our palms are spongy or velvet-like induced by rubbing of clasped hands across wires (the velvet-hand illusion [15], Fig. 1c). All these effects occur within seconds, which may be considered a long time, given that perception usually feels instantaneous, but it is also a rather short time for a bodily experience to undergo a profound change.

270

Tatjana Seizova-Cajic´ et al.

In his classic work on touch, Weber described how perceived distance varies across different parts of the skin, which he studied using a purpose-made compass. “As compasses are moved from the side of the face to the middle of the lips, they seem to grow farther and farther apart; but as they are moved from the middle of the lips to the side of the face, they seem to shrink closer and closer together” (p. 38, [16], see also Chapter 5, this volume). The effect became known as Weber’s illusion [17, 18]. Weber proposed the regional inhomogeneity in innervation to be its cause, making an early attempt at physiological explanation of the illusory effect. Studies of tactile illusions have also been inspired by vision. In 1906, Jaensch investigated a tactile analog of the visual illusion of filled versus unfilled space [19], and in 1934, Re´ve´sz described tactile analogs of the Mu¨ller-Lyer (where perceived length of a line segment varies dramatically with orientation of the arrowheads at its two ends), Opel-Kundt (filled space appears to have greater extent than unfilled space), and other geometric illusions [20]. Another classic author in touch, Katz, described a visual-tactile interaction: “when one cuts into some soft wood with a knife under a strong magnifying glass, the resulting visual enlargement gives rise to the impression that one is cutting deeply into a soft mass, such as cork” (p. 232 in [21]). Many other illusions are described in classical [16, 21, 22] and recent [23, 24] books on touch and the body, in journal articles and online sources. These include illusions of movement of body parts [7, 25–27] and tactile (also known as ‘haptic’) illusions [5, 12, 13, 28]. Some effects also involve vision or audition. The illusions are not necessarily labeled as such, especially in older literature or if they represent adaptation effects. For example, Schilder describes postural aftereffects similar to that described in Fig. 1a as ‘persistence of tone’ ([22], p. 75). There are many ways to classify illusions. Lederman and Jones [12] distinguish illusions of objects and their properties (further divided into material and geometric illusions) and illusions of haptic space (body space; external space). Hayward [13] proposes 15 categories, including, for example, illusions of numerosity or weight, audio tactile interactions, geometric illusions, and distal attribution. We propose one broad distinction based on the aspect of tactile reality that is illusory: illusions of property or pattern versus illusions of source. The velvet hand illusion mentioned earlier is an illusion of property: the softness of the skin is illusory, while the sources of input – the wires, and one’s palms pressed against each other – are correctly perceived as such. In illusions of source, by contrast, we misperceive who or what is touching us. In the rubber hand illusion [8], we feel that the brush touching the artificial hand touches us. Other examples are the Ann Boleyn illusion [29], where we

Somatosensory Illusions

271

see another person’s finger touching wood, but feel that it touches us, and the whose hand illusion (WHI), described further below, in which touching another person is perceived as self-touch. Illusions of source are applied in virtual reality to simulate the tactile presence of external objects. A tactile stimulus is delivered using various devices such as a haptic glove, pin array, or ultrasound (see [28] for overview of devices). As the hand moves freely, tactors are activated in one specific area of external space, creating the illusion of a physical presence of an object occupying the space [30]. Weber described the underlying principle in 1846: variations in tactile sensations contingent on bodily movements lead to external (distal) attribution. He explained that a source of heat near a face will affect different parts of the face when the head turns, whereas “if the source of the heat were in our skin, it would move along with the skin and maintain its position relative to it” (p. 144 in [16]). Once we intentionally create an illusion, or stumble upon one, how do we measure it? We now turn to this topic. 1.3 Illusions and Measurement

“Count what is countable, measure what is measurable and what is not measurable make measurable.” (Galileo Galilei, quoted in Rossi, 2014 [31])

In this quotation, Galileo distinguishes between things measurable and those we need to make so. Phenomenology and psychophysics make perception – be it veridical or illusory – measurable. They concern private phenomena and are in that sense subjective (for other meanings of the subjective-objective dichotomy in psychophysics, see p. 28 in [32]). Subjective measures were only accepted by the scientific community of the 19th century as a ‘proper’ (quantitative) form of measurement when measurement was redefined to include inner comparison, ranking, and magnitude estimation of one’s own subjective states, linked to stimuli that caused them (Chapter 1 in [31]). They have since proven their worth through the replicability of results when different people are exposed to similar conditions of stimulation, and because they helped create a coherent body of knowledge about perception. Experimental studies of illusions also rely on objective measures, such as measurements of brain activity, physiological parameters, or motor action, not to replace subjective measures but to complement them. Phenomenological description plays a crucial part in perception science, although its value is sometimes played down due to its qualitative nature. Researchers elicit descriptions of experience through open-ended or focused questions. Questions with rating scales attached to them allow an ordinal level of measurement. Many other quantitative ways to probe into the relationship between the stimulus and illusory percept entail elaborate psychophysical procedures. They can be classified into ‘performance-based’ and ‘appearance-based’ (Fig. 3.1 in [32]). Procedures directly

272

Tatjana Seizova-Cajic´ et al.

measuring an illusory feature are appearance-based (acknowledging it is difficult to define, Kingdom and Prins describe appearance as “the apparent magnitude of a stimulus dimension”, p 21 in [32]), and include matching, scaling and other procedures, implemented through a forced-choice or other task (an example of the matching procedure is described in Note 3.1). Research on illusions, like any research, is open to methodological threats. One such threat is known as demand characteristics; it is the propensity to accept suggestions or respond to social and other cues that reveal research expectations [33], as recently analyzed in studies of the rubber hand illusion [34, 35]. It may lead to reporting bias, or even influence the percept itself, if stimuli are noisy or ambiguous ([36] and see Note 3.3). To empirically investigate their effect, participants’ expectations need to be uncorrelated – unconfounded – from the sensory stimuli of interest. An apt analogy is placebo control in clinical research [37], but just like controlling the placebo effect is not always possible, it may be very difficult or impossible to experimentally control for demand characteristics, because participants also form expectations based on experimental manipulations themselves [33–35]. Careful wording of instructions and questions is important in any case, and researchers should use a ‘funneling approach’ when asking questions, if possible. In the funneling approach, the experimenter elicits spontaneous answers before introducing any suggestions, by ordering questions from open-ended (‘What did you experience?’) to more specific (‘Can you describe where exactly your tongue was touching your fingers?’) and very focused (Please rate these statements: ‘I felt as if my tongue was split in two separate parts’; ‘I felt a hole in my tongue’). Although fundamentals of measurement apply to all types of percepts, some methodological issues might arise more often, be more serious, or even unique to illusion research. We now focus on those. 1.3.1 Use of Indirect Measures

Some measures are motivated by the proposition that ‘if illusion X occurs, then effect Y should or may occur as well’. For example, in the rubber hand illusion (RHI), ‘if the participants feel that their hand is brushed by the brush touching the artificial hand, then they should feel their hand to be located where the artificial hand is’. Such a relationship should be carefully tested rather than assumed, because it may not hold [38]. Our perceptual system tolerates dissociations and inconsistencies that cannot occur in the real world; a perfectly valid inference considering real, physical relationships may not translate into the relationships between the corresponding perceptual attributes. For example, objects that move in the real world must by definition change position, but we can see objects move without seeing much position change [39], and we can feel that our limbs move without feeling a

Somatosensory Illusions

273

corresponding change in posture [40]. Percept-percept dissociations and illusions are intrinsically linked: where such dissociations occur, one or both features are incorrectly perceived (if I feel that my arm is moving but not changing position, then at least one of the impressions must be illusory). Similar caution applies when using movement responses. They may be based on a different stimulus attribute than the illusory percept. Pointing responses, for example, rely on the relationship between the hand starting position and the target location, while conscious percepts reflect a broader stimulus configuration [41]. 1.3.2 Knowing That the Percept Is Illusory

Some illusions cannot pass as reality, or hardly so; study participants will realize that the experience is non-veridical. Illusions concerning our body are likely to fall into this category because we know what our body normally feels like. On the other hand, there are illusions that will not be recognized as such until the participant is explicitly informed. In the cutaneous rabbit illusion (see Note 3.1), taps delivered to the forearm are erroneously felt in between the locations of delivery, but these illusory locations are entirely plausible. But even in such cases, the observer may be informed, and knowing that the experience is not veridical may create a response bias. Some participants may hesitate to report illusory effects because they view them as error, failure, or weird (as communicated to us on different occasions); others may exaggerate, prompted by the demand characteristics of the study. Bias may also alter the percept itself if the sensory signal is noisy or ambiguous, as mentioned earlier.

1.3.3 Dealing with Unique, Rare, or Conflicting Experiences

Unique or rare experiences such as the vastness of space seen from a satellite, or a deep underwater environment, may be difficult to describe for a simple lack of vocabulary. This may also apply to somatosensory illusions, especially those that violate anatomical features and boundaries (such as the tongue splitting [2] or the elbow bending beyond 180 degrees [14]). Participants call upon their imagination to find words for such experiences, experimenters risk biasing them when they offer descriptions and rating scales, and tasks where the illusory experience is matched to other experiences may not feel right. Complex illusions may also take longer to develop or to ‘settle’ into an organized percept because their different aspects are incompatible, analogous to impossible figures in vision [42]. Some of the above challenges can be addressed by instructions, as illustrated by an unlikely source – children’s games. Traditional games that involve illusions [2] have features similar to experimental procedures, including the roles of ‘director’ and ‘actor’ (experimenter and participant, respectively), and narratives, lyrics, or chants to prepare for the illusion, and/or to accompany the performance (instructions). Some chants call for concentration (‘Concentration, concentration!’), which, according to Barker and Rice [2],

274

Tatjana Seizova-Cajic´ et al.

make the child “understand (consciously or unconsciously) that they must submit their consciousness to the imaginary situation of the verbal component in order for the illusion to be successful.” (p. 455). Suggestion is an inherent part of the narrative, and in children’s games it presents no problem; rather, it helps to frame and communicate the experience, to everyone’s delight (recall the ‘digging’ of a hole underneath a child’s legs as a preparation for the postural aftereffect). In research, we likewise want to encourage open-mindedness and direct participants to attend to the relevant aspect of the experience but would like to avoid suggestion. Our ‘narratives’ – experimental instructions – need to strike a fine balance. To summarize, there are no special illusion-tools in perception researchers’ toolboxes, but some issues may arise more often in illusion research and warrant attention. They include insufficiently tested assumptions about links between illusory and other experiences, possibility of reporting bias in participants aware of the illusory nature of the experience, and extraordinary experiences that are difficult to communicate and measure. We now describe detailed methods for the study of two example illusions. They are both illusions of source, in that there is an error in perceiving who or what is touching us. The RHI in its classical form involves synchronous input from vision and touch, where the seen touches of the brush delivered to the artificial hand and the felt touches of one’s own hand are combined into a single percept [8]. In the WHI, perception changes from ‘touching you’ (veridical percept) to ‘touching myself’ (an illusion) during an unchanging physical contact with hands of another person [43].

2

Materials and Methods for Two Example Illusions

2.1 Rubber Hand Illusion (RHI)

In the RHI, an artificial hand is placed on a table in front of the participants while their real hand is hidden from view. When the experimenter simultaneously strokes the artificial hand and the participants’ real hand in approximately the same location (Fig. 2a), the participants often report feeling the brushstrokes in the location where they see the artificial hand being touched. Strikingly, although to a somewhat lesser extent [44], participants also report the illusory feeling that the artificial hand belongs to their body [8]. The illusion can also be induced by moving the participant’s finger and the artificial finger at the same time (Fig. 2b, c), conveyed via vision and touch [45, 46], or via proprioception [47]. This fascinating illusion has extended our knowledge about the significance of correlated sensory (or sensorimotor) inputs and plasticity of body ownership. The paradigm has also been applied to other body parts: the foot [48], the face [49], and the entire body [50, 51].

Somatosensory Illusions

275

Fig. 2 Three different methods for inducing the rubber hand illusion (RHI). In each panel, the hand underneath the occluder is the participant’s hand. (a) The classical setup involves simultaneous brushing of the participant’s hand and the artificial hand; (b) Participants actively move their hand, resulting in simultaneous movements of the artificial hand; (c) The experimenter moves the fingers of artificial and real hands at the same time using a rigid probe 2.1.1

Materials

The basic RHI setup requires three components (Fig. 2): an artificial hand (the viewed object should look like a hand and be in a plausible orientation and location compared to participant’s own hand [52–55]), an occluder that hides the participant’s hand from their view but allows the experimenter to touch both hands, and two paintbrushes or other paired tools to stimulate the hands. Experimenter and participant are seated opposite each other. Artificial hand: Medical prostheses, artist’s wooden hand models covered with a rubber glove, or plastic hand mannequin models (as used for training in hand nail art) can be used. An artificial hand can also be presented using virtual reality, robotics, or video projections. For the RHI to work it is important that the artificial hand looks like a hand, but at the same time, it can differ from the participant’s actual hand in terms, for example, of size, shape, hairiness, or skin tone [56, 57]. The occluder: Anything that occludes the participant’s hand from their sight will do – a bench, box, shelf, or curtain (for sideby-side arrangement). It is common practice to cover the area between the artificial hand and observer’s body using a sleeve or hairdresser’s gown, to conceal the fact that the two are not connected [58]. Two paintbrushes or other paired stimulators: When tactile stimuli are used to induce the illusion, two paintbrushes, experimenter’s fingers, or other paired stimulators stroke the artificial and real hands, controlling the relative timing of strokes. When using movement to induce the illusion, a mechanical device can transfer the movement of the participant’s finger to the artificial hand, or the real hand movement gets captured (using a camera or data glove) and projected onto a screen or virtual object [59].

276 2.1.2

Tatjana Seizova-Cajic´ et al. Methods

Set-up and basic procedure: The artificial hand is clearly visible, placed in front of the participant, usually beside or above their real (occluded) hand. The experimenter then strokes the fingers of the artificial and real hands at the same time, in a matching location and in a similar manner in terms of direction and velocity of the sweeps. Therefore, whenever participants feel a touch, they see the artificial hand being touched in a way that appears plausible to have induced the (felt) touch. During this induction period (typically 60–120 s of stroking), participants are instructed to look at the artificial hand and keep their own hand still. Control conditions: Synchronous stimulation is often compared to a condition where touches are presented asynchronously, with 1–2 s delay between the touches to the participants’ hand and the artificial hand (temporal discrepancy beyond 300 ms is detrimental to the illusion [60]). In addition to this classic control condition, others have been used, such as synchronous stroking of objects clearly deviating from a bodily shape (e.g., a wooden block [61]), or in an anatomically impossible position (e.g., an artificial hand rotated by 180 degrees so that the fingers are pointing towards the participant [62]). Dependent variables: The RHI is commonly measured directly, using a questionnaire, and indirectly, by indicating the felt location of the own hand or via physiological responses to body threat. Questionnaires are direct measures of the RHI, often relying on rating scales. They ask about different aspects of the experience: about feeling the touch where it is seen – known as remapping of touch (“It seemed as if I was feeling the touch of the paintbrush in the location where I saw the fake hand touched”), about the sense of body ownership (“It felt as if the artificial hand was my own hand”), and, for variants of the RHI involving active or passive movements (Fig. 2b, c), about the feeling of agency over the artificial hand (“It felt as if I was controlling the movements of the artificial hand”). Issues to consider when designing or choosing an RHI questionnaire include comparability across studies, and across dimensions of experience. Comparability across studies depends on a range of factors, from the type of rating scale used to the choice of the experimental and ‘control’ items. For example, a rating scale ranging from 1 to 7 [63] may give a different outcome from a -3 to 3 scale [64], because the middle value of 4 on the former might not be seen to be as ‘neutral’ as the middle value of 0 on the latter scale. Questionnaire items should also be comparable across dimensions of experience such as ownership and agency, if the intention is to determine which is impacted more. However, ownership questionnaire items are often patently false (e.g., “... the artificial hand was my own hand”), while items targeting agency are often plausible (e.g., “... I caused the movements of the artificial hand”), because the set-up in the active RHI often enables participants to directly

Somatosensory Illusions

277

cause the movements of the artificial hand. (For a psychometric analysis of the scales, see [65, 66], and for variation of control items between studies, [67].) A commonly used indirect measure is ‘proprioceptive drift’, defined as a change in the perceived location of the participant’s hidden hand induced by the RHI [8, 68]. The perceived location is measured behaviorally, for example, by pointing to the hidden hand using the contralateral index finger. A shift of around 1 to 3 cm typically occurs towards the location of the artificial hand after the RHI is induced. Other methods to assess proprioceptive drift include judging one’s hand position using a ruler placed on top of a box [68, 69], and rapid pointing movements with one’s own hand – the one subjected to the illusion – towards visual targets [70, 71]. It is unclear how comparable these assessment methods are, or which aspects of the RHI they capture (for detailed discussions, see [46, 67, 70]). It is also important to note that several studies have revealed dissociations between subjective ownership ratings and proprioceptive drift, suggesting that they might be driven by different mechanisms [38, 69, 72, 73]. RHI studies also measure physiological responses to bodily threats directed at the artificial hand using skin conductance [74] or blood oxygen level-dependent (BOLD) responses from functional magnetic resonance imaging [75] (see Chapter 18, [76]). Several studies also recorded the skin temperature changes of the real hand because Moseley et al. [77] showed a cooling effect during the RHI; however, this effect was not replicated [78]. The effects of the RHI were also examined using a cross-modal congruency task, in which participants make speeded responses to discriminate the location of tactile stimuli on the actual hand (e.g., touch to index finger or middle finger), while visual distractor stimuli are presented to different locations in the space near the rubber hand (e.g., a light above the index or middle finger). Synchronous stroking leads to more interference from the visual stimuli on the tactile task (i.e., more multisensory interactions) compared to asynchronous stroking [79, 80]. 2.2 Whose Hand Illusion (WHI)

Touching a body part of another person feels very different from touching our own. It may thus be surprising that an illusion of selftouch arises after only a few minutes (or less) of touching someone else’s fingertips, with eyes closed. Observers are asked to focus on the feeling of whose hand they are touching, their partner’s or their own, while maintaining the posture as depicted in Fig. 3a, b (see Note 3.4). The resulting illusion of self-touch has been named the whose hand illusion (WHI). The median onset time in the formal study [43] was approximately 3 min, and a majority of participants experienced the illusion (9/12). Comparable times were observed on a large number of people at public events. When the WHI occurs, the observer’s two hands feel to be opposite each other, as they would have to be for the fingers of opposite hands to self-

278

Tatjana Seizova-Cajic´ et al.

Fig. 3 Whose hand illusion (WHI) and hand postural illusion (HPI) setup and tasks. The posture that elicits the two illusions is shown: (a) From a side view and (b) from a bird’s eye view. (c) Two interleaved tasks are used to test for the presence of WHI and HPI, respectively: the participant reports whose hand she feels to be touching, and also the perceived relative position between two taps applied from time to time to the dorsum of her hands (small black circles in (b)). During practice, participants report the angle between the taps while looking at a printout of the clockface (shown as a transparent inset in (b)). During the experimental run, participants’ eyes are closed and they rely on memory. Using the example shown in (b), the correct description of taps’ relative position is either 1 o’clock or 7 o’clock, depending on which hand was tapped last (‘1’ if participant’s right hand was tapped last). Given that the perceived hand position is subject to the illusory bias (HPI), perceived relative position of taps may also be biased. In data analysis, angles are expressed relative to a reference, e.g., 3 o’clock is recorded as 90 deg., and 10:30 as 45 deg. relative to the mid-sagittal plane. Which hand was tapped last can be ignored in data analysis

touch. It is tempting to think that this latter effect - the hand postural illusion (HPI) - is a necessary condition of the WHI and perfectly correlated with it, but their relationship should be investigated rather than assumed (see Note 3.5). In the WHI, observers very well know, and initially correctly perceive, that they are touching someone else. The ambiguity seems to arise gradually, ending in a profound qualitative switch that resembles switching from one to another visual percept of the Necker cube, or during binocular rivalry. The switch to the illusory experience can often be read from the volunteers’ surprised and amused facial expressions. Switching back to the veridical percept occasionally occurs, possibly caused by movements of the partner’s hand. The WHI itself, and its stability, reveal that we are biased to perceive self-touch, presumably because it is the kind of touch we most often experience. It is a more common ingredient in our sensory diet (see Note 3.3) than touching others. A different illusion of self-touch while touching others, named the phantom nose illusion [81], occurs when one repetitively strokes another person’s nose while one’s own nose is also being stroked (12/18 naı¨ve participants experienced this illusion, [82] pp. 452–3). This illusion relies on simultaneous stroking – the coincidence of sensory inputs – just like the RHI. By contrast, the WHI occurs during stable skin contact and in the absence of other tactile stimulation, which is methodologically beneficial if neural

Somatosensory Illusions

279

correlates of perceived self-touch are explored using brain imaging. Since any signal is easier to detect against reduced noise, a neural ‘switch’ corresponding to the perceptual switch from ‘touching you’ to ‘touching myself’ should be easier to identify in the relative absence of other sensory stimulation than during the repetitive stimulation that characterizes phantom nose illusion and RHI paradigms (see Note 3.6). 2.2.1

Materials

Two people able to assume and maintain the required posture are sufficient to create the WHI and the HPI. In a formal study, one is the experimenter’s aid and the other, study participant. Materials required for testing the WHI are a blindfold for the participant, a recorder for verbal responses, and pre-recorded prompts for the experimenter and participant. The testing of HPI requires, in addition to the above materials, a pair of tactile stimuli that can be applied to the back of the participant’s hands, and a printed clockface for practice in reporting angles. The tactile stimuli can be handdelivered (using e.g., brushes) or computer-controlled (e.g., vibrators or piezoelectric stimulators).

2.2.2

Methods

Setup: The helper and the study participant touch each other’s fingertips for 5 min or longer, trying to be still and avoiding touches elsewhere on the hands and forearms (Fig. 3a, b). The fingertips of the participant’s two hands are close to their mid-sagittal plane, one hand further away from their face than the other; their elbows rest on the table comfortably, approximately shoulder-width apart. Which hand is closer to the face can be counterbalanced across participants. The helper’s left hand touches the participant’s right hand, and vice versa. The posture of the helper’s arms is adjusted to accommodate the participant’s posture. The participant is instructed to use consistent, light pressure when touching the helper’s fingertips. If the participant’s hand is larger than the helper’s, or vice versa, flexing fingers forward – rather than keeping them fully extended as shown in Fig. 3 - will help them establish fingertip-to-fingertip contact with each other. Twitches and other small movements of the fingers will occur; it is possible that they interfere with and delay the onset of the illusion, but they do not prevent it. Measurement of the WHI: Participants receive timed prompts to report in one word whose hand they feel to be touching (‘yours’ or ‘mine’, Fig. 3c). Done every 30 s or so, this forced-choice task focuses the participant’s attention on the relevant aspect of the experimental situation. Measurement of the HPI: The perceived position of the hands may deviate from their real position. Relative hand position is measured indirectly: answers about taps applied to the hands are used to infer where the hands feel to be relative to each other. This draws attention away from the hands in order to reduce the

280

Tatjana Seizova-Cajic´ et al.

potential for bias. Bias may occur if the participant reasons, for example, that ‘given that I feel I am touching my own fingertips, it makes sense to report that my hands are opposite each other’. At set time intervals (e.g., every 10 s), the experimenter briefly touches the middle of the hand dorsum, proximal to the third knuckle, in quick succession and randomly varying the order of touches from trial to trial (e.g., left hand first vs. right first). This variation ensures that the participant pays attention to the stimuli rather than repeating the same answer over and over. Alternatively, two computer-controlled tactors can be attached to the same locations, one on each hand (black circles in Fig. 3b), and activated at the set interval. The participant reports the relative location of taps (‘Where are the two taps relative to each other?’). An angular reporting scale is presented in the form of a printed clockface, placed horizontally on the table during practice trials only (Fig. 3b, c). During practice, the experimenter rearranges the participant’s relative hand position a few times and applies the taps each time; the participant’s task is to look at the clockface and estimate the position of taps relative to each other. A couple of practice trials using different postures are sufficient for participants to understand the instructions and memorize the clockface. During the experimental run, the participant is blindfolded and relies on their memory of a clockface. This measure alternates with the self-report of the WHI illusion described above. At the end of the session, open-ended questions can be used to explore the experience, such as the following: (1) Can you tell us how it felt when you said you were touching the experimenter’s fingers, and how it felt when it was your own fingers or hand? (2) Where did you feel your hands were relative to each other? (3) Is there anything else you would like to describe? Instructions to foster open-mindedness: In addition to the specific instructions regarding experimental procedure, participants not used to perception studies are encouraged to report what they experience, regardless of what they know, or think, is true. A color aftereffect can be used to illustrate the desired attitude (see Subheading 1 and Note 3.7). Demand characteristics in WHI: The participant is asked to choose one of the two answers - ‘yours’ or ‘mine’ (meaning ‘it feels like I am touching you’, vs. ‘it feels like I am touching myself’). This makes it clear that the experimenter expects both percepts to occur. One way to address the risk of bias this may create is to clearly state that only the participant’s experience is of value, regardless of what anyone expects (see Note 3.7). Some leverage against the bias also comes from using the same task in control conditions, although - as indicated earlier - participants may hold different beliefs about different experimental conditions. Asking for a spontaneous, self-paced report (‘Let us know what you feel’; ‘Let us know if you feel anything unusual’) would avoid suggestion but

Somatosensory Illusions

281

does not control the attentional focus, which is required to organize the percept in a particular way and to constantly query what one perceives, both desirable when investigating WHI. Control conditions: WHI seems to occur due to the symmetrical finger-to-finger contact: each of participant’s fingers touches a finger – although not their own. The control condition can alter this in many ways, without altering other aspects of the situation. For example, the helper may press her closed fists into the participant’s outstretched palms. The participant’s posture and task remain the same (they report whose hand they feel to be touching). Analysis: Different measures can be used to characterize the WHI: the number of participants who experience the illusion at different time points, its cumulative duration relative to the total time, time-to-illusion onset, etc. HPI can be quantified as the angular change over time (see Note 3.8). The relationship between WHI and HPI can be tested by comparing their relative strengths across different experimental conditions, and/or over time.

3

Notes

3.1 Illusory Perception of Different Stimuli as Equal

In the example study, six equally spaced contactors were placed along the left arm, and received one pulse each at equal time intervals, resulting in apparent motion. Apparent motion is itself illusory, but in this study, the focus was on another illusion, known as sensory saltation or the cutaneous rabbit illusion [81]. This illusion was created on the right arm: it had only 3 contactors, but the timing of the 6 pulses was adjusted to yield the same apparent spacing as the reference stimulus on the left arm (pp. 50–51 in [81]).

3.2 Keywords for Database Searches

Useful search terms are the following: illusions of touch; tactile illusions; haptic illusions; illusions of body movement or motor illusions; proprioceptive or kinesthetic illusions; bodily illusions.

3.3 Role of Knowledge in Perception

Demand characteristics are knowledge and expectations about the current experimental situation. A different meaning of knowledge can be described statistically, as a function of past stimulation [36]: our sensory responses have been shaped by exposure to sensory diets over a range of time scales, from evolutionary (‘light comes from above’) to minutes- or seconds-long (e.g., dark adaptation), allowing perception to be a ‘fast and dumb’ mechanism [82, 83]. Noise or ambiguity in the stimulus call for greater reliance on this ‘statistical’ knowledge (a 2-D circle which is shaded to appear lighter in its top half will be seen as a bump rather than a hole because ‘light comes from above’), but knowledge of the experimental situation may also help people to resolve ambiguous stimuli – an observer primed to think about dogs might see a

282

Tatjana Seizova-Cajic´ et al.

Dalmatian dog in a black-and-white picture more readily than someone not thus primed. 3.4 Variants of the Whose Hand Illusion

Pilot tests suggest that other postures involving symmetrical touch also lead to the illusion of self-touch. For example, two partners sitting opposite each other, their forearms alternately laid on top of each other like layers of a cake, with hands wrapped around elbows, may experience that they are touching themselves rather than their partner.

3.5 Whose Hand and Hand Postural Illusions

One may think that WHI causes or facilitates the shift in the hand’s perceived position (HPI), but the opposite direction of influence is also plausible, because HPI occurs even in the absence of WHI [43]. Note the possible similarity with proprioceptive drift in the RHI, which was thought to occur due to the RHI, but when probed more deeply, it was also found in experimental conditions in which RHI did not occur [38].

3.6 Switching Between Bistable Percepts as a Tool to Study Consciousness

A perceptual switch in WHI is analogous to changes in the visual awareness that occur during relatively constant visual input. One example of the latter is the much-studied binocular rivalry, deemed to be of special importance to the visual neuroscience “because of [the] profound dissociation between physical stimulation and perceptual experience”; it “affords a paradigmatic case for the study of conscious visual awareness” ([84], p. 182). In a similar vein, WHI affords a case for the study of tactile awareness.

3.7 Using Color Aftereffect to Encourage OpenMindedness

In a simple demonstration of the color aftereffect, the participant fixates a 5 × 5 cm red square presented on a white background for 30 s, followed by a gaze shift to a uniformly painted wall. The experimenter explains: ‘You know that this wall is of a uniform color, but we are interested in what you perceive and not what you know or think is true. So, if you see a colored shape on the wall, that’s what you need to report. If you don’t see it, that’s fine too. We are interested in what you experience, and nothing else.’

3.8 Processing of Clockface Responses

Original responses expressed as the time on the clockface need to be processed before converting them to angles, if physical posture is counterbalanced across participants (right hand closer to face vs. left hand closer to face). When the left hand is closer to the face, the physical relationship between the two taps, that is, the two hands, corresponds to the line connecting 1 and 7 o’clock (Fig. 3b); when the right hand is closer to the face, the corresponding line is mirror-reversed (11 and 5 o’clock). To express the responses of the two groups on the same scale, the responses of one group need to be mirror-reversed. After this, they can be converted to angles to indicate the relative position between the hands, regardless of the order of tactile stimuli; no angle is thus greater than 180°.

Somatosensory Illusions

283

Acknowledgments We thank Dejan Todorovic´ for comments on the manuscript, and Sandra Rose Brand and Marta Beauchamp for their artistic contributions to the figures. References 1. Spillmann L (2009) Phenomenology and neurophysiological correlations: two approaches to perception research. Vis Res 49:1507–1521 2. Barker KB, Rice C (2019) Folk illusions: children, folklore, and sciences of perception. Indiana University Press 3. Shapiro A, Todorovic´ D (2017) The Oxford compendium of visual illusions. Oxford University Press, New York 4. Todorovic´ D (2020) What are visual illusions? Perception 49(11):1128–1199 5. Hayward V (2015) Tactile illusions. Scholarpedia 10(3):8245 6. Goodwin GM, McCloskey DI, Matthews PB (1972) The contribution of muscle afferents to Kinaesthesia shown by vibration induced illusions of movement and by the effects of paralysing joint afferents. Brain 95(4):705–748 7. Kavounoudias A, Blanchard C, Landelle C, Chancel M (2023) Muscle tendon vibration: a method for estimating kinaesthetic perception. Ch 3, present volume 8. Botvinick M, Cohen J (1998) Rubber hands ’feel’ touch that eyes see. Nature 391(6669): 756 9. Israr A, Poupyrev I (2011) Tactile brush: drawing on skin with a tactile grid display. In: Proceedings of the 2011 annual conference on human factors in computing systems - CHI 11. ACM Press, New York, pp 2019–2028 10. Seizova-Cajic T, Fuchs X, Brooks J (2023) Creating tactile motion. Ch 4, present volume 11. Visell Y et al (2008) A Vibrotactile device for display of virtual ground materials in walking. In: Ferre M (ed) Haptics: perception, devices and scenarios. EuroHaptics 2008. Lecture notes in computer science, vol 5024. Springer, Berlin, Heidelberg 12. Lederman SJ, Jones LA (2011) Tactile and haptic illusions. IEEE Trans Haptics 4(4): 273–294 13. Hayward V (2008) A brief taxonomy of tactile illusions and demonstrations that can be done in a hardware store. Brain Res Bull 75(6): 742–752

14. Craske B (1977) Perception of impossible limb positions induced by tendon vibration. Science 196(4285):71–73 15. Mochiyama H et al (2005) Haptic illusions induced by moving line stimuli. First joint Eurohaptics conference and symposium on haptic interfaces for Virtual environment and teleoperator systems. World Haptics Conference, Pisa, Italy, 2005, pp 645–648 16. Weber EH, Ross HE, Murray DJ (1978) E. H. Weber: the sense of touch. (De Tactu, 1934, translated by HE Ross; Der Tastsinn und das Gemeingefuhl, 1946, translated by DJ Murray). Academic Press 17. Goudge ME (1918) A qualitative and quantitative study of Weber’s illusion. Am J Psychol 29:81–119 18. Longo MR, Haggard P (2011) Weber’s illusion and body shape: anisotropy of tactile size perception on the hand. JEP: HPP 37:720– 726 € 19. Jaensch E (1906) Uber T€auschungen des Tastsinns. (Im Hinblick auf die geometrischoptischen T€auschungen). Zeitschr Psychol 41: 280–294 20. Revesz G (1934) System der optischen und haptischen Raumtauschungen. Z Psychol Bd 131:296–375 21. Katz D (1925/1989) The world of touch. Lawrence Erlbaum Associates 22. Schilder P (1935/1999) The image and appearance of the human body. Kegan Paul, Oxford 23. Gentaz E, Hatwell Y (2008) Haptic perceptual illusions. In: Grunwald M (ed) Human haptic perception: basics and applications Birkh€auser. Basel 24. Gallace A, Spence C (2014) In touch with the future. Oxford University Press 25. Jones L (1988) Motor illusions: what do they reveal about proprioception? Psychol Bull 103(1):72–86 26. Lackner J (1988) Some proprioceptive influences on the perceptual representation of body shape and orientation. Brain 111:281– 297 27. Taylor MW, Taylor JL, Seizova-Cajic´ T (2017) Muscle vibration-induced illusions: review of

284

Tatjana Seizova-Cajic´ et al.

contributing factors, taxonomy of illusions and user’s guide. Multisens Res 30(1):25–63 28. Culbertson H, Schorr SB, Okamura AM (2018) Haptics: the present and future of artificial touch sensation. Annu Rev Control Robot Auton Syst 1:385–410 29. Newport R, Wong DY, Howard EM, Silver E (2016) The Anne Boleyn illusion is a six-fingered salute to sensory remapping. Iperception 7(5):2041669516669732 30. Hoshi T, Abe D, Shinoda H (2009) Adding tactile reaction to hologram. RO-MAN The 18th IEEE international symposium on robot and human interactive communication, Toyama, pp 7–11, https://doi.org/10.1109/ ROMAN.2009.5326299 31. Rossi GB (2014) Measurement and probability: a probabilistic theory of measurement with applications. Springer Netherlands, Dordrecht 32. Kingdom F, Prins N (2014) Psychophysics: a practical introduction, 2nd edn. Academic Press 33. Orne MT (1962) On the social psychology of the psychological experiment: with particular reference to demand characteristics and their implications. Am Psychol 17:776–783 34. Lush P (2020) Demand characteristics confound the rubber hand illusion. Collabra Psychol 6(1):1–10 35. Seth A, Roseboom W, Dienes Z, Lush P (2021) What’s up with the Rubber Hand Illusion? psyarxiv.com/b4qcy 36. de Lange FP, Heilbron M, Kok P (2018) How do expectations shape perception? Trends Cogn Sci 22(9):764–779 37. Orne M, Scheibe KE (1964) The contribution of nondeprivation factors in the production of sensory deprivation effects: the psychology of the “panic button”. J Abnorm Soc Psychol 68: 3–12 38. Rohde M, Luca M, Ernst M (2011) The rubber hand illusion: feeling of ownership and proprioceptive drift do not go hand in hand. PLoS One 6(6):e21659 39. Nishida S, Johnston A (1999) Influence of motion signals on the perceived position of spatial pattern. Nature 397:610–612 40. McCloskey DI (1973) Differences between the senses of movement and position shown by the effects of loading and vibration of muscles in man. Brain Res 63:119–131 41. Smeets JB, Brenner E (2002) Illusions in action: consequences of inconsistent processing of spatial attributes. Exp Brain Res 147: 135–144 42. Penrose LS, Penrose R (1958) Impossible objects: a special type of illusion. Br J Psychol 49:31–33

43. Seizova-Cajic´ T, Stuart E, Taylor J (2017) Whose hand? An illusion of self-touch and stubborn proprioceptive drift. Science of the self: the Agency & Body Representation Research Forum, Sydney, Australia 44. Reader AT, Trifonova VS, Ehrsson HH (2021) The relationship between referral of touch and the feeling of ownership in the rubber hand illusion. Front Psychol 12:629590 45. Kalckert A, Ehrsson HH (2014) The moving rubber hand illusion revisited: comparing movements and visuotactile stimulation to induce illusory ownership. Conscious Cogn 26:117–132 46. Riemer M, Kleinbo¨hl D, Ho¨lzl R, Trojan J (2013) Action and perception in the rubber hand illusion. Exp Brain Res 229:383–393 47. Walsh LD, Moseley GL, Taylor JL, Gandevia SC (2011) Proprioceptive signals contribute to the sense of body ownership. J Physiol 589: 3009–3021 48. Lenggenhager B, Hilti L, Brugger P (2015) Disturbed body integrity and the “Rubber Foot Illusion”. Neuropsychology 29:205–211 49. Tsakiris M (2008) Looking for myself: current multisensory input alters self-face recognition. PLoS One 3:e4040 50. Ehrsson HH (2007) The experimental induction of out-of-body experiences. Science 317: 1048 51. Lenggenhager B, Tadi T, Metzinger T, Blanke O (2007) Video ergo sum: manipulating bodily self-consciousness. Science 317:1096– 1099 52. Makin TR, Holmes NP, Ehrsson HH (2008) On the other hand: dummy hands and peripersonal space. Behav Brain Res 191:1–10 53. Tsakiris M (2010) My body in the brain: a neurocognitive model of body-ownership. Neuropsychologia 48:703–712 54. Kilteni K, Maselli A, Kording KP, Slater M (2015) Over my fake body: body ownership illusions for studying the multisensory basis of own-body perception. Front Hum Neurosci 9: 141 55. Riemer M, Fuchs X, Bublatzky F et al (2014) The rubber hand illusion depends on a congruent mapping between real and artificial fingers. Acta Psychol 152C:34–41 56. Farmer H, Tajadura-Jimenez A, Tsakiris M (2012) Beyond the colour of my skin: how skin colour affects the sense of bodyownership. Conscious Cogn 21(3): 1242–1256 57. Longo MR, Schu¨u¨r F, Kammers MP et al (2009) Self awareness and the body image. Acta Psychol 132:166–172

Somatosensory Illusions 58. Perez-Marcos D, Sanchez-Vives MV, Slater M (2012) Is my hand connected to my body? The impact of body continuity and arm alignment on the virtual hand illusion. Cogn Neurodyn 6: 295–305 59. Zopf R, Friedman J, Williams MA (2015) The plausibility of visual information for hand ownership modulates multisensory synchrony perception. Exp Brain Res 233:2311–2321 60. Shimada S, Fukuda K, Hiraki K (2009) Rubber hand illusion under delayed visual feedback. PLoS One 4:e6185 61. Tsakiris M, Carpenter L, James D, Fotopoulou A (2010) Hands only illusion: multisensory integration elicits sense of ownership for body parts but not for non-corporeal objects. Exp Brain Res 204:343–352 62. Ehrsson HH, Spence C, Passingham RE (2004) That’s my hand! activity in premotor cortex reflects feeling of ownership of a limb. Science 305:875–877 63. Heed T, Gru¨ndler M, Rinkleib J et al (2011) Visual information and rubber hand embodiment differentially affect reach-to-grasp actions. Acta Psychol 138:263–271 64. Riemer M, Wolbers T, Kuehn E (2019) Preserved multisensory body representations in advanced age. Sci Rep 9:2663 65. Longo MR, Schu¨u¨r F, Kammers MP et al (2008) What is embodiment? A psychometric approach. Cognition 107:978–998 66. Romano D, Maravita A, Perugini M (2021) Psychometric properties of the embodiment scale for the rubber hand illusion and its relation with individual differences. Sci Rep 11: 5029 67. Riemer M, Trojan J, Beauchamp M, Fuchs X (2019) The rubber hand universe: on the impact of methodological differences in the rubber hand illusion. Neurosci Biobehav Rev 104:268–280 68. Tsakiris M, Haggard P (2005) The rubber hand illusion revisited: Visuotactile integration and self-attribution. JEP: HPP 31:80–91 69. Riemer M, Bublatzky F, Trojan J, Alpers GW (2015) Defensive activation during the rubber hand illusion: ownership versus proprioceptive drift. Biol Psychol 109:86–92 70. Kammers MPM, de Vignemont F, Verhagen L, Dijkerman HC (2009) The rubber hand illusion in action. Neuropsychologia 47(1): 204–211 71. Zopf R, Truong S, Finkbeiner M, Friedman J, Williams MA (2011) Viewing and feeling touch modulates hand position for reaching. Neuropsychologia 49(5):1287–1293 72. Holmes NP, Snijders HJ, Spence C (2006) Reaching with alien limbs: visual exposure to

285

prosthetic hands in a mirror biases proprioception without accompanying illusions of ownership. Percept Psychophys 68(4):685–701 73. Holle H, McLatchie N, Maurer S, Ward J (2011) Proprioceptive drift without illusions of ownership for rotated hands in the “rubber hand illusion” paradigm. Cogn Neurosci 2(3–4):171–178 74. Armel KC, Ramachandran VS (2003) Projecting sensations to external objects: evidence from skin conductance response. Proc R Soc B Biol Sci 270:1499–1506 75. Ehrsson HH, Wiech K, Weiskopf N, Dolan RJ, Passingham RE (2007) Threatening a rubber hand that you feel is yours elicits a cortical anxiety response. Proc Natl Acad Sci U S A 104(23):9828–9833 76. Puckett AM, Sanchez Panchuelo RM (2023) Imaging somatosensory cortex: human functional magnetic resonance imaging (fMRI), present volume 77. Moseley GL, Olthof N, Venema A, Don S, Wijers M, Gallace A (2008) Psychologically induced cooling of a specific body part caused by the illusory ownership of an artificial counterpart. Proc Natl Acad Sci U S A 105(35): 13169–13173 78. de Haan AM, Van Stralen HE, Smit M, Keizer A, van der Stigchel S, Dijkerman HC (2017) No consistent cooling of the real hand in the rubber hand illusion. Acta Psychol 179: 68–77 79. Zopf R, Savage G, Williams MA (2010) Crossmodal congruency measures of lateral distance effects on the rubber hand illusion. Neuropsychologia 48(3):713–725 80. Zopf R, Savage G, Williams MA (2013) The crossmodal congruency task as a means to obtain an objective behavioural measure in the rubber hand illusion paradigm. J Visual Exp 77:e50530 81. Ramachandran VS, Hirstein W (2019) Three laws of qualia: what neurology tells us about the biological functions of consciousness. J Conscious Stud 4:429–457 82. Geldard F (1975) Sensory saltation: Metastability in the perceptual world. Wiley 83. Fodor JA (1983) Modularity of mind: an essay on faculty psychology. MIT Press, Cambridge, MA 84. Nakayama K (2005) Modularity in perception, its relation to cognition and knowledge. In: Goldstein EB (ed) Blackwell handbook of sensation and perception 85. Blake R, O’Shea RP (2009) Binocular Rivalry. In: Squire LR (ed) Encyclopedia of neuroscience. Academic Press, pp 179–187

Chapter 14 Sensory Substitution: Visual Information via Haptics Jack Brooks, A´rni Kristja´nsson, and Runar Unnthorsson Abstract Sensory substitution devices compensate for the lack of one sense, such as vision, by conveying information about a person’s environment through the stimulation of another sense, like touch or audition. In this chapter, we introduce a newly developed sensory substitution device (SSD), the Sound of Vision (SOV) device, which acquires visual information via a 3D camera system and conveys it to people with visual impairments through touch and audition. We focus mainly on the haptic stimulation that the system uses, introducing the relevant research and hardware. We raise important design issues for producing functionally relevant sensory substitution, such as training on the SSD, which is vital for externalization and navigation with the SOV system. All-in-all, the SOV system is a highly promising candidate for a successful SSD and is innovative in its simultaneous use of both haptic and auditory information. Key words Multisensory perception, Neuroplasticity, Remapping, Training, Vibrotactile, Vision to touch, Visual impairment.

1

Introduction For those with a sensory deficit or impairment, one possibility for improved mobility and awareness about the peripheral environment is to create a device to convey the missing environmental information via another sense. The main components required to “translate” sensory information to another modality are sensors to acquire the information, a system to remap this information into inputs for the substituting modality, and a device to stimulate the substituting modality. The combined hardware and software solutions fulfilling these functions are called Sensory Substitution Devices (SSD). Functional SSDs have been devised, but it remains a challenge to develop a device that is immediately intuitive to use, minimally disruptive to the user, and cost-effective. Within the SSD research field, there has been an emphasis on recovering visual sensation, as there are around 250 million people worldwide with visual impairments [1]. The possibility of using another sense to restore vision could lead to enormous improvements in the quality

287

288

Jack Brooks et al.

of life of the visually impaired, which can be severely affected by their impoverished visual acuity. But SSDs could also enhance intact sensory capabilities. For instance, a pilot could receive tactile information in place of the cockpit display, firefighters could use haptic feedback to navigate when smoke causes zero visibility, or tactile information could create immersion in virtual reality gaming. This chapter will first focus on methodological considerations for devices that substitute tactile information for vision, and secondly explain the key materials and methods for the Sound of Vision (SOV) device (reviewed in [2–4]). The somatosensory system is a strong candidate to target for a device aimed at recreating a down-sampled sense of vision. From the peripheral nervous system to the visual and tactile sensory cortices, the two systems have many commonalities. Mechanoreceptors in the periphery are capable of encoding touch information with sufficient bandwidth across the dimensions of pressure, spatial frequency, and temporal frequency. A large range of complex frequencies (50–800 Hz) can be encoded [5], consistent with experience from our everyday lives of being able to perceive a plethora of different textures through touch. As information passes through the nervous system, in both vision and touch, increasingly complex spatial properties of the stimuli are extracted, starting with features like edges and orientation and eventually shape [6–8]. The tactile system is also a good candidate for SSDs for visual impairment as compensatory increases in tactile acuity are known to occur in the blind [8]. Further, there is substantial cross-modal cortical plasticity following training on a spatial task using a tactile visual SSD [10], showing potential for beneficial effects of long-term use of such devices [11]. Nevertheless, there are also notable challenges to using the tactile system for sensory substitution. First, the tactile system is concerned with sensing touch to the body and not events at a distance as in vision, and it has an impoverished spatial resolution when compared to vision [4, 12]. Second, attentional capacity for complex spatial patterns is limited in the tactile modality [13] (see also Chapter 4, this volume). For example, in a task involving three subsequent stimulations to the index and middle fingers [13] response accuracy to the second of the three was reduced, showing an ‘attentional blink’, that is well-known from research in visual perception [14, 15]. Nevertheless, the overall evidence [4, 12] suggests that the somatosensory system has the necessary bandwidth and adaptability to target with SSDs to partially fulfil the functions of normal vision. 1.1 Tactile Visual Devices

One of the earliest and most ubiquitous support devices for navigation developed for the visually impaired is the white cane. Critically, users can often externalize (see Note 4.1) the sensation felt through the cane onto the outside world [16]. Although the white cane

Sensory Substitution

289

improves the mobility of the visually impaired, it is cognitively taxing to use as it requires active probing of the environment, and this takes a while to master (for discussion of cognitive load see Note 4.2). Another obvious limitation is its small reach, further compounded by the limitation that it cannot (safely) be used to detect obstacles that may be at head height. These shortcomings have prompted the development of devices providing information about the world with a larger radius of reach so that users can navigate efficiently. Bach-y-Rita and colleagues [17] developed a tactile visual sensory substitution (TVSS) device for remapping information from a camera to a 20 × 20 grid of solenoids positioned on the user’s back. The technical details regarding the design are discussed in a separate paper [18]. After tens of hours of training, expert users reached impressive levels of performance in recognizing spatial patterns. However, the device was not portable. Performance was later improved by instead electrically stimulating the tongue, which has better spatial acuity [19]. Unfortunately, users cannot usually talk or eat whilst using this version of the device. TVSS devices have also been developed for holding with the hand [20] or for attachment to the fingertips [21, 22], but a drawback is that this restricts the usage of the hands. A promising way of using SSDs to convey information about the visual scene for navigation purposes is to provide vibration to the user through a belt [23]. Van Erp and colleagues [24] introduced a haptic belt, where direction was translated into vibration location and distance was encoded in the pulse width of vibration. Navigation performance was improved by using location to cue direction, but was not further improved with the addition of distance information (for other implementations see [25, 26]). This result suggested that encoding distance in the pulse width was ineffective. A number of other TVSS devices were developed in the decades following Bach-y-Rita and colleagues’ original SSD device. Attempts were made to engineer these devices to improve their usability and portability. Throughout this time, scientists were also characterizing basic features (detection and discrimination limits for intensity, spatial, and temporal stimulus properties) of the somatosensory system and its potential plasticity [27–29] (see Chapters 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10, this volume). In parallel, there was also substantial development in technology for encoding and interpreting visual information using computer vision, in conjunction with increased fidelity of devices delivering complex patterns to the skin surface. Future devices will capitalize on our increased knowledge of somatosensation and take advantage of these technological innovations.

290

Jack Brooks et al.

1.2 Design Considerations for a Tactile Visual SSD 1.2.1 Selecting and Encoding Task-Relevant Visual Information

1.2.2 Remapping from Optical Sensors to Skin

A major challenge for any SSD is devising methods to rapidly select and encode task-relevant information [29]. The visual system and computer vision both share the challenge of comprehending rich environmental information. One approach is to detect and identify all objects that are in the camera’s view of the visual scene. Ideally, the system filters information to a level set by the user, so they only see potentially dangerous objects that they might otherwise collide with, or ground-level changes that could hinder them. The SOV project initially implemented such an approach using tactile and sound stimuli to represent the surrounding environment. At first, the user was tasked with identifying the location of multiple objects, without needing to perceive the object features. However, this task became overwhelming for the user as the number of objects in the scene increased. The approach was therefore modified towards identifying a specific set of objects and features, only a handful of them in the beginning: walls, holes, changes in elevation, stairs, and doors. A large amount of training data is required for learning algorithms to solve this problem (this is, for example, more complex than self-driving cars – as their cameras are fixed). If another approach is taken that does not involve any processing, the visual information is best represented using acoustics [30]. This approach is “natural” for the users and they can become very good at using it. Given that the technology is available to provide flexibility in what is encoded from the visual scene, how should objects that could pose a danger to visually impaired people be encoded with haptics? The world contains a lot of information that is redundant, emphasizing the importance of selecting the relevant information for further perceptual processing [30] and considering the limited bandwidth of the senses that convey the information [31, 32]. Loomis and colleagues [29] suggested that failure to consider these hard-wired limitations has resulted in a number of unusable devices. For example, the spatial resolution of the tactile sense is about four orders of magnitude lower than that of the visual system [33–35]. Note, however, that traditional measurements of acuity, such as two-point thresholds, may not set the upper limit for haptic resolution when richer and larger-scale stimulation is considered (Holmes & Tame, Chapter 1, this volume, raise other concerns with this method). Such hyperacuity arises through the integration of information from different receptors [11, 36]. Different approaches can be taken to transform spatial information to the user depending on the required resolution, the acuity of the skin, and the tactors chosen. If an SSD conveys information about the 2D spatial layout of a scene, the device can down-sample the resolution of the image, rescale it, and actuate a grid of tactors. Other aspects of vision, such as color or depth information do not have an analogous intuitive dimension in touch. In these instances,

Sensory Substitution

291

the information is remapped onto a feature-based dimension such as pressure or the frequency and amplitude of a vibrotactile waveform. The best way of conveying particular types of information from the environment is a complex topic beyond the scope of the current chapter (for discussion see [33, 35]). 1.2.3 Tactor Selection and Arrangement

The choice of tactors can be guided by general findings regarding the sensitivity of different body regions [37–39], stimulus properties [39], and the characteristics of perceptual interpretation of stimulation [40]. However, often there is little information available about acuity for the complexity of patterns that SSDs deliver. The choice of tactors has consequences for the richness of information that can be delivered. All tactors have the basic dimensions of stimulus location and timing; however, on their own, these cannot convey meaningful information. Electrocutaneous stimulation has been used, but often results in diffuse sensation and requires direct contact with the skin. Vibrotactile stimulation can be generated using electromechanical actuators, piezoelectric motors, or other devices. Vibrotactile stimulation is an attractive choice as it provides large bandwidth along the frequency and amplitude dimensions (although pairings of frequency and amplitude can be perceptually redundant). Further, vibrotactile information is easily felt through clothing with minimal loss of signal quality, making this a good choice for a portable system [3]. Selecting the skin region to convey the visual scene can be difficult as there are many competing factors that need to be considered [41]. Overall, human tactile perception is typically spatially imprecise as the skin is sparsely innervated [42], especially when compared to visual perception. As a rule of thumb, one can refer to estimates of the innervation and simple measures of acuity across the body for simple stimuli [38]. Tactile spatial acuity spans the range of mm to cm depending on body region and stimulus type. The fine spatial acuity of the hands would make them ideally suitable, but for the fact that they could then not be used for their normal function. The back and torso are reasonable alternatives as the large surface area can make up for lower spatial acuity (and a belt or vest with tactors is easily worn without interrupting other functions or activities). The arrangement of the tactors can also be guided by studies exploring different parameters of the layout of tactors of that stimulation type. Studies have explored parameters such as stimulus spacings and arrangement, and perceptual measures of spatial acuity, diffuseness of the stimulus, and motion [4, 40, 43–45]. There are also many instances in the literature of complex spatiotemporal interactions, such as sensory saltation (the “cutaneous rabbit”) [46, 47] and newly discovered haptic illusions [39, 48] (see Chapter 13, this volume). The SOV device uses vibrotactile

292

Jack Brooks et al.

stimulation to convey information about the surroundings. Although there is a considerable amount of research available on spatial acuity for static tactile pressure [37, 49, 50] it is less well described for vibration [4]. Firstly, in contrast to pressure stimuli, vibrotactile stimulation spreads well beyond the actual contact area which may cause signal interference between the tactile stimulators [43]. Additionally, pressure and vibration cause varied patterns of activation across different mechanoreceptors.

2

Materials: The Sound of Vision Device In this section, we cover the SOV device (Fig. 1), developed to help navigation for the blind and visually impaired [2, 51]. The device uses custom tactile and auditory stimulation in a complementary manner as a substitution device for vision. The system is intended to utilize the multimodal nature of observers‘representations of the environment arising from multisensory integration [52–54]. The practical relevance of the device has been proven as, within a few hours of training, some measures of navigation reach the benchmark of the white cane [54]. Cameras

The system acquires 640 by 480 pixel depth maps at a frame rate of 10 Hz. The depth maps span the camera’s entire field of view (58° horizontal by 45° vertical, having a 0.4 to 3.5 m depth range). Additionally, depth information is obtained from the Structure Sensor (Occipital, www.structure.io).

2.2 Tactile Stimulation Device

The tactile information is conveyed through stimulation from a haptic belt with a matrix of 6 by 10 vibrating motors placed on the abdomen, and a processing unit carried on the back, currently a powerful laptop (see Note 4.3 for discussion of portability).

2.1

Fig. 1 The Sound of Vision system. (a) headgear, including stereo camera and structure sensors. (b) Information about the visual scene can be given by headphones. (c) Or via tactors

Sensory Substitution

293

Our main focus here will be on the tactile belt that includes 60 eccentric rotating mass (ERM) tactors. Each ERM (9 × 25 mm) can provide frequencies in the range of 40–250 Hz, with amplitude up to 7 g. Increases in the applied voltage drive proportional increases in frequency, and drives the amplitude by a square rule. The ERMs are arranged in a 6 by 10 motor array, with 30 mm between them. 2.3

3

Acoustic Stimuli

The SOV system allows the user to select whether the information about the environment is delivered as a combination of haptic and/or acoustic stimuli. The acoustic information is delivered via open headphones that do not block the environmental sounds. This feature is important as environmental sounds already provide clues about one’s environment. Considerable effort was put into designing multi-speaker headphones that could deliver a natural sense of obstacle elevation. The results were, however, disappointing and due to time pressure the effort was aborted and it was decided to encode the elevation information in the acoustic information. Because many visually impaired individuals also suffer from hearing impairment, the final user testing in Iceland was carried out using a hearing aid made by Oticon. The results showed that the SOV system works well with the hearing aid. The hearing aid can possibly be used to further process the environmental sounds to help the user.

Methods

3.1 Encoding Optical Information

In the SOV system, 3D information from the environment is acquired with two cameras that can be easily attached to various types of headgear designs. The stereo camera uses principles similar to binocular vision to measure the distance to an object and is best suited to well-lit conditions. The structure light camera actively emits infrared light patterns and can reconstruct the distance (and shape) of objects based on how these objects distort the light pattern. As it uses infrared light, it provides excellent resolution in low lighting conditions but is less reliable outdoors due to the contamination of infrared light from the sun. Importantly, the acquisition system also integrates information from an internal measurement unit that measures the orientation of the head. Intensive image processing, using a different approach for the different data types delivered by the 3D depth camera and the Structure Sensor is performed. To represent the 3D layout of the environment, a 3D reconstruction of the environment is generated for the identified elements of interest using the data points from the cameras. This reconstruction includes positive and negative obstacles. Positive obstacles are more readily detectable and include the ground, walls, ceiling, generic obstacles, doors, and textual signs.

294

Jack Brooks et al.

Negative obstacles where a supporting surface is lacking (potholes, ditches, and stairs) can be difficult to detect. Overall, the system uses various algorithms for 3D reconstruction, including computer vision tools for segmentation and ground detection [54–56]. The detected objects are then encoded using custom audio and haptic models. For discussion on negative obstacles, see Note 4.4, and for moving objects see Note 4.5. As the environment contains multitudes of varying objects, there is a danger that a user could become disoriented if all these objects would be encoded and conveyed. This factor is also a concern when the bandwidths of the information channels do not match, as is the case when vision is substituted for tactile information as discussed in Subheading 1. Therefore, an important feature of the SOV device is that users can configure it to their liking, choosing how many objects should be encoded and how to choose the most relevant ones, based on the size, depth, and deviation from the view direction (see Note 4.6). The user can then gradually add to this as experience is gained with the device or change the settings depending on the context and requirements. To try to ensure that the SOV system can be used by as many user groups as possible (including those with hearing impairments, where haptics should presumably play a larger role), the system provides both full scene encodings and tools that can modify the main encodings. Overall, the system is intended to be highly customizable. 3.2 Auditory and Tactile Stimulation

The auditory and haptic information is rendered in real time with two encoding mechanisms. For audition, users were free to select from a number of models, of which the fluid flow model was most often chosen. The fluid flow sound model is designed for constantly changing scenarios, relying on continuous input of information about depth from the cameras. Complex liquid sounds are created from a population of bubble sounds defined from an empirical phenomenological model of bubble statistics, with nearby objects resulting in more bubbles that sound closer to the user [57]. The haptic encoding of the information from the cameras is based on the users’ relative body and head positions. These were obtained by processing data from inertial measurement units (IMUs) that were positioned on the headgear and backpack. This resulted in the haptic representation of the location of stationary objects being kept constant when users stay in place but only rotate their heads. In the closest point haptic model (Fig. 2), the direction of the surface nearest to the user within a 3.5 m radius is haptically represented, through the activation of the spatially corresponding motors on the haptic belt. Distance is encoded with frequency and amplitude (intensity) of the activated motor, which is inversely proportional to object distance: The closer the object is to the

Sensory Substitution

a

b

c

1

1

2

2

3

3

4

4

5

5

6

6 1 2 3 4 5 6 7 8 9 10

1

5 depth (m)

10

5

10

1 2 3 4 5 6 7 8 9 10

e Frequency/amplitude

d

295

1 2 3 4 5 6 1 2 3 4 5 6 7 8 9 10

0

Depth

Fig. 2 The closest point haptic model for remapping the nearest surface from depth information to vibrotactile stimulation. (a) The user is confronted with a visual scene with varying obstacles. (b) Depth map data of the scene in front of the user is continuously collected and then downsampled* The depth being the distance in metres of the nearest surface from the user at that heading. (c) The nearest surface is identified*. (d) The tactor corresponding to the nearest surface is stimulated*, and augmented by stimulating the immediately surrounding tactors. (e) The vibration and amplitude of the vibrotactile stimulus are inversely proportional to the distance of the object from the user. *arbitrary axes units

user, the higher the vibration intensity. Additionally, to amplify the vibrotactile feedback, the depth map is augmented through the activation of neighboring motors. The method used to remap depth information into tactile stimulation where the aim is to aid navigation within the scene was informed by studies characterizing vibrotactile acuity on the torso [3, 38]. In brief, vibrotactile acuity was investigated with a discrimination task in which participants judged if the second of two stimuli were to the left or right of the first. Participants were able to confidently detect differences as small as 13 mm [4]. As the skin is highly sensitive to the vibrotactile waveform, depth information can be transformed into combinations of amplitude and frequency. The rule for transformation was that the depth and vibrotactile intensity were inversely related.

296

Jack Brooks et al.

3.3 Training and Testing

Training on the device is almost certainly a highly integral part of using any SSD [58]. A rigorous active training program was developed for the SOV device, and the performance of visually impaired individuals with the SOV device was then tested following the training. We strongly recommend active training [58–60] since active engagement with tools leads to better externalization [61]. In the training study for the SOV device reported in [2], six visually impaired participants received 8 h of training with the SOV system, half of the time in a virtual reality training environment and half in a real-world setting. We compared the performance of visually impaired users with the SOV system with their performance with the white cane. After the training, visually impaired people successfully avoided collisions in a difficult navigation task just as well as they were able to when they only used the white cane, although they took longer to complete the task. Our training has so far only involved 8 h which is a very short time for a device designed for lifelong use. We expect that months and years of experience will allow the SOV encodings to become very efficient. Regarding sensory substitution, training environments such as virtual mazes or video games can strongly enhance learning and use of SSDs [62, 63], a testament to the ability of the nervous system to reorganize [64]. The first 4 h of training on the SOV device were performed in a virtual environment allowing participants to get acquainted with the SOV system through exploration without distractors or the need to coordinate their body [65]. Initially, participants learned to distinguish the size, direction, and distance of single objects, followed by the properties of multiple objects. Subsequently, participants actively navigated through virtual scenes of increasing difficulty by using a joystick, (e.g., finding one obstacle and passing between two or more obstacles). Navigation performance and usability of a device can be tested on a number of measures. Functional measures of how well an SSD is working can include the number of collisions when users travel through an environment, the obstacles identified, the time needed to navigate through a space and the recollection of areas. The device should also be benchmarked against a standard, such as navigation performance with the white cane (see [2, 9] for discussion). If sufficient performance levels are reached, testing can be extended to increasingly complex real-world situations. Users can be surveyed on all aspects of the device, and the feedback used iteratively to improve its functionality. A further measure of interest is long-term follow-up of users’ performance, to see if they still use the device and to test for learning effects and any potential cortical changes accompanying them. There is good evidence that sensory substitution can cause neural reorganization [10, 11, 64]. For example, if the blind are trained on a tactile visual SSD, increased activation of the occipital cortex is observed, as if visual information

Sensory Substitution

297

is channeled from the somatosensory to visual system. Importantly, brain regions traditionally considered devoted to one particular sense may also serve supramodal functions [66, 67]. Following training in virtual reality (VR), device users were trained in increasingly realistic but safe environments which initially consisted of navigating an empty room. With time, obstacles consisting of cardboard boxes of varying dimensions were added to the scene. Five assessments (including baseline) were conducted in the same setting at different times during the training period. The constants in the testing were that there was a 15 m distance from the starting point to the target in a 1.9 m wide straight corridor with ten rectangular column obstacles, where three were high (1.8 m) and seven were low (1.2 m). Their width and length were 0.4 m by 0.4 m. Obstacle location was varied randomly between scenes. The only restriction was that we ensured that there was always at least one free passageway of at least 1.0 m between different obstacles and between obstacles and walls. Figure 3 shows how collision frequency on the test decreased and the number of identified obstacles increased with training. In real-world test conditions, SOV users had to navigate to a test speaker positioned opposite them in the room. Performance was measured by obstacle recognition (by pointing) and obstacle avoidance (by navigating around them without contact). Improvement was seen on both measures, as expected: As more obstacles were successfully identified there was a corresponding decrease in collisions (Fig. 3). Within 8 h of training, users’ performance accuracy with the device converged with their performance with the white cane (which users were already adept at using). Interestingly, the time to navigate was unchanged with training, potentially reflecting a speed-accuracy trade-off. In any case, no time

Fig. 3 Multipanel figure depicting performance with the SOV for 6 participants (black line, mean; grey shading, standard error). Frequency of collisions decrease (a) and obstacles identified (10 were presented per scene) increase (b) with training on the SOV system. (c) Users of the SOV system reach performance level comparable to the white cane after only a few hours of training. (Figure redrawn from [1])

298

Jack Brooks et al.

Strongly disagree

Disagree

Neutral

Agree

Strongly agree

I would like to use this device frequently I found the device unnecessarily complicated The device was easy to use, set up, and operate I would need technical support to use this device The functions in this device were well integrated There was too much inconsistency in this device Most people would learn to use this device quickly I found the device very cumbersome to use I felt very confident using the device I needed to learn a lot of things to get going with this device This device would enhance leisure time activities

% of respondents

Fig. 4 SOV device user ratings from 6 participants

restrictions were implemented, especially since many errors would have been highly counterproductive and frustrating for our users. Manipulating speed and accuracy payoffs would, on the other hand, be of interest for future studies. Judging from our experience of SSD training for spatial navigation tasks, it is better to focus initially on accuracy and only subsequently on speed. 3.4

4

User Feedback

Overall, participants rated the SOV system’s usability quite favorably on an 11-item questionnaire that was based on the System Usability Scale to assess effectiveness, efficiency, and satisfaction [68]. These ratings are shown in Fig. 4.

Notes

4.1 Externalization of Touch

For an SSD to aid navigation, a key consideration is whether the touch felt on the skin is externalized into the outside world. It is essential that SSDs can become externalized [10, 69], so that the haptic pressure picked up by mechanical receptors of the body is experienced as occurring “outside“the body, making use of the so-called “out-of-body” illusion [70] (see also Chapter 13, this

Sensory Substitution

299

volume). Externalization seems to require considerable training. Visual-tactile interactions can extend into space beyond observers‘reach [71], and representations of peripersonal space can be extended with active tool use [70]. For example, Nagel and colleagues [72] used haptic stimulation with a belt to convey location information from a magnetic compass. Following training, their participants began to consider this an extra sense, almost like a “sixth” sense. 4.2 Sensory and Cognitive Load

Tests are needed to assess potential multimodal sensory overload. Measures of the cognitive load involved in using an SSD to navigate can provide an indication of how viable the device might be to use in complex real-world settings. Subjective ratings of workload can be gathered using a questionnaire like the NASA Task Load Index [73]. Objective assessments of situational awareness can be gathered by measuring responses to a stimulus (from any modality) that occurs at random intervals such as in the Detection Response Task [74]. Tests of cognitive load are also being tailored to devices for the visually impaired [75].

4.3

Portability

Further hardware developments are required to improve device portability (Subheading 2.2). Currently, the device involves a backpack with a laptop and battery. Future iterations could incorporate a more portable control unit. With a switch to coils that enable independent control of amplitude and frequency, and require less power, the device will be more compact.

4.4

Surface Texture

The nature of a surface is an important issue. For example, will the surface be supportive, like Tarmac, or an uneven terrain with vegetation? When vision is present, we can usually tell how supportive a surface will be from our prior experience of surfaces with a particular texture and shape. With an SSD, it is a considerable challenge to display textural information to the user.

4.5 Moving Obstacles

The SOV device is yet to be challenged by tests involving dynamic obstacles, although some testing has been done outdoors with obstacles such as other pedestrians [76]. Dynamic obstacles could include other pedestrians, bikes, cars, and any devices the user might interact with that have moving parts.

4.6 Object Identification

Although mobility is the primary goal, object identification has not been addressed sufficiently well, yet (e.g., is this 4-legged animal a cat or a dog?). Object identification is challenging because, without fine detail, some objects can look exactly the same as others (e.g., two animals such as a cat and a dog). In itself, this can be challenging as objects can be viewed from many perspectives, and an animal can take up a range of postures. These challenges have also been addressed in driverless car systems and require large data sets (https://waymo.com/open/).

300

Jack Brooks et al.

References 1. World Health Organization (WHO) Media Centre — Visual Impairment and Blindness Fact Sheet No. 282. [Internet]. [cited 2018 May 31]. Available from: http://www. who.int/mediacentre/factsheets/ fs282/en/ ´, 2. Hoffmann R, Spagnol S, Kristja´nsson A Unnthorsson R (2018) Evaluation of an audio-haptic sensory substitution device for enhancing spatial awareness for the visually impaired. Optom Vis Sci 95(9):757–765 3. Strumillo P, Bujacz M, Baranski P, Skulimowski P, Korbel P, Owczarek M et al (2017) Different approaches to aiding blind persons in mobility and navigation in the “naviton” and “sound of vision” projects. In: Pissaloux E, Velazquez R (eds) Mobility of visually impaired people: fundamentals and ICT assistive Technologie. Springer, pp 435–468 ´ I, Hoffmann R, Valgeirsdo´ttir 4. Jo´hannesson O VV, Unnthorsson R, Moldoveanu A, Kristja´ns´ (2017) Relative vibrotactile spatial acuity son A of the torso. Exp Brain Res 235(11): 3505–3515 5. Weber AI, Saal HP, Lieber JD, Cheng J-W, Manfredi LR, Dammann JF et al (2013) Spatial and temporal codes mediate the tactile perception of natural textures. Proc Natl Acad Sci 110(42):17107–17112 6. Apkarian-Stielau P, Loomis JM (1975) A comparison of tactile and blurred visual form perception. Percept Psychophys 18(5):362–368 7. Loomis JM, Apkarian-Stielau P (1976) A lateral masking effect in tactile and blurred visual letter recognition. Percept Psychophys 20(4): 221–226 8. Sampaio E, Maris S, Bach-y-Rita P (2001) Brain plasticity: ‘visual’ acuity of blind persons via the tongue. Brain Res 908(2):204–207 9. Sathian K (2000) Practice makes perfect: sharper tactile perception in the blind. Neurology 54(12):22034 10. Ptito M, Moesgaard SM, Gjedde A, Kupers R (2005) Cross-modal plasticity revealed by electrotactile stimulation of the tongue in the congenitally blind. Brain 128(3):606–614 ´ , Moldoveanu A, Jo´hannesson 11. Kristja´nsson A ´ I, Balan O, Spagnol S, Valgeirsdo´ttir VV O et al (2016) Designing sensory-substitution devices: principles, pitfalls and potential. Restor Neurol Neurosci 34(5):769–787 12. Ernst MO, Banks MS (2002) Humans integrate visual and haptic information in a statistically optimal fashion. Nature 415(6870): 429–433

13. Dell’Acqua R, Jolicœur P, Sessa P, Turatto M (2006) Attentional blink and selection in the tactile domain. Eur J Cogn Psychol 18(4): 537–559 14. Raymond JE, Shapiro KL, Arnell KM (1992) Temporary suppression of visual processing in an RSVP task: an attentional blink? J Exp Psychol Hum Percept Perform 18(3):849 ´ , Nakayama K (2002) The atten15. Kristja´nsson A tional blink in space and time. Vis Res 42(17): 2039–2050 16. Serino A, Bassolino M, Farne A, Ladavas E (2007) Extended multisensory space in blind cane users. Psychol Sci 18(7):642–648 17. Bach-y-Rita P, Collins CC, Saunders FA, White B, Scadden L (1969) Vision substitution by tactile image projection. Nature 221(5184): 963–964 18. Collins CC (1970) Tactile televisionmechanical and electrical image projection. IEEE Trans Man-Machine Syst 11(1):65–71 19. Stevens JC, Choo KK (1996) Spatial acuity of the body surface over the life span. Somatosens Mot Res 13(2):153–166 20. Hugues O, Fuchs P, Brunet L, Megard C (2015) Evaluation of a Vibrotactile device for outdoor and public transport pedestrian navigation using virtual reality. In: ACHI 2015: the eighth international conference on advances in computer-human interactions, vol 1, pp 24–32 21. Koo IM, Jung K, Koo JC, Nam J-D, Lee YK, Choi HR (2008) Development of softactuator-based wearable tactile display. IEEE Trans Robot 24(3):549–558 22. Shah C, Bouzit M, Youssef M, Vasquez L (2006) Evaluation of RU-netra-tactile feedback navigation system for the visually impaired. In: 2006 international workshop on virtual rehabilitation. IEEE, pp 72–77 23. McDaniel T, Krishna S, Balasubramanian V, Colbry D, Panchanathan S (2008) Using a haptic belt to convey non-verbal communication cues during social interactions to individuals who are blind. In: 2008 IEEE international workshop on haptic audio visual environments and games. IEEE, pp 13–18 24. Van Erp JBF, Van Veen HAHC, Jansen C, Dobbins T (2005) Waypoint navigation with a vibrotactile waist belt. ACM Trans Appl Percept 2(2):106–117 25. Johnson LA, Higgins CM (2006) A navigation aid for the blind using tactile-visual sensory substitution. In: 2006 international conference of the IEEE engineering in medicine and biology society. IEEE, pp 6289–6292

Sensory Substitution 26. Segond H, Weiss D, Kawalec M, Sampaio E (2013) Perceiving space and optical cues via a visuo-tactile sensory substitution system: a methodological approach for training of blind subjects for navigation. Perception 42(5): 508–528 27. ten Donkelaar HJ, Broman J, van Domburg P (2020) The somatosensory system. In: ten Donkelaar HJ (ed) Clinical Neuroanatomy. Springer, pp 171–255 28. Delhaye BP, Long KH, Bensmaia SJ (2018) Neural basis of touch and proprioception in primate cortex. Compr Physiol 8(4): 1575–1602 29. Loomis JM, Klatzky RL, Giudice NA (2012) Sensory substitution of vision: importance of perceptual and cognitive processing. In: Manduchi R, Kurniawan S (eds) Assistive technology for blindness and low vision. CRC Press, pp 162–191 30. Spagnol S, Hoffmann R, Herrera Martı´nez M, Unnthorsson R (2018) Blind wayfinding with physically-based liquid sounds. Int J Hum Comput Stud 115:9–19 31. Elli GV, Benetti S, Collignon O (2014) Is there a future for sensory substitution outside academic laboratories? Multisens Res 27(5–6): 271–291 32. K€archer SM, Fenzlaff S, Hartmann D, Nagel SK, Ko¨nig P (2012) Sensory augmentation for the blind. Front Hum Neurosci 6:37 ´ , Egeth H (2020) How feature 33. Kristja´nsson A integration theory integrated cognitive psychology, neurophysiology, and psychophysics. Atten Percept Psychophys 82(1):7–23 34. Kokjer KJ (1987) The information capacity of the human fingertip. IEEE Trans Syst Man Cybern 17(1):100–102 35. Schmidt RF (1981) Somatovisceral sensibility. In: Schmidt RF (ed) Fundamentals of sensory physiology. Springer, pp 81–125 36. Bach-y-Rita P, Kercel SW (2003) Sensory substitution and the human-machine interface. Trends Cogn Sci 7(12):541–546 37. Weinstein S (1968) Intensive and extensive aspects of tactile sensitivity as a function of body part, sex and laterality. In: Kenshalo DR (ed) The skin senses 38. Corniani G, Saal HP (2020) Tactile innervation densities across the whole body. J Neurophysiol 124(4):1229–1240 39. Hoffmann R, Valgeirsdo´ttir VV, Jo´hannesson ´ (2018) ´ I, Unnthorsson R, Kristja´nsson A O Measuring relative vibrotactile spatial acuity: effects of tactor type, anchor points and tactile anisotropy. Exp Brain Res 236(12):3405–3416

301

40. Hoffmann R, Brinkhuis MAB, Unnthorsson R, ´ (2019) The intensity order illuKristja´nsson A sion: temporal order of different vibrotactile intensity causes systematic localization errors. J Neurophysiol 122(4):1810–1820 41. Dakopoulos D, Bourbakis NG (2010) Wearable obstacle avoidance electronic travel aids for blind: a survey. IEEE Trans Syst Man, Cybern Part C Appl Rev 40(1):25–35 42. Bolanowski SJ, Gescheider GA, Verrillo RT (1994) Hairy skin: psychophysical channels and their physiological substrates. Somatosens Mot Res 11(3):279–290 43. Collins AA, Brill JC, Cholewiak RW (2001) Spatial factors in vibrotactile pattern perception. In: Eurohaptics 2001, p 41 44. Cholewiak RW, Collins AA (2000) The generation of vibrotactile patterns on a linear array: influences of body site, time, and presentation mode. Percept Psychophys 62(6):1220–1235 45. Cholewiak RW, Brill JC, Schwab A (2004) Vibrotactile localization on the abdomen: effects of place and space. Percept Psychophys 66(6):970–987 46. Geldard FA, Sherrick CE (1972) The Cutaneous “Rabbit”: a perceptual illusion. Science 178(4057):178–179 47. Brooks J, Trojan J (2017) The cutaneous rabbit effect: phenomenology and saltation. Scholarpedia 12(12):52363 48. Brooks J, Seizova-Cajic T, Taylor JL (2019) Biases in tactile localization by pointing: compression for weak stimuli and centering for distributions of stimuli. J Neurophysiol 121(3):764–772 49. Mancini F, Bauleo A, Cole J, Lui F, Porro CA, Haggard P et al (2014) Whole‐body mapping of spatial acuity for pain and touch. Ann Neurol 75(6):917–924 50. Gibson GO, Craig JC (2005) Tactile spatial sensitivity and anisotropy. Percept Psychophys 67(6):1061–1079 ´ I, Balan O, Unnthorsson R, 51. Jo´hannesson O ´ (2016) The Moldoveanu A, Kristja´nsson A sound of vision project: on the feasibility of an audio-haptic representation of the environment, for the visually impaired. Brain Sci 6(3): 20 52. Driver J, Spence C (2000) Multisensory perception: beyond modularity and convergence. Curr Biol 10(20):R731–R735 53. Wallace MT, Meredith MA, Stein BE (1993) Converging influences from visual, auditory, and somatosensory cortices onto output neurons of the superior colliculus. J Neurophysiol 69(6):1797–1809

302

Jack Brooks et al.

54. Skulimowski P, Strumillo P (2015) Verification of visual odometry algorithms with an OpenGL-based software tool. J Electron Imag 24(3):1–9 55. Caraiman S, Zvoristeanu O, Burlacu A, Herghelegiu P (2019) Stereo vision based sensory substitution for the visually impaired. Sensors 19(12):2771 56. Caraiman S, Morar A, Owczarek M, Burlacu A, Rzeszotarski D, Botezatu N et al (2017) Computer vision for the visually impaired: the sound of vision system. In: Proceedings of the IEEE international conference on computer vision workshops, p 1480–9 57. Spagnol S, Baldan S, Unnthorsson R (2017) Auditory depth map representations with a sensory substitution scheme based on synthetic fluid sounds. In: 2017 IEEE 19th international workshop on multimedia signal processing (MMSP). IEEE, pp 1–6 58. Stronks HC, Nau AC, Ibbotson MR, Barnes N (2015) The role of visual deprivation and experience on the performance of sensory substitution devices. Brain Res 1624:140–152 59. Proulx MJ, Stoerig P, Ludowig E, Knoll I (2008) Seeing ‘where’ through the ears: effects of learning-by-doing and long-term sensory deprivation on localization based on imageto-sound substitution. PLoS One 3(3):e1840 60. White BW, Saunders FA, Scadden L, Bach-YRita P, Collins CC (1970) Seeing with the skin. Percept Psychophys 7(1):23–27 61. Maravita A, McNeil J, Malhotra P, Greenwood R, Husain M, Driver J (2003) Prism adaptation can improve contralesional tactile perception in neglect. Neurology 60(11):1829–1831 62. Levy-Tzedek S, Maidenbaum S, Amedi A, Lackner J (2016) Aging and sensory substitution in a virtual navigation task. PLoS One 11(3):e0151593 63. Maidenbaum S, Amedi A (2015) Blind in a virtual world: mobility-training virtual reality games for users who are blind. IEEE Virtual Real 2015:341–342 64. Merabet LB, Pascual-Leone A (2010) Neural reorganization following sensory loss: the opportunity of change. Nat Rev Neurosci 11(1):44–52 65. Moldoveanu ADB, Ivascu S, Stanica I, Dascalu M, Lupu R, Ivanica G et al (2017) Mastering an advanced sensory substitution device for visually impaired through innovative

virtual training. In: 2017 IEEE 7th international conference on consumer electronics Berlin (ICCE-Berlin), p 120–5 66. Ricciardi E, Bonino D, Pellegrini S, Pietrini P (2014) Mind the blind brain to understand the sighted one! Is there a supramodal cortical functional architecture? Neurosci Biobehav Rev 41:64–77 67. Proulx MJ, Brown DJ, Pasqualotto A, Meijer P (2014) Multisensory perceptual learning and sensory substitution. Neurosci Biobehav Rev 41:16–25 68. Brooke J (1996) SUS: a quick and dirty usability scale. In: Jordan PW, Thomas B, McClelland IL, Weerdmeester B (eds) Usability evaluation in industry. CRC Press, London 69. Hartcher-O’Brien J, Auvray M (2014) The process of distal attribution illuminated through studies of sensory substitution. Multisens Res 27(5–6):421–441 70. Maravita A, Spence C, Driver J (2003) Multisensory integration and the body schema: close to hand and within reach. Curr Biol 13(13): R531–R539 71. Maravita A, Husain M, Clarke K, Driver J (2001) Reaching with a tool extends visual– tactile interactions into far space: evidence from cross-modal extinction. Neuropsychologia 39(6):580–585 72. Nagel SK, Carl C, Kringe T, M€artin R, Ko¨nig P (2005) Beyond sensory substitution—learning the sixth sense. J Neural Eng 2(4):R13–R26 73. Hart SG, Staveland LE (1988) Development of NASA-TLX (Task Load Index): results of empirical and theoretical research. In: Advances in psychology, vol 52. North Holland, pp 139–183 74. ISO DIS 17488 (2015) Road Vehicles - Transport information and control systems – Detection Response Task (DRT) for assessing selective attention in driving. Draft International Standard, ISO TC 22/SC39/WG8 75. Saitis C, Parvez MZ, Kalimeri K (2018) Cognitive load assessment from EEG and peripheral biosignals for the design of visually impaired mobility aids. Wirel Commun Mob Comput 2018:8971206 76. Herghelegiu P, Burlacu A, Caraiman S (2017) Negative obstacle detection for wearable assistive devices for visually impaired. In: 2017 21st international conference on system theory, control and computing (ICSTCC), pp 564–570

Part IV The Somatosensory Nervous System

Chapter 15 Microneurography: Recordings from Single Neurons in Human Peripheral Nerves Rochelle Ackerley and Roger Holmes Watkins Abstract Over the last 50 years, the technique of microneurography has provided us with valuable insights into activity in human nerves. Microneurography allows us to record from the axons of single neurons in various peripheral nerves across the body. This approach has provided a wealth of information on peripheral nerve signals, especially the encoding of touch in mechanoreceptive afferents in humans. As these single-neuron recordings are performed in conscious, healthy human participants, their activity can be directly linked with the resultant perceptual processes. This chapter will focus on how to setup, use, and apply microneurography, to make single-unit recordings from mechanoreceptive afferents to study the sense of touch. Only a handful of laboratories in the world carry out single-unit microneurography, due to the many challenges faced when recording from human nerves (e.g., ethical, practical, mental), and it takes extensive training to become competent in the technique. The co-founder of the technique (Åke Vallbo) rightly states that the method is safe, as long as the experiments are performed with care and consideration, and he emphasizes that a highly considerate attitude is required during microneurography (Vallbo, J Neurophysiol 120:1415– 1427. https://doi.org/10.1152/jn.00933.2017, 2018). Overall, we will explore how microneurography developed into the practice it is today, how to use it safely and effectively, and provide avenues for future development and investigations with the technique. Key words Microneurography, Single unit, Mechanoreceptor, Mechanoreceptive afferent, Tactile, Touch, Human, Peripheral nerve

1

Introduction Microneurography is, in principle, a very simple technique, but in practice is quite demanding. (Åke Vallbo [1])

Microneurography began in 1965 at the University Hospital in Uppsala, Sweden, where Karl-Erik Hagbarth and Åke Vallbo established a technique where axonal activity in human peripheral nerves could be registered using percutaneously inserted, metal, needle electrodes [2, 3]. It was not the first time that human nerves had been recorded from, yet previous attempts involved other

305

306

Rochelle Ackerley and Roger Holmes Watkins

approaches, such as surgical nerve exposure [4] or using glass electrodes [5]. Although the technology for data acquisition and processing has advanced since these early recordings (e.g., computers), the practical aspects of the approach (electrodes, amplifiers) remain virtually unchanged. This is in part due to the attention to detail and dedication of these two scientists. They tested and refined the technique extensively on themselves, assessing various nerves and electrode types to be used for recording. The initial excitement of hearing their own neuronal activity—albeit it was an extremely weak signal—must have been extraordinary. Still to this day, microneurographers take great pleasure and privilege in listening to such responses. Hagbarth and Vallbo went through a long process of overcoming conceptual, medical, ethical, and technical obstacles including electrode types and insertion techniques, as covered in the comprehensive history documented by Vallbo [1]. Multi-unit microneurography (e.g., recording efferent sympathetic activity [6]) is conducted in a number of laboratories across the world, where the activity of many axons is recorded simultaneously, while only a small number of laboratories conduct the much more challenging approach of recording from a single axon. Microneurography is a highly manual skill; there are many technically demanding factors to consider, and it takes years to learn to practice it safely and effectively (see Note 4.1). Additionally, there is a human being at the end of your electrode, making interpersonal skills and effective communication with the participant critical to its success (see Note 4.1). Thorough training is essential in learning microneurography, where it is best to work for a number of years in an established group [1] to learn how to conduct the technique efficiently and safely, and there are always new situations and difficulties that are encountered. It is a “clean” procedure, but it is not conducted under sterile conditions, as there is no surgery or cutting involved. Rather, a fine needle electrode is gently manually inserted through the skin of a participant by light pushing. This is usually pain-free, although occasionally, the electrode may generate a sensation of pressure on the skin, which occasionally causes some discomfort. This may be avoided by shifting the insertion site a few millimeters. Once the electrode has been successfully placed within a nerve, single-unit responses are searched for from individual myelinated Aβ or unmyelinated C-fiber axons (see Note 4.2). There are many steps to consider in a microneurography experiment, from a well-organized experimental setup to neural signal analysis (see Notes 4.1 and 4.3–4.9). There are also many stages where the technique can fail to generate usable data, which is one of the reasons why so few single-unit microneurography research groups exist. One issue, before starting any experiments, is to have ethical approval. This is an extremely important step, for the safety of both the participant and experimenter. As the skin barrier

Single Unit Microneurography

307

is crossed, microneurography is treated as an invasive approach that poses some level of risk (e.g., transmission of infection, tissue damage), which often entails high levels of ethical consideration, including medical surveillance, auditing (e.g., of consent forms, data collected), and insurance. Here we aim to outline the potential ethical and technical difficulties encountered during microneurography, as although it can produce insightful data, it is notoriously challenging. Many peripheral nerves can be accessed, yet some are easier to locate and achieve a stable intraneural position for recording (e.g., the median nerve at the wrist is superficial, relatively large, and at a clearly defined anatomical location see Notes 4.10 and 4.11). Deeper and smaller nerves, such as the antebrachial nerves innervating the forearm, can be located, but this is extremely challenging (see Note 4.11). Figure 1 shows a diagrammatic representation of

Fig. 1 Nerves typically accessed in single unit microneurography. The median nerve of the arm and the peroneal nerve of the leg commonly accessed in microneurography, as these are large nerves, relatively easy to find, and have many sensory afferents. Note that a lot of microneurography is carried out on the left side of the body and these locations are for demonstration. Example references for each nerve: facial (cranial nerve VI, abducens) [62], supraorbital [14], infraorbital [7, 63, 64], lingual [65–68], inferior alveolar [67–70], median—axilla [71], upper arm [37, 43, 72–77], elbow [78, 79], wrist [33, 53, 54, 71, 76, 77, 80], ulnar— upper arm [26, 72, 81], elbow [28], wrist [71, 80, 81], radial—upper arm [22, 57, 82–86], superficial branch at the wrist [28, 59, 76, 80, 87], lateral antebrachial [12, 13, 34, 36, 88–91], dorsal antebrachial [34, 36, 88, 89], medial antebrachial [72], lateral cutaneous femoral [15], common peroneal (fibular) [28, 57, 92–99], superficial peroneal [29], tibial [28, 98–101], posterior tibial [102], saphenous [103, 104], and sural [105, 106]

308

Rochelle Ackerley and Roger Holmes Watkins

many of the nerves that have been accessed in microneurography. Some nerves, such as the median and peroneal, are commonly used, but for others, such as the nerves in the face, only a few papers have documented recordings (e.g., [7]). Single-unit microneurography includes the recording of individual afferent mechanoreceptive, thermoreceptive, and nociceptive signals, as well as muscle and joint afferents (see also Chapters 1, 2, 6, 7, 9, 10, and 16 of this volume). Even some single-unit efferent sympathetic [8, 9] and motor [10] recordings have been published. However, the majority of single-unit studies have focused on cutaneous mechanoreceptive afferents and the sense of touch in the hand, which is predominantly conveyed by thickly myelinated Aβ axons (see Note 4.2, for a classic review, see [11]), although mechanoreceptors in hairy skin have also been investigated [12–15]. There is some single-unit work on cutaneous thermoreceptive and nociceptive afferents, predominantly from unmyelinated C-fibers (for a review, see [16]) or occasionally thinly myelinated Aδ afferents [17, 18] (see Note 4.2). Further, research has also focused on subcutaneous muscle and joint afferent signals, using recordings from both the upper and lower limbs (e.g., [19– 26], for further information, see also [27]). One of the reasons that many studies have recorded from cutaneous mechanoreceptive afferents is that they are numerous, relatively easy to identify and maintain recordings from, and play an important role in interacting with our external environment in our sense of touch. Fundamentally, the practical approach to microneurography has not changed much over the years from its initial development in the 1960s [28], but there is potential for development. Nerves are typically found through experience in human anatomy (see Notes 4.4 and 4.9) and/or in using electrical skin stimulation to find the course of the nerve and the best access point in a given participant. One recent development is the use of ultrasoundguided microneurography, where an ultrasound system can be used to visualize the nerves before implantation and even during intraneural manipulation [29–32]. Accessing nerves in microneurography experiments can take many hours (see Note 4.7), but ultrasound can dramatically cut down this time, directly visualizing nerves and electrode implantation [29]. Further technical developments have been made in the recording apparatus, in low-amplitude nerve stimulation (even during neuroimaging) [33], and in the robotic stimulators compatible with nerve recording [19, 34–43]. The application of single-unit intraneural microstimulation (INMS), which is an extension of single-unit microneurography, has also driven the development of microneurography systems that may be used concurrently with neuroimaging [33]. Experiments in the 1980s showed that it was possible to selectively stimulate one single myelinated Aβ mechanoreceptive afferent axon, using a few

Single Unit Microneurography

309

microamps of current, and produce a ‘quantal’ tactile percept. It is possible to demonstrate that a quantal evoked sensation can be linked to the highly controlled stimulation of a single mechanoreceptive afferent, when appropriate methodology is used [44– 47]. Very few papers exist using single-unit INMS, and this technique has been questioned over the years [48–50], but it continues to be used and has been extended into investigating not only the perceptual quality of the quantal sensation evoked (e.g., vibration in a fast-adapting type I afferent, or pressure in a slowly adapting type I afferent), but also to image brain responses. Initial studies combined electroencephalography (EEG, Chapter 19, this volume) with single-unit INMS [51] and 3 Tesla (T) functional magnetic resonance imaging (fMRI, Chapter 18, this volume) with singleunit INMS [52]. These studies paved the way for further investigations using magnetoencephalography (MEG) [33, 53] and ultrahigh field 7T fMRI [33, 54]. These neuroimaging environments are challenging to conduct INMS in. Strong magnetic fields require minimizing the use of metal in the fMRI scanner and dealing with problems associated with long cables. The MEG environment has a surprisingly high level of electrical interference, mitigated only by careful positioning of participants. These challenges led to the development of a system by Glover and colleagues [33] that allows high-quality and high-resolution single-unit recordings, low-current intraneural stimulation, and many added safety features. Although detailed information about single-unit INMS is beyond the remit of the current chapter, this is an extension of the recording technique detailed here, and the above references provide information on the equipment and approach required for such experiments (see Chapter 16, this volume). Many microneurography laboratories have relied on homebuilt hardware, to amplify, filter, and acquire signals; however, these systems are now many years old and safety requirements are more stringent, meaning that new systems are required. There are few options to purchase a commercially available microneurography system, but the NeuroAmp EX (AD Instruments; New South Wales, Australia) is available, and many multi-unit microneurographers use the Nerve Traffic Analyzer from Absolute Design (Solon, IA; formerly Iowa Biosystems). The SC/Zoom system, used for numerous experiments in Sweden, was developed at the University of Umea˚, but it is not available commercially. Other openly available options include using a modified Open Ephys system (Open Ephys initiative, https://open-ephys.org [29]) or a system with custom-designed amplifiers and stimulators initially constructed to use during microneurography with neuroimaging [33]. These systems are currently being combined and adapted for various types of open microneurography by the authors, as previous systems have had somewhat limited functionality for these kinds of specialist experiments. Making these systems standardized, simple to use in

310

Rochelle Ackerley and Roger Holmes Watkins

combination with ultrasound, and more widely available should help increase adoption of the technique, by flattening the learning curve and decreasing the technical demands.

2

Materials

2.1 Experimental Setup

• An adjustable, comfortable chair (e.g., a dental chair), preferably reclinable (see Note 4.6). • A sturdy, solid frame to stably support the investigated body region (e.g., arm/leg rest), supplemented with padding and vacuum cushion supports as appropriate. • Preparation equipment: disinfectant (e.g., 70% alcohol, which may be combined with other substances, such as 2% chlorhexidine or iodine), tissue, small compresses, skin tape (~1 cm diameter), very fine marker pens. • Means of electrode sterilization, for example by dry heat, gas, or UV light. • Environmental requirements (see Note 4.12). Microneurography is best conducted in a temperature-controlled environment, under standard conditions (e.g., ~22 °C). This aids in keeping the participant and their skin at a comfortable temperature. • Electrical shielding. It is ideal to conduct microneurography in a shielded room (e.g., metal in the walls to provide a Faraday cage) with good mains grounding. This helps minimize external electromagnetic interference, as small signal amplitudes are recorded in microneurography. This is not a strict requirement, and microneurography is possible in more “hostile” electrical environments, including in MEG or MRI scanners where there are more sources of interference [33, 52–54]. However, interference from these and from other nearby equipment may be unpredictable and make identification of single-unit recordings more challenging due to masking of the nerve signals.

2.2 Experimental Equipment 2.2.1

Electrodes

An insulated, high impedance, active recording electrode (length dependent on the nerve investigated (see Notes 4.9–4.11) is required, in conjunction with an uninsulated electrode as a reference (~3 cm long) that is inserted subcutaneously nearby (~5 cm away). An uninsulated stimulating electrode which is the same length as the recording electrode may be useful during an electrical search for the nerve, where the identified nerve depth can be a precise reference for the recording electrode insertion. This approach can help find deep nerves, where the nerve is initially located via this stimulating electrode (using up to 2 mA current), then the recording electrode is inserted nearby, parallel and slightly distal to this, which avoids searching using high current stimulation

Single Unit Microneurography

311

with the recording electrode (typically limited at 250 μA), which decreases its impedance if used repetitively. FHC (Bowdoin, ME) is the only commercial maker of microneurography electrodes, although earlier experiments used homemade electrodes [1]. Electrodes can be bought with specific requirements (e.g., length, pin termination, impedance), but it is also possible to modify them by cutting to size and attaching pin terminations. Even if they are sterilized, it is not recommended to re-use electrodes between participants. 2.2.2 Microneurography System

The nerve activity should be amplified and filtered, with an audio output of the signal sent to speakers and/or on-ear headphones (which are useful to hear the activity clearly; see Note 4.12) and the signal displayed via recording hardware/software that allows realtime processing and visualization of signals, along with data saving. The microneurography system should optionally incorporate an electrical stimulator for nerve localization or intraneural stimulation [33, 55see below]. Participant safety: As with other physiological recording and stimulation techniques (e.g., EEG, EMG, or brain stimulation Chapters 19 and 20, this volume), participant isolation is important in microneurography. Participants should always be adequately electrically isolated from the external sampling or stimulation equipment. This can be done, for example, by optically isolating signals at the preamplifier head stage (e.g., in the SC/Zoom system), by isolating and battery-powering the entire recording system [33], or by having a specific optically isolated output of the nerve signal in the main amplifier [55]. Grounding: As in standard electrophysiological methods, such electromyography (EMG), a participant ground is used to reduce electromagnetic interference, such as via a silver plate next to the skin with a saline-soaked tissue or a gel-based pad adhesive electrode. Additionally, it is important that the experimenter, who will be touching the electrodes and the participant, is connected to the participant ground, to avoid the possibility of static shocks. Amplification: Microneurography signals are very low amplitude. It is uncommon to see nerve signals in excess of 50 μV peak amplitude. Signals must therefore be amplified significantly. A signal gain of more than 100,000 times, usually at 10 μV/V, is necessary to visualize potentials on most recording systems. Filtering: Filtering allows the effective removal of external interference with minimal effects on nerve signals. Bandpass filtering typically between 0.2 and 10 kHz is generally adequate, and a typical bandpass range is 0.3–3 kHz. Additional notch filters at the mains supply frequency (50/60 Hz) may be useful when the system is used in combination with mains-powered equipment. Filtering should ideally be implemented in hardware for real-time output [55]. Alternatively, this can be performed in software [33], but

312

Rochelle Ackerley and Roger Holmes Watkins

precise, real-time output is difficult to achieve, which can make listening to and visualizing the signals challenging, as a small, yet noticeable, delay is introduced. Audio: An audio system of sufficient quality to hear the activity in real time greatly helps with identification of nerve signals and with identifying/troubleshooting interference. Using high-fidelity audio equipment (speakers or headphones; see Note 4.12), nerve activity (high-frequency “whooshing”), electrical interference (commonly 50/60 Hz mains hum), and concomitant motor unit activity (2–5 Hz “thumping” noise) are easily and readily distinguishable by an experienced experimenter, more simply than by visualization alone. Additionally, when performing a detailed characterization and mapping of receptive fields, it is not possible to simultaneously watch a display. Signal visualization: It is advantageous to display and process signals in real time, in addition to having the audio signal. On-line processing can be used to detect individual spikes based on their shape and amplitude. A simple method for doing this is using a histogram of local peak amplitudes [56]. This allows a simple effective display of the goodness-of-separation from the noise floor or between afferents. After appropriate thresholding of spike activity, a triggered window showing the detected spikes can be useful to distinguish between different afferents. If multiple afferents are present, these may differ slightly in shape or amplitude, but this difference may not be audible. Unlike cortical neuronal recordings, where spike shapes may differ significantly between neurons, in peripheral nerves, this may not always be the case. Additional signs can be used to identify of multiple afferents in a recording, including the amplitude summation that can occur when multiple afferents fire synchronously or detected spikes firing at non-physiological instantaneous firing rates with intermittent extremely high-frequency instantaneous spikes (>500 Hz). These signs, along with the identification of different receptive properties (e.g., adaptation, threshold) or spatial separation of receptive fields can help identify different afferents that may not be easily separable during the analysis of the signals. Special care should be taken to note indicators of the presence of any additional afferents, even if the amplitude or shape separation seems sufficient, since this can be invaluable during the analysis. Data acquisition: After amplification and filtering, nerve signals are normally saved digitally. A sampling frequency of >10 kHz is adequate for visualizing and differentiating impulses, since the recordings are only used as a method to time-stamp the spikes and produce a binary output of the response. An important point to consider is the simultaneous acquisition of other signals alongside nerve activity. Nerve signals can have sub-millisecond timing accuracy, and this accuracy may be necessary for some stimulus-

Single Unit Microneurography

313

linked sampling. Many acquisition systems are capable of performing such simultaneous acquisition of external signals alongside the nerve activity at sufficient frequencies, including Spike2 (CED, Cambridge, UK), SC/Zoom, LabChart (AD Instruments; New South Wales, Australia), Nerve Traffic Analyzer, National Instruments DAQ with LabView (NI, Austin, TX) MATLAB (Mathworks, Natick, MA) [53] or Python, and Open Ephys [29]. 2.2.3 Equipment for Finding the Nerve

Extraneural electrical stimulation: A constant current stimulator (up to 2 mA) that is suitable for nerve localization (i.e., finding the approximate trajectory of the nerve) may be used, either via transcutaneous electrodes or a percutaneously uninsulated needle electrode. This can either be integrated with the microneurography amplifier (e.g., SC/Zoom system) or use an external stimulator (e.g., AD Instruments). Such a large current is too great to use to achieve nerve penetration, as this could be painful if applied near the nerve. Therefore, it is advisable to have another means of electrical stimulation with a maximum output of 250 μA to avoid this possibility. This lower range of electrical stimulation can be used with a percutaneously implanted electrode (an uninsulated stimulation electrode or the insulated recording electrode); however, great care should be taken when using electrical stimulation close to the nerve. Search currents that are required for extraneural stimulation with high-impedance recording electrodes require high voltages (up to 50 V) or may not be deliverable, so caution should be taken in system design and use. An important principle always to adhere to is to decrease the stimulation current to 0 μA before starting stimulation, then increase this gradually to the point of sensation. The window between no activation of any nerve fibers and the recruitment of the majority of the nerve can be small. If the electrodes (needle or transcutaneous) are moved while stimulating, this could lead to a sudden, surprising, and sometimes painful recruitment of a large number of fibers. When using the needle electrodes special care should be taken, since current thresholds can vary more substantially than transcutaneous stimuli and participant movement from a shock will be more problematic. What is a barely perceivable stimulus extra-neurally (80–100 μA) would be intolerable intraneurally and this risks involuntary contractions and potential nerve damage. Additionally, for safety, the system should be easily controllable/reducible or switched off by the experimenter. Intraneural electrical stimulation: An intraneural stimulation system may be used to perform INMS, within the nerve. This system should be designed to deliver a small current (up to ~10 μA) and special care needs to be taken in its design. The system must have the capability to switch between recording and stimulating modes, easily, quickly, and safely, with full control to reduce or turn off the stimulation by the experimenter. It is possible to have a

314

Rochelle Ackerley and Roger Holmes Watkins

system that is suitable for both nerve search (up to ~250 μA) and intraneural stimulation [46]. However, because of this wide range in current thresholds, such systems need to be well-designed to safely switch between these ranges depending on the intended stimulus. Electrical artifacts can be generated from switching between recording and stimulation [33], which can be painful for the participant when in the nerve. Precautions can be taken to avoid this, such as “soft” analog switches that gradually switch the electrodes from the amplifier to stimulator. In general, using highimpedance electrodes for recording single-unit activity requires voltages of 25 kHz) to represent each spike snippet with enough points (>80) for reliable sorting. Chronic recording sessions are particularly prone to mains interference (50 or 60 Hz). As such, the behavioral chamber should be within a Faraday cage and grounded well. It is usually advised to bring all power lines to one wall socket to prevent ground loops [175]. Additionally, unnecessary shielding or hook-up wires should be avoided on various pieces of equipment. Reducing noise is more art than science, and requires patience and an iterative approach. On certain days, SNR will be lower than usual because of unknown electromagnetic interference from the environment. It is important to note that almost all shielding techniques are for electric fields, but not magnetic fields, which indeed appear concurrently and are very difficult to avoid. Spike sorting can be performed online or offline, depending on the purpose of the experiment. For example, a BMI application would require quick online sorting. The first step for sorting is to identify candidate spikes. This is usually done by thresholding. Artifact

Electrophysiology in Rats

355

rejection often accompanies this initial step, as too large waveforms are probably due to noise in the headstage cable or perhaps due to static discharge as the animal subject touches a metal surface. It is also not unlikely to pick up noise due to muscle activity, especially when rodents chew rigorously. Some of the movement artifacts can be reduced by band-pass filtering the amplified voltage traces in the software. The candidate spike waveforms are analyzed based on a predefined data point before and after threshold crossing. There are several sorting techniques, using either principal component analysis (PCA) [176], template matching [177], wavelet transforms [178] or superparamagnetic clustering [179]. Over the years, we have tested plenty of algorithms based on those techniques, such as the offline sorter by Plexon, MountainSort [180], Klusta [181], KiloSort [182], and WaveClus [179, 183]. At best, we have found up to 4 units isolated from each electrode channel. It has also been observed, however, that with the default settings, many sorted units actually come out as multi-units (i.e., containing spikes from different neurons), because some inter-spike time intervals are shorter than the refractory period. Depending on the purpose of the study, multi-unit spikes may be adequate; otherwise, further criteria may be used to identify single-unit spikes. Spike waveform snippets form clusters in the feature space. Many metrics exist to isolate these clusters to obtain single-unit data. For example, a higher Bayesian Information Criterion (BIC) value means a higher likelihood of correctly classified data points in a cluster [184]. The F statistic of multivariate analysis of variance (MANOVA), J1 (compactness), J2 (separation), and J3 (well-separated and compact clusters) are other metrics used to accept or reject a cluster [155, 185]. Some spikes are also left out as outliers, that is, those that do not belong to any cluster. Once spike sorting is complete, spike times are extracted (usually at the time of the spike peak or trough) for each single- or multi-unit dataset. Next, basic analyses based on spike times and counts proceed as discussed previously. Raster plots (Fig. 5g), PSTHs (Fig. 5h), and SPHs (Fig. 5i) are standard methods to study vibrotactile responses in the S1 cortex. The cortical recording shown in Fig. 5 is from an awake behaving rat and the spikes are from a multiunit. Psychophysical Task

Psychophysical tasks are rather important to understand tactile perception in rodents. For rats, we have mainly focused on stimulating the glabrous skin of the hindpaws (Fig. 2b), in which the mechanoreceptors are similar to those in humans, and it is easier to apply well-controlled stimuli when the animal’s head is fixed in the stereotaxic frame. However, we have also developed a conditioning chamber to stimulate the glabrous skin of forepaws and hindpaws during psychophysical tasks performed by rats with electrode implants [22]. There are many behavioral paradigms in the

356

Sevgi O¨ztu¨rk et al.

Fig. 7 Psychophysical task and cortical recording in the awake rat. The animal performs a yes/no detection task in a conditioning chamber while cortical activity is recorded by a multichannel electrode. The rat has started the current trial by pressing the middle lever (ML) (8). The LED light in the chamber is lit, and this is signaled on the computer screen as a red symbol (2). Depending on the presentation of a vibrotactile stimulus from the bottom of the chamber, the rat presses either the right lever (RL) (9) for detecting or the left lever (LL) (7) for not detecting the stimulus. In the figure, the rat has pressed RL (9) and a green symbol (6) appears on the computer screen for that lever press. Computer screen indicators: (1) left LED, (2) middle LED, (3) right LED, (4) LL, (5) ML, (6) RL; operant chamber: (7) LL, (8) ML, (9) RL, (10) headstage cable

literature where recording and/or stimulation are performed in awake rodents. For example, in a recent study, recording was obtained from mice during a visual discrimination task to predict action, engagement, and choice [186]. In Fig. 7, we describe the chamber controls used in a yes/no stimulus detection task. The rat starts a trial by pressing the middle lever (8), and an LED light is lit in the chamber, which is signaled on the computer screen as a red symbol (2). Depending on the presentation of a vibrotactile stimulus from the bottom of the chamber, the rat presses either the right lever (RL) (9) for detecting or the left lever (LL) (7) for not detecting the stimulus. In Fig. 7, the rat presses RL (9) and a green symbol (6) appears on the computer screen for that lever press. For a stimulus-on trial,

Electrophysiology in Rats

357

pressing RL (9) is a hit and pressing LL (7) is a miss. For a stimulusoff trial, pressing RL and LL are false alarm and correct rejection, respectively. These behavioral outcomes are analyzed to assess sensitivity indices (i.e., d’ or A’) based on signal detection theory. The headstage and its cable (10) are also seen in Fig. 7. For correct responses, the rat is rewarded by water given by the spout under the middle lever (8). The pattern of dots on the ground plate are the holes where the probes of the electrodynamic shaker pass to stimulate the paws. Because of the animal’s posture, mostly the hindpaw soles receive the vibrotactile stimuli. As can be seen in the photograph, the animal is not restrained or fixed during this task. It voluntarily starts each trial. However, the rat’s movements are continuously recorded by an infrared camera to verify the time periods where it chooses not to continue the task or to watch out for technical problems such as a detached or twisted cable. The neural data are also continuously streamed to the computer hard drive and displayed on screen. 3.4 Electrical Microstimulation of Neural Tissue

Current-controlled charge injection is the most widespread method used in electrical stimulation of the nervous system [187]. In this method, the charge is delivered to the tissue, usually by current pulses. The current level (measured in μA or mA) is monitored and kept constant during a pulse. Either monophasic or biphasic pulses have been used in the literature [187]. In monophasic stimulation, either negative (cathodic) or positive (anodic) pulses are delivered to the tissue (Fig. 8a). However, this causes accumulation of products of Faradaic reactions because these reactions are mostly irreversible (see above). On the other hand, both consecutive cathodic and anodic pulses are delivered to the tissue in the biphasic stimulation (Fig. 8b). In this way, some of the Faradaic reactions that occurred during the first phase may be reversed by proper balancing. Yet, this may reduce the efficiency of the stimulation in triggering spikes compared to the monophasic stimulation [117], because the second phase reduces the effects of the first phase by reversing the flow of some ions in the medium. In order to prevent this, the amount of current delivered in the second phase can be reduced but presented with a longer duration (Fig. 8c). Alternatively, a brief delay can be placed between the first phase and the second phase in order to let the effects of the first phase take place (Fig. 8d). The latter two techniques can also be combined. The efficiency of the current pulses to trigger action potentials in neurons decreases as the distance between the electrode and the neuron increases [117]. However, increasing the current level above a limit is not desired because of the possible damage to the tissue. The safety limits for the current pulse amplitude and width can be predicted with the Shannon equation, which is applicable for surface electrodes (mainly of Pt) of certain diameter range (see Note 4.3 for further discussion of stimulation intensity). However, this

358

Sevgi O¨ztu¨rk et al.

Fig. 8 Commonly used pulse patterns for electrical microstimulation. (a) Monophasic pulses. (b) Biphasic pulses. (c) Asymmetric pulses. (d) Pulses with an interphase interval/delay

equation is mainly empirical and also does not incorporate material properties. Therefore, assessing long-term safety requires the critical knowledge of electrochemistry of the electrode-tissue interface. Lapicque’s empirical equation allows one to find a threshold of nerve excitation around the vicinity of the electrode by co-varying current pulse amplitude and width. The required pulse amplitude decreases as the width increases until a minimum, called rheobase, is reached [188]. For the peripheral nerves, it was found that increasing the frequency of brief stimulation pulses increases the spike rate in the excited fiber population in the nerve. By contrast, increasing the pulse width recruits more fibers. It is also well known that largediameter fibers are first recruited as the pulse amplitude is increased. Non-myelinated, small-diameter fibers require the highest pulse amplitudes for excitation.

Electrophysiology in Rats

359

The electrical and geometrical properties of the excitable elements are very different in the cortex given the elaborate structural and functional connectivity. Intracortical microstimulation (ICMS) has been successfully used in basic neuroscience experiments and neural engineering applications to induce excitation or inhibition in the networks. In some studies, ICMS is applied through one of the electrodes in the array, while other electrodes record the neural activity [189]. It is possible to stimulate and record from the same electrode, but this requires special headstages which decouple the amplifier or the stimulator circuits while one is active [190, 191]. Another technical issue is the electrical artifact of the ICMS on the recorded signal. The neural activity is usually lost during the stimulation window because the artifact voltages are too high in neighboring recording electrodes. Special electronic precautions should be taken not to saturate the amplifiers of the neighboring electrode channels. Therefore, the neural activity is typically studied in pre- and post-stimulation windows. ICMS generally elicits an unnatural synchronous population activity which is followed by a period with no neural activity [189]. Additionally, elongated ICMS (i.e., between 15 min and 2 or more hours) may alter the network structure by means of plasticity [192–194]. The network is thought to recover back to its initial connectivity when the ICMS is ceased. When S1 cortex neurons are excited by ICMS in awake subjects, they may experience artificial tactile sensations. Although the actual perception of the animal is not known, this novel sensory information can be used in behavioral tasks. Rodents can be conditioned to respond (i.e., licking a water spout or pressing a lever) when the ICMS-cue is present. The detection threshold changes with ICMS parameters [187]. Additionally, if an animal was conditioned with a natural stimulus (i.e., mechanical vibration), then it can quickly generalize the task on the information from the new artificial sensation elicited by ICMS [33]. Therefore, the artificial senses elicited by ICMS can be studied effectively in a psychophysical framework. For this purpose, different conditioning strategies have been used in the literature [32, 33, 187, 195, 196]. For example, the rodent can be first trained to lick a water spout when a mechanical stimulus (i.e., whisker movement or mechanical vibration on skin) is presented. Then, the mechanical stimulus is replaced with ICMS pulse trains. Although the response rate may drop in the first sessions with ICMS, it generally recovers back to levels achieved with a natural stimulus. If the response of the animal (i.e., corrected hit rate or A’ (nonparametric sensitivity index)) is recorded with varying amplitude levels of the mechanical stimulus and ICMS, a psychometric equivalence can be established mathematically between the parameters of the two stimulation modalities [33].

360

Sevgi O¨ztu¨rk et al.

3.5 Microinjection of Drugs into the Somatosensory Cortex

Two main methods of microinjection, iontophoresis and pressure injection, were previously described. Both have been used for more than 70 years [197–200]. Iontophoresis was originally used for neuromuscular injection, then later developed to deliver drugs into local regions of the brain, including somatosensory cortex [201, 202]. There are four issues which should be considered case-by-case for every experiment: (1) the effects of direct application to the vicinity of neurons to prevent diffusion barriers and limiting biochemical effects (i.e., enzymatic breakdown), (2) precise control of drug volume injected, (3) making a smaller assembly to prevent damage in the brain, and (4) a suitable electrode to ensure better recording quality. In microiontophoresis, charged compounds drift down from the pipettes by applied voltage gradients and flow into the surrounding area. Anodic iontophoresis is used for delivery of positively charged molecules, whereas cathodic iontophoresis is for negatively charged compounds [203]. Although there are several advantages of microiontophoresis as a delivery method, such as rapid application time, low solution volume, bypassing bloodbrain barrier, and specific delivery regions [204], ejection quantities are difficult to predict. It is possible to monitor ejection volume by using a voltammetric microelectrode assembled together to measure passing current [205]. There are also theoretical approaches to estimate the ejection quantity [127, 129, 206, 207]. In scientific articles, in addition to the initial concentrations of the chemicals in the pipettes, iontophoresis current and duration are reported to enable comparisons between studies. In pressure microinjection, it is possible to determine ejected volumes with high accuracy after calibration. Any soluble substances (not necessarily charged) can be ejected into the desired regions of the brain by means of air compression or mechanical pressure [124]. Apart from experiments with anesthesia, this method is also used with awake animals, both extracellularly [208] and intracellularly [209, 210]. Glass micropipettes of the combination electrode assembly are connected to a pneumatic pump via soft catheter tubing and high-pressure tubing. Additionally, a valve manifold is required to route the pump pressure to individual barrels as desired in the experiment. Pulse duration and pressure can be set for optimum results. In our laboratory, we adjust the valves for the selected barrels during an experiment according to the randomization of drug conditions previously planned. Since there are slight differences in the initial volumes of the solutions in different barrels and air volume in the tubing, each barrel needs to be calibrated before the experiment to measure the amount of drug delivered (Fig. 4g). For this purpose, several hundred injection pulses can be applied while the combination electrode tip is immersed in saline, and the weight change of the filled barrel can be measured by a precision balance afterward.

Electrophysiology in Rats

361

Based on the lost volume and initial concentration of the chemical, the amount delivered by each pulse can be calculated. It is critical to have good connections along the tubing line. In addition to tight fittings, we found that it is convenient to use hot silicone glue for airtight connections. Pressure leaks not only disrupt the ejection but cause mechanical artifacts during electrophysiological recording. For both microinjection methods to be reliable, the following criteria need to be met [129, 199]: The pH of all substances should be close to the physiological range so as not to disrupt the homeostasis of the brain area. (we typically dissolve drugs in artificial cerebral spinal fluid (aCSF) adjusted to pH 7.4); drug concentrations should be dilute, preferably in micromolar range; low volumes should be applied to prevent the accumulation of the drug and large changes in osmolarity. See Note 4.4 includes further discussion about the recording during microinjection.

4

Notes This chapter covered commonly used electrophysiological techniques to study tactile perception in rodents, especially in rats. There are common pitfalls which warrant further discussion in this section.

4.1 Placement of Mechanical Stimulators on the Target Organ

Primary afferents innervating the whiskers are sensitive to direction of whisker movement. Cutaneous afferents in the glabrous skin, by contrast, are mostly stimulated with mechanical stimulation perpendicular to the skin surface. In any case, skin-probe coupling should be well maintained during the experiments; otherwise, the stimulus at the mechanoreceptor level will change considerably, confounding the results [38]. Due to its mechanical properties, the skin may not follow the movement of the contactor, especially for higher frequencies and high-amplitude levels, even if it is pre-indented. Some researchers have tried gluing the probe to the skin, but this approach interferes with the mechanical properties of the skin and may create unnatural stimuli (e.g., pulling the skin away) not usually observed during tactile exploration. Additionally, when stimulating the skin, the limb should not move. This can be achieved by fixing the lower limb with modeling clay. If a relative movement occurs, this effectively reduces the stimulation of mechanoreceptors.

4.2 Penetration of the Recording Electrodes

The most challenging part of recording from the cortex is the penetration of the electrodes after removing the dura. Capillary bleeding is inevitable, but puncturing larger surface vessels can be prevented by careful observation under a surgical microscope. Still, significant trauma is caused below the surface. If there is too much

362

Sevgi O¨ztu¨rk et al.

bleeding, the neurons will die and the activity will deteriorate quickly. Since there is large inter-subject variability in the brain anatomy, there is always a chance of misplacement of the electrodes. As opposed to classical microwire arrays, modern electrode designs such as Michigan probes cause less trauma per unit recording yield. It is also rather difficult to adjust each microwire in the array to avoid vessels during insertion. More expensive electrodes with on-board microdrives can mitigate the long-term effects of trauma and foreign body reaction by readjusting the depths of the recording sites after recovery. 4.3 Stimulation Intensity

ICMS in S1 or other cortical surface stimulation can initially be started at low charge densities. Then, depending on the animal’s behavior, it can be increased up to the material or Shannon limit [211] or up to the level of discomfort, whichever comes first. The upper limit and the lower psychophysical limit can specify the range of current pulse parameters, which can vary in an experiment. Unfortunately, establishing the lower psychophysical limit—sensory threshold—may be difficult because it requires the animal to learn a task and perform it with enough trials so that a psychometric function can be estimated. In some studies, a motor threshold can be established by observing a change in the animal’s behavior such as a brief stop of chewing.

4.4 Recording During Microinjection

In addition to the difficulty of estimating the amount of drug delivered, there is another drawback for microiontophoresis. Because of the electrical artifact produced by the applied current, it is not possible to record while injecting drugs into the cortex. Therefore, the activity is studied before and after injection. Furthermore, microiontophoresis may not be suitable for compounds with a low charge-to-mass ratio. In all microinjection methods, there may be some common problems, such as diffusion-based leakage from the tip of the micropipette, channel blockages, and physical displacement of surrounding tissue.

Acknowledgments € ˙ TAK Grant 117F481 within the This work was supported by TUBI European Union’s FLAG-ERA JTC 2017 project GRAFIN. References 1. Corniani G, Saal HP (2020) Tactile innervation densities across the whole body. J Neurophysiol 124:1229–1240. https://doi.org/ 10.1152/jn.00313.2020 2. Gu¨c¸lu¨ B (2021) Introduction to somatosensory neuroprostheses. In: Gu¨c¸lu¨ B

(ed) Somatosensory feedback for neuroprosthetics. Elsevier, pp 3–40 3. Handler A, Ginty DD (2021) The mechanosensory neurons of touch and their mechanisms of activation. Nat Rev Neurosci 22:521–

Electrophysiology in Rats 537. https://doi.org/10.1038/s41583-02100489-x 4. Middleton SJ, Perez-Sanchez J, Dawes JM (2021) The structure of sensory afferent compartments in health and disease. J Anat 241: 1186. https://doi.org/10.1111/joa.13544 5. Fox K, Woolsey T (2008) Barrel Cortex. Cambridge University Press, Cambridge 6. Rice FL, Kinnman E, Aldskogius H et al (1993) The innervation of the mystacial pad of the rat as revealed by PGP 9.5 immunofluorescence. J Comp Neurol 337:366–385. https://doi.org/10.1002/cne.903370303 7. Guzun L, Fortier-Poisson P, Langlais J-S, Smith AM (2021) Tactile sensitivity in the rat: a correlation between receptor structure and function. Exp Brain Res 239:3457– 3469. https://doi.org/10.1007/s00221021-06193-7 8. Wai V, Roberts L, Michaud J et al (2021) The anatomical distribution of mechanoreceptors in mouse hind paw skin and the influence of integrin α1β1 on meissner-like corpuscle density in the footpads. Front Neuroanat 15: 628711. https://doi.org/10.3389/fnana. 2021.628711 9. Coste B, Mathur J, Schmidt M et al (2010) Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels. Science 330:55–60. https://doi.org/10. 1126/science.1193270 10. Ebner FF, Kaas JH (2015) Somatosensory system. In: The rat nervous system. Elsevier, pp 675–701 11. Sachdev RN, Sellien H, Ebner FF (2000) Direct inhibition evoked by whisker stimulation in somatic sensory (SI) barrel field cortex of the awake rat. J Neurophysiol 84(3): 1497–1504 12. Martin C, Berwick J, Johnston D et al (2002) Optical imaging spectroscopy in the unanaesthetised rat. J Neurosci Methods 120: 25–34 13. Gener T, Reig R, Sanchez-Vives MV (2009) A new paradigm for the reversible blockage of whisker sensory transmission. J Neurosci Methods 176:63 14. Walker JL, Monjaraz-Fuentes F, Pedrow CR, Rector DM (2011) Precision rodent whisker stimulator with integrated servo-locked control and displacement measurement. J Neurosci Methods 196:20–30. https://doi.org/ 10.1016/j.jneumeth.2010.12.008 15. Krupa DJ, Brisben AJ, Nicolelis MA (2001) A multi-channel whisker stimulator for producing spatiotemporally complex tactile stimuli. J Neurosci Methods 104:199

363

16. Topchiy IA, Wood RM, Peterson B et al (2009) Conditioned lick behavior and evoked responses using whisker twitches in head restrained rats. Behav Brain Res 197:16–23. https://doi.org/10.1016/j.bbr.2008. 07.032 17. Rousche PJ, Petersen RS, Battiston S et al (1999) Examination of the spatial and temporal distribution of sensory cortical activity using a 100-electrode array. J Neurosci Methods 90:57–66 18. Ewert TA, Vahle-Hinz C, Engel AK (2008) High-frequency whisker vibration is encoded by phase-locked responses of neurons in the rat’s barrel cortex. J Neurosci 28:5359–5368 19. Tsytsarev V, Pope D, Pumbo E, Garver W (2010) Intrinsic optical imaging of directional selectivity in rat barrel cortex: application of a multidirectional magnetic whisker stimulator. J Neurosci Methods 189:80–83 20. Adibi M, Arabzadeh E (2011) A comparison of neuronal and behavioral detection and discrimination performances in rat whisker system. J Neurophysiol 105:356–365. https:// doi.org/10.1152/jn.00794.2010 21. Tahon K, Wijnants M, De Schutter E (2011) The RAT-ROTADRUM: a reaction time task depending on a continuous stream of tactile sensory information to the rat. J Neurosci Methods 200:153–163. https://doi.org/10. 1016/j.jneumeth.2011.06.031 22. Knutsen PM, Pietr M, Ahissar E (2006) Haptic object localization in the vibrissal system: behavior and performance. J Neurosci 26(33):8451–8464 23. Prigg T, Goldreich D, Carvell GE, Simons DJ (2002) Texture discrimination and unit recordings in the rat whisker/barrel system. Physiol Behav 77:671–675 24. Wiest MC, Thomson E, Pantoja J, Nicolelis MAL (2010) Changes in S1 neural responses during tactile discrimination learning. J Neurophysiol 104:300–312. https://doi.org/10. 1152/jn.00194.2010 25. Mehta SB, Whitmer D, Figueroa R et al (2007) Active spatial perception in the vibrissa scanning sensorimotor system. PLoS Biol 5: e15 26. Quairiaux C, Armstrong-James M, Welker E (2007) Modified sensory processing in the barrel cortex of the adult mouse after chronic whisker stimulation. J Neurophysiol 97(3): 2130–2147 27. Hulsey DR, Mian TM, Darrow MJ, Hays SA (2019) Quantitative assessment of cortical somatosensory digit representations after median and ulnar nerve injury in rats. Exp

364

Sevgi O¨ztu¨rk et al.

Brain Res 237:2297–2304. https://doi.org/ 10.1007/s00221-019-05593-0 28. Deveciog˘lu I˙, Gu¨c¸lu¨ B (2013) Asymmetric response properties of rapidly adapting mechanoreceptive fibers in the rat glabrous skin. Somatosens Mot Res 30:16–29. https://doi.org/10.3109/08990220.2012. 732128 29. Vardar B, Gu¨c¸lu¨ B (2020) Effects of basal forebrain stimulation on the vibrotactile responses of neurons from the hindpaw representation in the rat SI cortex. Brain Struct Funct 225:1761–1776. https://doi.org/10. 1007/s00429-020-02091-w 30. Hayashi A, Yoshida T, Ohki K (2018) Cell type specific representation of vibro-tactile stimuli in the mouse primary somatosensory cortex. Front Neural Circuits 12:109. https://doi.org/10.3389/fncir.2018.00109 31. Morandell K, Huber D (2017) The role of forelimb motor cortex areas in goal directed action in mice. Sci Rep 7(1):15759. https:// doi.org/10.1038/s41598-017-15835-2 32. Deveciog˘lu I˙, Gu¨c¸lu¨ B (2015) A novel vibrotactile system for stimulating the glabrous skin of awake freely behaving rats during operant conditioning. J Neurosci Methods 242:41– 51. https://doi.org/10.1016/j.jneumeth. 2015.01.004 33. Deveciog˘lu I˙, Gu¨c¸lu¨ B (2017) Psychophysical correspondence between vibrotactile intensity and intracortical microstimulation for tactile neuroprostheses in rats. J Neural Eng 14(1): 016010. https://doi.org/10.1088/17412552/14/1/016010 ¨ zturk S, Devecioglu I, Beygi M et al (2019) 34. O Real-time performance of a tactile neuroprosthesis on awake behaving rats. IEEE Trans Neural Syst Rehabil Eng 27:1053– 1062. https://doi.org/10.1109/TNSRE. 2019.2910320 35. Lechner SG, Lewin GR (2013) Hairy sensation. Physiology 28:142–150. https://doi. org/10.1152/physiol.00059.2012 36. Foffani G, Tutunculer B, Moxon KA (2004) Role of spike timing in the forelimb somatosensory cortex of the rat. J Neurosci 24:7266– 7 2 7 1 . h t t p s : // d o i . o r g / 1 0 . 1 5 2 3 / JNEUROSCI.2523-04.2004 37. Yıldız MZ, Gu¨c¸lu¨ B (2013) Relationship between vibrotactile detection threshold in the Pacinian channel and complex mechanical modulus of the human glabrous skin. Somatosens Mot Res 30:37–47. https://doi.org/ 10.3109/08990220.2012.754754

38. Cohen JC, Makous JC, Bolanowski SJ (1999) Under which conditions do the skin and probe decouple during sinusoidal vibrations? Exp Brain Res 129:211–217. https://doi. org/10.1007/s002210050891 39. Furuta T, Bush NE, Yang AE-T et al (2020) The cellular and mechanical basis for response characteristics of identified primary afferents in the rat vibrissal system. Curr Biol 30:815– 826. https://doi.org/10.1016/j.cub.2019. 12.068 40. Elyahoodayan S, Larson C, Cobo AM et al (2020) Acute in vivo testing of a polymer cuff electrode with integrated microfluidic channels for stimulation, recording, and drug delivery on rat sciatic nerve. J Neurosci Methods 336:108634 41. Mathews KS, Wark HA, Normann RA (2014) Assessment of rat sciatic nerve function following acute implantation of high density utah slanted electrode array (25 electrodes/ mm2) based on neural recordings and evoked muscle activity. Muscle Nerve 50(3):417–424 42. del Valle J, Rodrı´guez-Meana B, Navarro X (2021) Neural electrodes for long-term tissue interfaces. In: Gu¨c¸lu¨ B (ed) Somatosensory feedback for neuroprosthetics. Elsevier, pp 509–536 43. Leem JW, Willis WD, Chung JM (1993) Cutaneous sensory receptors in the rat foot. J Neurophysiol 69:1684–1699. https://doi. org/10.1152/jn.1993.69.5.1684 44. Yeager JD, Phillips DJ, Rector DM, Bahr DF (2008) Characterization of flexible ECoG electrode arrays for chronic recording in awake rats. J Neurosci Methods 173:279– 285. https://doi.org/10.1016/j.jneumeth. 2008.06.024 45. Spinks RL, Kraskov A, Brochier T et al (2008) Selectivity for grasp in local field potential and single neuron activity recorded simultaneously from M1 and F5 in the awake macaque monkey. J Neurosci 28:10961– 1 0 9 7 1 . h t t p s : // d o i . o r g / 1 0 . 1 5 2 3 / JNEUROSCI.1956-08.2008 46. Linde´n H, Tetzlaff T, Potjans TC et al (2011) Modeling the spatial reach of the LFP. Neuron 72:859. https://doi.org/10.1016/j.neu ron.2011.11.006 47. Buzsa´ki G, Anastassiou CA, Koch C (2012) The origin of extracellular fields and currents – EEG, ECoG, LFP and spikes. Nat Rev Neurosci 13:407–420 48. Watson BO, Ding M, Buzsa´ki G (2018) Temporal coupling of field potentials and action potentials in the neocortex. Eur J Neurosci

Electrophysiology in Rats 48:2482–2497. https://doi.org/10.1111/ ejn.13807 49. Burle B, Spieser L, Roger C et al (2015) Spatial and temporal resolutions of EEG: is it really black and white? A scalp current density view. Int J Psychophysiol 97:210–220. https://doi.org/10.1016/j.ijpsycho.2015. 05.004 50. Nunez PL, Srinivasan R (2006) Electric fields of the brain: the neurophysics of EEG. Oxford University Press, Oxford 51. Leeb R, Tonin L, Rohm M et al (2015) Towards independence: a BCI telepresence robot for people with severe motor disabilities. Proc IEEE 103(6):969–982. https:// doi.org/10.1109/JPROC.2015.2419736 52. Sellers EW, Vaughan TM, Wolpaw JR (2010) A brain-computer interface for long-term independent home use. Amyotroph Lateral Scler 11(5):449–455. https://doi.org/10. 3109/17482961003777470 53. Lesenfants D, Habbal D, Lugo Z et al (2014) An independent SSVEP-based brain-computer interface in locked-in syndrome. J Neural Eng 11(3):035002. https://doi.org/10. 1088/1741-2560/11/3/035002 54. Li Y, Long J, Yu T et al (2010) An EEG-based BCI system for 2-D cursor control by combining Mu/Beta rhythm and P300 potential. IEEE Trans Biomed Eng 57(10):2495–2505. https://doi.org/10.1109/TBME.2010. 2055564 55. De Vos M, Kroesen M, Emkes R, Debener S (2014) P300 speller BCI with a mobile EEG system: comparison to a traditional amplifier. J Neural Eng 11(3):036008. https://doi. org/10.1088/1741-2560/11/3/036008 56. Chauvette S, Soltani S, Seigneur J, Timofeev I (2016) In vivo models of cortical acquired epilepsy. J Neurosci Methods 260:185–201 57. Timofeev I, Steriade M (2004) Neocortical seizures: initiation, development and cessation. Neuroscience 123:299–336. https:// doi.org/10.1016/j.neuroscience.2003. 08.051 58. Yang T, Hakimian S, Schwartz TH (2014) Intraoperative electroCorticoGraphy (ECog): indications, techniques, and utility in epilepsy surgery. Epileptic Disord 16(3): 271–279. https://doi.org/10.1684/epd. 2014.0675 59. Hill NJ, Lal TN, Schroder M et al (2006) Classifying EEG and ECoG signals without subject training for fast BCI implementation: comparison of nonparalyzed and completely paralyzed subjects. IEEE Trans Neural Syst Rehabil Eng 14:183–186

365

60. Leuthardt EC, Schalk G, Wolpaw JR et al (2004) A brain–computer interface using electrocorticographic signals in humans. J Neural Eng 1:63–71. https://doi.org/10. 1088/1741-2560/1/2/001 61. Wilson JA, Felton EA, Garell PC et al (2006) ECoG factors underlying multimodal control of a brain-computer interface. IEEE Trans Neural Syst Rehabil Eng 14:246–250 62. Andersen RA, Musallam S, Pesaran B (2004) Selecting the signals for a brain-machine interface. Curr Opin Neurobiol 14(6):720–726 63. Ludwig KA, Miriani RM, Langhals NB et al (2009) Using a common average reference to improve cortical neuron recordings from microelectrode arrays. J Neurophysiol 101(3):1679–1689. https://doi.org/10. 1152/jn.90989.2008 64. Cogan SF (2008) Neural stimulation and recording electrodes. Annu Rev Biomed Eng 10:275–309. https://doi.org/10.1146/ annurev.bioeng.10.061807.160518 65. Kuzum D, Takano H, Shim E et al (2014) Transparent and flexible low noise graphene electrodes for simultaneous electrophysiology and neuroimaging. Nat Commun 5:17–19. https://doi.org/10.1038/ncomms6259 66. Venkatraman S, Hendricks J, King ZA et al (2011) In vitro and in vivo evaluation of PEDOT microelectrodes for neural stimulation and recording. IEEE Trans Neural Syst Rehabil Eng 19:307–316. https://doi.org/ 10.1109/TNSRE.2011.2109399 67. Khodagholy D, Gelinas JN, Thesen T et al (2015) NeuroGrid: recording action potentials from the surface of the brain. Nat Neurosci 18:310–315. https://doi.org/10. 1038/nn.3905 68. Garcia-Cortadella R, Schwesig G, Jeschke C et al (2021) Graphene active sensor arrays for long-term and wireless mapping of wide frequency band epicortical brain activity. Nat Commun 12:1–17. https://doi.org/10. 1038/s41467-020-20546-w 69. Humphrey DR, Schmidt EM (1990) Extracellular single-unit recording methods. In: Boulton AA, Baker GB, Vanderwolf CH (eds) Neurophysiological techniques: applications to neural systems. Springer 70. Loeb GE, Bak M, Salcman M, Schmidt E (1977) Parylene as a chronically stable, reproducible microelectrode insulator. IEEE Trans Biomed Eng:121–128 71. Schmidt E, McIntosh J, Bak M (1988) Longterm implants of Parylene-C coated microelectrodes. Med Biol Eng Comput 26:96– 101

366

Sevgi O¨ztu¨rk et al.

72. Plexon Inc – Neuroscience Technology. In: Plexon. https://plexon.com/. Accessed 6 Feb 2022 73. Jun JJ, Steinmetz NA, Siegle JH et al (2017) Fully integrated silicon probes for highdensity recording of neural activity. Nature 551:232–236. https://doi.org/10.1038/ nature2463 74. Wise KD (2005) Silicon microsystems for neuroscience and neural prostheses. IEEE Eng Med Biol Mag 24:22–29. https://doi. org/10.1109/MEMB.2005.1511497 75. NeuroNexus. https://www.neuronexus. com/2021-Probe-Catalog.pdf. Accessed 6 Feb 2022 76. Lehew G, Nicolelis MAL (2008) State-of-theArt microwire array design for chronic neural recordings in behaving animals. In: Nicolelis MAL (ed) Methods for neural ensemble recordings, 2nd edn. CRC Press, Boca Raton 77. Williams JC, Rennaker RL, Kipke DR (1999) Long-term neural recording characteristics of wire microelectrode arrays implanted in cerebral cortex. Brain Res Protocol 4:303–313. https://doi.org/10.1016/S1385-299X(99) 00034-3 78. Porada I, Bondar I, Spatz W, Kru¨ger J (2000) Rabbit and monkey visual cortex: more than a year of recording with up to 64 microelectrodes. J Neurosci Methods 95:13–28 79. McNaughton BL, O’Keefe J, Barnes CA (1983) The stereotrode: a new technique for simultaneous isolation of several single units in the central nervous system from multiple unit records. J Neurosci Methods 8:391–397 80. Chelaru MI, Jog MS (2005) Spike source localization with tetrodes. J Neurosci Methods 142:305–315. https://doi.org/10. 1016/j.jneumeth.2004.09.004 81. Gray CM, Maldonado PE, Wilson M, McNaughton B (1995) Tetrodes markedly improve the reliability and yield of multiple single-unit isolation from multi-unit recordings in cat striate cortex. J Neurosci Methods 63:43–54. https://doi.org/10.1016/01650270(95)00085-2 82. Harris KD, Henze DA, Csicsvari J et al (2000) Accuracy of tetrode spike separation as determined by simultaneous intracellular and extracellular measurements. J Neurophysiol 84: 401–414. https://doi.org/10.1152/jn. 2000.84.1.401 83. O’Keefe J, Recce ML (1993) Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus 3:317–330

84. Wilson MA, McNaughton BL (1993) Dynamics of the hippocampal ensemble code for space. Science 261:1055–1058 85. Buzsa´ki G (2004) Large-scale recording of neuronal ensembles. Nat Neurosci 7:446– 451. https://doi.org/10.1038/nn1233 86. Blanche TJ, Spacek MA, Hetke JF, Swindale NV (2005) Polytrodes: high-density silicon electrode arrays for large-scale multiunit recording. J Neurophysiol 93:2987–3000. https://doi.org/10.1152/jn.01023.2004 87. Schjetnan AGP, Luczak A (2011) Recording large-scale neuronal ensembles with silicon probes in the anesthetized rat. JoVE:3282. https://doi.org/10.3791/3282 88. Rousche PJ, Normann RA (1998) Chronic recording capability of the Utah Intracortical Electrode Array in cat sensory cortex. J Neurosci Methods 82(1):1–15. https://doi.org/ 10.1016/S0165-0270(98)00031-4 89. Campbell PK, Jones KE, Huber RJ et al (1991) A silicon-based, three-dimensional neural interface: manufacturing processes for an intracortical electrode array. IEEE Trans Biomed Eng 38:758–768. https://doi.org/ 10.1109/10.83588 90. Hoogerwerf AC, Wise KD (1994) A threedimensional microelectrode array for chronic neural recording. IEEE Trans Biomed Eng 41:1136–1146. https://doi.org/10.1109/ 10.335862 91. Normann RA, Maynard EM, Rousche PJ, Warren DJ (1999) A neural interface for a cortical vision prosthesis. Vis Res 39:2577– 2587. https://doi.org/10.1016/S00426989(99)00040-1 92. Rousche PJ, Normann RA (1992) A method for pneumatically inserting an array of penetrating electrodes into cortical tissue. Ann Biomed Eng 20:413–422 93. Scholten K, Meng E (2015) Materials for microfabricated implantable devices: a review. Lab Chip 15:4256–4272. https://doi.org/ 10.1039/C5LC00809C 94. Drake KL, Wise KD, Farraye J et al (1988) Performance of planar multisite microprobes in recording extracellular single-unit intracortical activity. IEEE Trans Biomed Eng 35: 719–732 95. Rivnay J, Wang H, Fenno L et al (2017) Nextgeneration probes, particles, and proteins for neural interfacing. Sci Adv 3:e1601649. https://doi.org/10.1126/sciadv.1601649 96. Ferna´ndez E, Greger B, House PA et al (2014) Acute human brain responses to

Electrophysiology in Rats intracortical microelectrode arrays: challenges and future prospects. Front Neuroeng 7:24. https://doi.org/10.3389/fneng.2014. 00024 97. McDermott MD, Zhang J, Otto KJ (2015) Improving the brain machine interface via multiple Tetramethyl Orthosilicate sol-gel coatings on microelectrode arrays. In: The 7th international IEEE/EMBS conference on neural engineering 98. Liu J, Fu T-M, Cheng Z et al (2015) Syringeinjectable electronics. Nano Lett 15(10): 6979–6984. https://doi.org/10.1038/ nnano.2015.115 99. Salcman M, Bak MJ (1976) A new chronic recording intracortical microelectrode. Med Biol Eng 14:42–50 100. Schmidt E, Bak M, McIntosh J (1976) Longterm chronic recording from cortical neurons. Exp Neurol 52:496–506 101. Somogyva´ri Z, Cserpa´n D, Ulbert I, E´rdi P (2012) Localization of single-cell current sources based on extracellular potential patterns: the spike CSD method. Eur J Neurosci 36:3299–3313. https://doi.org/10.1111/j. 1460-9568.2012.08249.x 102. Kipke DR, Shain W, Buzsaki G et al (2008) Advanced neurotechnologies for chronic neural interfaces: new horizons and clinical opportunities. J Neurosci 28:11830–11838. https://doi.org/10.1523/JNEUROSCI. 3879-08.2008 103. Delgado Ruz I, Schultz SR (2014) Localising and classifying neurons from high density MEA recordings. J Neurosci Methods 233: 115–128. https://doi.org/10.1016/j. jneumeth.2014.05.037 104. Obien MEJ, Deligkaris K, Bullmann T et al (2015) Revealing neuronal function through microelectrode array recordings. Front Neurosci 8:423. https://doi.org/10.3389/fnins. 2014.00423 105. Fia´th R, Mesze´na D, Somogyva´ri Z et al (2021) Recording site placement on planar silicon-based probes affects signal quality in acute neuronal recordings. Sci Rep 11:1–18 106. Du J, Blanche TJ, Harrison RR et al (2011) Multiplexed, high density electrophysiology with nanofabricated neural probes. PLoS One 6:e26204. https://doi.org/10.1371/ journal.pone.0026204 107. Raducanu BC, Yazicioglu RF, Lopez CM et al (2017) Time multiplexed active neural probe with 1356 parallel recording sites. Sensors (Basel) 17(10):2388. https://doi.org/10. 3390/s17102388

367

108. Steinmetz NA, Aydin C, Lebedeva A et al (2021) Neuropixels 2.0: a miniaturized high-density probe for stable, long-term brain recordings. Science 372(6539): eabf4588. https://doi.org/10.1126/sci ence.abf4588 109. Kipke DR, Vetter RJ, Williams JC, Hetke JF (2003) Silicon-substrate intracortical microelectrode arrays for long-term recording of neuronal spike activity in cerebral cortex. IEEE Trans Neural Syst Rehabil Eng 11: 151–155. https://doi.org/10.1109/ TNSRE.2003.814443 110. Ulyanova AV, Cottone C, Adam CD et al (2019) Multichannel silicon probes for awake hippocampal recordings in large animals. Front Neurosci 13:397. https://doi. org/10.3389/fnins.2019.00397 111. Priori A, Foffani G, Rossi L, Marceglia S (2013) Adaptive deep brain stimulation (aDBS) controlled by local field potential oscillations. Exp Neurol 245:77–86. https:// doi.org/10.1016/j.expneurol.2012.09.013 112. Kennedy SH, Giacobbe P, Rizvi SJ, Placenza FM, Nishikawa Y, Mayberg HS et al (2011) Deep brain stimulation for treatmentresistant depression: follow-up after 3 to 6 years. Am J Psychiatr 168:502–510. https://doi.org/10.1176/appi.ajp.2010. 10081187 113. Bianco MG, Pullano SA, Citraro R et al (2020) Neural modulation of the primary auditory cortex by intracortical microstimulation with a bio-inspired electronic system. Bioengineering 7(1):23 114. Berg JA, Dammann JF, Tenore FV et al (2013) Behavioral demonstration of a somatosensory neuroprosthesis. IEEE Trans Neural Syst Rehabil Eng 21:500–507. https://doi.org/10.1109/TNSRE.2013. 2244616 115. Fagg AH, Hatsopoulos NG, London BM et al (2009) Toward a biomimetic, bidirectional, brain machine interface. In: Conference proceedings: Annual international conference of the IEEE Engineering in Medicine and Biology Society 2009, pp 3376–3380 116. Rousche PJ, Otto KJ, Reilly MP, Kipke DR (2003) Single electrode micro-stimulation of rat auditory cortex: an evaluation of behavioral performance. Hear Res 179:62–71. https://doi.org/10.1016/s0378-5955(03) 00081-9 117. Merrill DR, Bikson M, Jefferys JGR (2005) Electrical stimulation of excitable tissue: design of efficacious and safe protocols. J

368

Sevgi O¨ztu¨rk et al.

Neurosci Methods 141:171–198. https:// doi.org/10.1016/j.jneumeth.2004.10.020 118. Kostarelos K, Vincent M, Hebert C, Garrido JA (2017) Graphene in the design and engineering of next-generation neural interfaces. Adv Mater 29:1700909. https://doi.org/10. 1002/adma.201700909 119. Spieth S, Schumacher A, Trenkle F et al (2014) Approaches for drug delivery with intracortical probes. Biomed Tech 59(4): 291–303. https://doi.org/10.1515/bmt2012-0096 120. Noudoost B, Moore T (2011) A reliable microinjectrode system for use in behaving monkeys. J Neurosci Methods 194:218– 223. https://doi.org/10.1016/j.jneumeth. 2010.10.009 121. Spieth S, Schumacher A, Kallenbach C et al (2012) The NeuroMedicator—a micropump integrated with silicon microprobes for drug delivery in neural research. J Micromech Microeng 22:065020. https://doi.org/10. 1088/0960-1317/22/6/065020 122. Gerhardt GA, Palmer MR (1987) Characterization of the techniques of pressure ejection and microiontophoresis using in vivo electrochemistry. J Neurosci Methods 22:147–159. https://doi.org/10.1016/0165-0270(87) 90009-4 123. Kang YN, Chou N, Jang J-W et al (2021) A 3D flexible neural interface based on a microfluidic interconnection cable capable of chemical delivery. Microsyst Nanoeng 7:66. https://doi.org/10.1038/s41378-02100295-6 124. Veith VK, Quigley C, Treue S (2016) A pressure injection system for investigating the neuropharmacology of information processing in awake behaving macaque monkey cortex. J Vis Exp: JoVE. https://doi.org/10. 3791/53724 125. Herr NR, Wightman RM (2013) Improved techniques for examining rapid dopamine signaling with iontophoresis. Front Biosci 5: 249–257. https://doi.org/10.2741/e612 126. Herr NR, Kile BM, Carelli RM, Wightman RM (2008) Electroosmotic flow and its contribution to iontophoretic delivery. Anal Chem 80:8635–8641. https://doi.org/10. 1021/ac801547a 127. Bevan P, Bradshaw CM, Pun RY et al (1979) The relative contribution of iontophoresis and electro-osmosis to the electrophoretic release of noradrenaline from multibarrelled micropipettes [proceedings]. Br J Pharmacol 67:478P–479P

128. Walker T, Dillman N, Weiss ML (1995) A constant current source for extracellular microiontophoresis. J Neurosci Methods 63: 127–136. https://doi.org/10.1016/01650270(95)00101-8 129. Lalley PM (1999) Microiontophoresis and pressure ejection. In: Windhorst U, Johansson H (eds) Modern techniques in neuroscience research. Springer, Berlin Heidelberg, pp 193–212 130. Thiele A, Delicato LS, Roberts MJ, Gieselmann MA (2006) A novel electrode-pipette design for simultaneous recording of extracellular spikes and iontophoretic drug application in awake behaving monkeys. J Neurosci Methods 158:207–211. https://doi.org/10. 1016/j.jneumeth.2006.05.032 131. Budai D (2010) Carbon fiber-based microelectrodes and microbiosensors. In: Somerset VS (ed) Intelligent and biosensors, pp 269–288 132. Chen J, Wise KD, Hetke JF, Bledsoe SC (1997) A multichannel neural probe for selective chemical delivery at the cellular level. IEEE Trans Biomed Eng 44:760–769. https://doi.org/10.1109/10.605435 133. John J, Li Y, Zhang J et al (2011) Microfabrication of 3D neural probes with combined electrical and chemical interfaces. J Micromech Microeng 21:105011. https://doi. org/10.1088/0960-1317/21/10/105011 134. Kobayashi R, Kanno S, Lee S et al (2009) Development of double-sided Si neural probe with microfluidic channels using wafer direct bonding technique. In: 2009 4th international IEEE/EMBS conference on neural engineering. IEEE, pp 96–99 135. Seidl K, Spieth S, Herwik S et al (2010) In-plane silicon probes for simultaneous neural recording and drug delivery. J Micromech Microeng 20:105006. https://doi.org/10. 1088/0960-1317/20/10/105006 136. Cheung KC, Djupsund K, Dan Y, Lee LP (2003) Implantable multichannel electrode array based on SOI technology. J Microelectromech Syst 12:179–184. https://doi.org/ 10.1109/JMEMS.2003.809962 137. Guo K, Pei W, Li X et al (2012) Fabrication and characterization of implantable silicon neural probe with microfluidic channels. SCIENCE CHINA Technol Sci 55:1–5. https:// doi.org/10.1007/s11431-011-4569-8 138. Pellinen DS, Moon T, Vetter RJ et al (2005) Multifunctional flexible Parylene-based intracortical microelectrodes. In: 2005 IEEE

Electrophysiology in Rats engineering in medicine and biology 27th annual conference. IEEE, pp 5272–5275 139. Takeuchi S, Ziegler D, Yoshida Y et al (2005) Parylene flexible neural probes integrated with microfluidic channels. Lab Chip 5:519– 523. https://doi.org/10.1039/b417497f 140. Ziegler D, Suzuki T, Takeuchi S (2006) Fabrication of flexible neural probes with built-in microfluidic channels by thermal bonding of Parylene. J Microelectromech Syst 15:1477– 1482. https://doi.org/10.1109/JMEMS. 2006.879681 141. Rubehn B, Wolff SBE, Tovote P et al (2013) A polymer-based neural microimplant for optogenetic applications: design and first in vivo study. Lab Chip 13:579. https://doi. org/10.1039/c2lc40874k 142. Metz S, Bertsch A, Bertrand D, Renaud P (2004) Flexible polyimide probes with microelectrodes and embedded microfluidic channels for simultaneous drug delivery and multichannel monitoring of bioelectric activity. Biosens Bioelectron 19:1309–1318. https:// doi.org/10.1016/j.bios.2003.11.021 143. Altuna A, Bellistri E, Cid E et al (2013) SU-8 based microprobes for simultaneous neural depth recording and drug delivery in the brain. Lab Chip 13:1422. https://doi.org/ 10.1039/c3lc41364k 144. Ferna´ndez LJ, Altuna A, Tijero M et al (2009) Study of functional viability of SU-8-based microneedles for neural applications. J Micromech Microeng 19:025007. https://doi. org/10.1088/0960-1317/19/2/025007 145. Hochberg LR, Serruya MD, Friehs GM et al (2006) Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 442:164–171. https://doi.org/10. 1038/nature04970 146. Rohatgi P, Langhals NB, Kipke DR, Patil PG (2009) In vivo performance of a microelectrode neural probe with integrated drug delivery. Neurosurg Focus 27:E8. https:// doi.org/10.3171/2009.4.FOCUS0983 147. Leem JW, Willis WD, Weller SC, Chung JM (1993) Differential activation and classification of cutaneous afferents in the rat. J Neurophysiol 70:2411–2424. https://doi.org/ 10.1152/jn.1993.70.6.2411 148. Vardar B, Gu¨c¸lu¨ B (2017) Non-NMDA receptor-mediated vibrotactile responses of neurons from the hindpaw representation in the rat SI cortex. Somatosens Mot Res 34: 189–203. https://doi.org/10.1080/ 08990220.2017.1390450 149. Shoykhet M, Doherty D, Simons DJ (2000) Coding of deflection velocity and amplitude

369

by whisker primary afferent neurons: implications for higher level processing. Somatosens Mot Res 17:171–180. https://doi.org/10. 1080/08990220050020580 150. Pawson L, Prestia LT, Mahoney GK et al (2009) GABAergic/glutamatergic-glial/ neuronal interaction contributes to rapid adaptation in Pacinian corpuscles. J Neurosci 29:2695–2705. https://doi.org/10.1523/ JNEUROSCI.5974-08.2009 151. Khalsa PS, Zhang C, Qin YX (2000) Encoding of location and intensity of noxious indentation into rat skin by spatial populations of cutaneous mechano-nociceptors. J Neurophysiol 83:3049–3061 152. Gu¨c¸lu¨ B (2005) Maximizing the entropy of histogram bar heights to explore neural activity: a simulation study on auditory and tactile fibers. Acta Neurobiol Exp 65(4):399–407 153. Chapin JK, Lin C-S (1984) Mapping the body representation in the SI cortex of anesthetized and awake rats. J Comp Neurol 229:199–213. https://doi.org/10.1002/ cne.902290206 154. Oliveira LMO, Dimitrov DF (2008) Surgical techniques for chronic implantation of microwire arrays in rodents and primates. In: Nicolelis MAL (ed) Methods for neural ensemble recordings, 2nd edn. CRC Press, Boca Raton 155. Nicolelis MAL, Dimitrov D, Carmena JM et al (2003) Chronic, multisite, multielectrode recordings in macaque monkeys. Proc Natl Acad Sci 100:11041–11046. https:// doi.org/10.1073/pnas.1934665100 156. Venkatachalam S, Fee MS, Kleinfeld D (1999) Ultra-miniature headstage with 6-channel drive and vacuum-assisted micro-wire implantation for chronic recording from the neocortex. J Neurosci Methods 90:37–46 157. Bai Q, Wise KD, Anderson DJ (2000) A highyield microassembly structure for threedimensional microelectrode arrays. IEEE Trans Biomed Eng 47:281–289 158. Hatsopoulos NG, Ojakangas CL, Paninski L, Donoghue JP (1998) Information about movement direction obtained from synchronous activity of motor cortical neurons. Proc Natl Acad Sci 95:15706–15711 159. Maynard E, Hatsopoulos N, Ojakangas C et al (1999) Neuronal interactions improve cortical population coding of movement direction. J Neurosci 19:8083–8093 160. Usoro JO, Dogra K, Abbott JR et al (2021) Influence of implantation depth on the performance of intracortical probe recording sites. Micromachines 12:1158. https://doi. org/10.3390/mi12101158

370

Sevgi O¨ztu¨rk et al.

161. Decharms RC, Blake DT, Merzenich MM (1999) A multielectrode implant device for the cerebral cortex. J Neurosci Methods 93: 27–35 162. Hetke JF, Lund JL, Najafi K et al (1994) Silicon ribbon cables for chronically implantable microelectrode arrays. IEEE Trans Biomed Eng 41:314–321 163. Owen JH, Laschinger J, Bridwell K et al (1988) Sensitivity and specificity of somatosensory and neurogenic-motor evoked potentials in animals and humans. Spine 13(10): 1111–1118 164. Sharma HS, Winkler T, Sta˚lberg E et al (1991) Evaluation of traumatic spinal cord edema using evoked potentials recorded from the spinal epidural space: an experimental study in the rat. J Neurol Sci 102:150–162 165. Fehlings MG, Tator CH, Linden RD, Piper IR (1988) Motor and somatosensory evoked potentials recorded from the rat. Electroencephalogr Clin Neurophysiol 69:65–78. https://doi.org/10.1016/0013-4694(88) 90036-3 166. Fehlings MG, Tator CH, Linden RD (1989) The relationships among the severity of spinal cord injury, motor and somatosensory evoked potentials and spinal cord blood flow. Electroencephalogr Clin Neurophysiol/Evoked Potentials Section 74:241–259. https://doi. org/10.1016/0168-5597(89)90055-5 167. Fehlings MG, Tator CH (1992) The effect of direct current field polarity on recovery after acute experimental spinal cord injury. Brain Res 579:32–42. https://doi.org/10.1016/ 0006-8993(92)90738-U 168. Boyes WK, Cooper GP (1981) Acrylamide neurotoxicity: effects on far-field somatosensory evoked potentials in rats. Neurobehav Toxicol Teratol 3:487–490 169. Edwards MSB, Powers SK, Baringer RA et al (1983) Evoked potentials in rats with misonidazole neurotoxicity. J Neuro-Oncol 1:115– 123 170. Oro J, Haghighi SS (1992) Effects of altering core body temperature on somatosensory and motor evoked potentials in rats. Spine 17: 498–503. https://doi.org/10.1097/ 00007632-199205000-00005 171. Mouraux A, Iannetti GD (2008) Across-trial averaging of event-related EEG responses and beyond. Magn Reson Imaging 26:1041– 1054. https://doi.org/10.1016/j.mri.2008. 01.011 172. Malekpour S, Gubner JA, Sethares WA (2018) Measures of generalized magnitudesquared coherence: differences and

similarities. J Franklin Inst 355(5): 2932–2950. https://doi.org/10.1016/j. jfranklin.2018.01.014 173. Lachaux JP, Rodriguez E, Martinerie J, Varela FJ (1999) Measuring phase synchrony in brain signals. Hum Brain Mapp 8(4): 194–208 174. Lowet E, Roberts MJ, Bonizzi P et al (2016) Quantifying neural oscillatory synchronization: a comparison between spectral coherence and phase-locking value approaches. PLoS One 11(1):e0146443. https://doi. org/10.1371/journal.pone.0146443 175. Golabchi A, Kipke I Chronic penetrating arrays. In: NeuroNexus https://www.neuro n e x u s . c o m / fi l e s / e l e c t r o d e a r r a y s / penetrating/ChronicPenetratingArrays.pdf 176. Quiroga R (2007) Spike sorting. Scholarpedia 2:3583. https://doi.org/10.4249/ scholarpedia.3583 177. Kim S, McNames J (2007) Automatic spike detection based on adaptive template matching for extracellular neural recordings. J Neurosci Methods 165:165–174. https://doi. org/10.1016/j.jneumeth.2007.05.033 178. Mallat SG (1989) A theory for multiresolution signal decomposition: the wavelet representation. IEEE Trans Pattern Anal Machine Intell 11:674–693. https://doi.org/10. 1109/34.192463 179. Quiroga RQ, Nadasdy Z, Ben-Shaul Y (2004) Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering. Neural Comput 16:1661–1687. https:// doi.org/10.1162/089976604774201631 180. Chung JE, Magland JF, Barnett AH et al (2017) A fully automated approach to spike sorting. Neuron 95:1381–1394.e6. https:// doi.org/10.1016/j.neuron.2017.08.030 181. Rossant C, Kadir SN, Goodman DFM et al (2016) Spike sorting for large, dense electrode arrays. Nat Neurosci 19:634–641. https://doi.org/10.1038/nn.4268 182. Pachitariu M, Steinmetz N, Kadir S et al (2016) Kilosort: realtime spike-sorting for extracellular electrophysiology with hundreds of channels. bioRxiv. https://doi.org/10. 1101/061481 183. Chaure FJ, Rey HG, Quian Quiroga R (2018) A novel and fully automatic spikesorting implementation with variable number of features. J Neurophysiol 120:1859–1871. https://doi.org/10.1152/jn.00339.2018 184. Pelleg D, Moore AW, others (2000) X-means: Extending k-means with efficient estimation of the number of clusters. In: Proceedings of

Electrophysiology in Rats the 17th international conference on machine learning. pp 727–734 185. Wheeler BC, Nicolelis M (1999) Automatic discrimination of single units. In: Nicolelis M (ed) Methods for neural ensemble recordings. CRC Press, Boca Raton, pp 61–77 186. Steinmetz NA, Zatka-Haas P, Carandini M, Harris KD (2019) Distributed coding of choice, action and engagement across the mouse brain. Nature 576:266–273. https:// doi.org/10.1038/s41586-019-1787-x 187. Koivuniemi AS, Otto KJ (2012) The depth, waveform and pulse rate for electrical microstimulation of the auditory cortex. In: Annual international conference of the IEEE Engineering in Medicine and Biology Society. IEEE, San Diego, CA, pp 2489–2492 188. Boinagrov D, Loudin J, Palanker D (2010) Strength–duration relationship for extracellular neural stimulation: numerical and analytical models. J Neurophysiol 104:2236–2248. https://doi.org/10.1152/jn.00343.2010 189. Butovas S, Hormuzdi SG, Monyer H, Schwarz C (2006) Effects of electrically coupled inhibitory networks on local neuronal responses to intracortical microstimulation. J Neurophysiol 96:1227–1236. https://doi. org/10.1152/jn.01170.2005 190. Azin M, Guggenmos DJ, Barbay S et al (2011) A miniaturized system for spiketriggered intracortical microstimulation in an ambulatory rat. IEEE Trans Biomed Eng 58: 2589–2597. https://doi.org/10.1109/ TBME.2011.2159603 191. Venkatraman S, Elkabany K, Long JD et al (2009) A system for neural recording and closed-loop intracortical microstimulation in awake rodents. IEEE Trans Biomed Eng 56: 15–22. https://doi.org/10.1109/TBME. 2008.2005944 192. Dinse HR, Ragert P, Pleger B et al (2003) Pharmacological modulation of perceptual learning and associated cortical reorganization. Science 301:91–94. https://doi.org/ 10.1126/science.1085423 193. Recanzone GH, Merzenich MM, Dinse HR (1992) Expansion of the cortical representation of a specific skin field in primary somatosensory cortex by intracortical microstimulation. Cereb Cortex 2:181–196. https://doi.org/10.1093/cercor/2.3.181 194. Spengler F, Godde B, Dinse HR (1995) Effects of ageing on topographic organization of somatosensory cortex. Neuroreport 6: 469–473 195. Butovas S, Schwarz C (2007) Detection psychophysics of intracortical microstimulation

371

in rat primary somatosensory cortex. Eur J Neurosci 25:2161–2169. https://doi.org/ 10.1111/j.1460-9568.2007.05449.x 196. Semprini M, Bennicelli L, Vato A (2012) A parametric study of intracortical microstimulation in behaving rats for the development of artificial sensory channels. Conference proceedings: Annual international conference of the IEEE Engineering in Medicine and Biology Society 2012:799–802 197. Gottschaldt K, Vahle-Hinz C, Hicks TP (1983) Electrophysiological and micropharmacological studies on mechanisms of inputoutput transformation in single neurones of the somatosensory thalamus. In: Macchi G (ed) Somatosensory integration in the thalamus. Elsevier Science, Amsterdam, pp 199–216 198. Philip Hicks T (1984) The history and development of microiontophoresis in experimental neurobiology. Prog Neurobiol 22:185– 240. https://doi.org/10.1016/0301-0082 (84)90019-4 199. Curtis DR, Eccles RM (1958) The effect of diffusional barriers upon the pharmacology of cells within the central nervous system. J Physiol 141:446–463. https://doi.org/10. 1113/jphysiol.1958.sp005988 200. Curtis DR, Eccles RM (1958) The excitation of Renshaw cells by pharmacological agents applied electrophoretically. J Physiol 141: 435–445. https://doi.org/10.1113/ jphysiol.1958.sp005987 201. Bucher ES, Fox ME, Kim L et al (2014) Medullary norepinephrine neurons modulate local oxygen concentrations in the bed nucleus of the stria terminalis. J Cereb Blood Flow Metab 34:1128–1137. https://doi. org/10.1038/jcbfm.2014.60 202. Windhorst U, Johansson H (1999) Modern Techniques in Neuroscience Research. Springer Berlin Heidelberg, Berlin, Heidelberg 203. Kirkpatrick DC, Walton LR, Edwards MA, Wightman RM (2016) Quantitative analysis of iontophoretic drug delivery from micropipettes. Analyst 141:1930–1938. https:// doi.org/10.1039/c5an02530c 204. Kova´cs P, De´nes V, Kelle´nyi L, Herna´di I (2005) Microiontophoresis electrode location by neurohistological marking: comparison of four native dyes applied from current balancing electrode channels. J Pharmacol Toxicol Methods 51:147–151. https://doi. org/10.1016/j.vascn.2004.08.002 205. Herr NR, Daniel KB, Belle AM et al (2010) Probing presynaptic regulation of

372

Sevgi O¨ztu¨rk et al.

extracellular dopamine with iontophoresis. ACS Chem Neurosci 1:627–638. https:// doi.org/10.1021/cn100056r 206. Purves RD (1979) The physics of iontophoretic pipettes. J Neurosci Methods 1:165– 178. https://doi.org/10.1016/0165-0270 (79)90014-1 207. Kasting GB (1992) Theoretical models for iontophoretic delivery. Adv Drug Deliv Rev 9:177–199. https://doi.org/10.1016/ 0169-409X(92)90023-J 208. Dias EC, Segraves MA (1997) A pressure system for the microinjection of substances into the brain of awake monkeys. J Neurosci Methods 72:43–47. https://doi.org/10. 1016/s0165-0270(96)00154-9

209. Woody CD, Bartfai T, Gruen E, Nairn AC (1986) Intracellular injection of cGMPdependent protein kinase results in increased input resistance in neurons of the mammalian motor cortex. Brain Res 386:379–385. https://doi.org/10.1016/0006-8993(86) 90175-7 210. Szente MB, Baranyi A, Woody CD (1990) Effects of protein kinase C inhibitor H-7 on membrane properties and synaptic responses of neocortical neurons of awake cats. Brain Res 506:281–286. https://doi.org/10. 1016/0006-8993(90)91262-f 211. Shannon RV (1992) A model of safe levels for electrical stimulation. IEEE Trans Biomed Eng 39:424–426. https://doi.org/10. 1109/10.126616

Chapter 17 Imaging Somatosensory Cortex in Rodents Mariangela Panniello, Severin A. C. Limal, and Michael M. Kohl Abstract The rodent somatosensory cortex has been investigated using a range of electrophysiological techniques, from intracellular recordings to electroencephalography. Nonetheless, their accessible location on the dorsal surface of the brain has more recently made the somatosensory areas popular models for the imaging-based investigation of cortical function. In this chapter, we will outline the general principles of two-photon microscopy applied to the functional study of the rodent somatosensory cortex. This technique allows recording the activity of hundreds of individual neurons simultaneously, with single-cell precision and while knowing their relative positions in the brain. We will place particular emphasis on long-term calcium imaging procedures on awake behaving mice and will introduce advantages and limitations of this technique. Our specific aim is to provide the reader with useful information regarding equipment and experimental procedures, from the choice of the calcium indicator to the post hoc analysis of imaging time series. Key words Mouse, Barrel cortex, Somatosensory, Whiskers, Two photon, Chronic imaging, Fluorescence microscopy, Calcium imaging

1

Introduction A fundamental question in neuroscience concerns the link between neuronal activity and our perception of, and interaction with, the external world. Tactile stimuli are mostly received by mechanical receptors on the surface of our body and readily converted into electrical impulses which travel through the somatosensory pathways until they reach the cerebral cortex—one of the most complex of brain structures, where disparate sensory inputs are thought to be converted into coherent percepts. Here, hundreds of thousands of individual neurons act in concert to encode sensory information in specific spatio-temporal patterns of electrical activity. Decoding these patterns is crucial to advance our understanding of brain function as well as to recognize when and how the underlying mechanisms are altered due to pathological conditions. In order to succeed in this endeavor, we must record the activity of large

373

374

Mariangela Panniello et al.

populations of neurons with single-cell precision and while the experimental subject is undergoing a controlled sensory experience. This is now possible thanks to the recent development of a range of optical and genetic approaches to brain function (see also Chapters 16, 18, and 19, this volume). The imaging technique we will describe in this chapter, namely in vivo two-photon (2p) calcium imaging, allows monitoring the activity of hundreds of genetically identified individual neurons simultaneously in the cortex of animals performing a sensory task over a time period ranging from seconds to weeks. Relying on fluorescent molecules as probes, 2p calcium imaging uses the increase in the intracellular concentration of free calcium as a correlate for spiking activity [1]. By measuring fluorescence changes concurrently with sensory stimulation in awake animals, neural activity can therefore be linked to perception. Owing to its small size and the possibility to combine cell-specific recordings with complex sensory-learning paradigms, the mouse is the ideal model for studying cortical function at the fine spatial scale with 2p imaging. 1.1 The Rodent Barrel Cortex as an Experimental Model for In Vivo Imaging

The somatosensory system of the mouse has been well characterized in its structural and functional properties. Great attention has traditionally been given to the vibrissal primary somatosensory cortex (vS1). This area is the first in the cortex to receive bottomup inputs originating in the tactile receptors on the mystacial vibrissae (or whiskers)—the sensory organs actively moved by rodents for object detection, discrimination, and localization [2]. Mirroring the spatial arrangement of the facial whiskers, vS1 is characterized by a somatotopic organization [3, 4], where sensory information from each individual whisker is represented in a discrete anatomical unit, termed a “barrel,”, hence the alternative name for vS1, “barrel cortex.” Each cortical barrel defines an area of approximately 150 μm in diameter and contains about 10,000 neurons [5]. Its topographic organization allows discrete tactile stimuli to result in easily localizable activity during electrophysiological or optical recordings. Besides the topographic organization, its accessible location on the dorsal surface of the brain has made vS1 a popular model for the 2p imaging-based investigation of cortical function. Imaging has been paired with a wide range of tactile stimulation paradigms, from textured sandpapers [6, 7] to objects presented at different positions [8, 9], tactile virtual realities [10], and vibrating piezoelectric drivers deflecting threaded whiskers at different frequencies [11]. Barrel cortex was first imaged with 2p microscopy nearly 20 years ago in a pioneering study introducing this technique to the in vivo investigation of the mammalian cortex. Here, individual neurons in superficial layers two and three (LII-LIII) responded to whisker deflection in anesthetized mice [12]. A few years later, Sato

Imaging Rodents

375

and colleagues [13] disclosed the potential of 2p calcium imaging in capturing fine-scale functional details of barrel cortex. In their report, the authors confirmed the overall somatotopic organization of vS1, but they nonetheless revealed a high variability of whisker preference among neighboring neurons within one barrel. This fractured topography, inaccessible to traditional microelectrode recordings which average neuronal activity across hundreds of micrometers (see Chapter 16, this volume), was later known as a “salt and pepper organization,” and emerged in 2p studies of the primary cortical areas of a range of species [14–18]. Early in vivo 2p imaging studies in anesthetized models had the advantage of being less technically challenging than research in awake preparations, and they have revealed fundamental aspects of S1 function [19, 20]. Nonetheless, to understand how specialized circuits represent tactile information that is ultimately used to inform behavioral choices and actions, 2p calcium imaging, combined with genetic tools, was later performed in longitudinal studies on awake animals actively sensing tactile stimuli, and in some cases engaged in learning tasks. In these studies, individual S1 neurons were modulated by whisker movement, object touch [21, 22], object location [8], multisensory stimuli [23], and locomotion speed of the animal [7, 10]. Moreover, as primary cortical areas are traditionally considered as purely sensory stations, it is perhaps surprising that recent reports have described correlates of behavioral choice [24] and stimulus-outcome associations [25] in vS1 neurons, particularly in expert mice carrying out whisker-based sensory discrimination tasks. The evidence reported so far has emerged from imaging calcium dynamics at the level of neuronal somata, which will be the focus of this chapter. However, it is worth noting that the combination of 2p imaging with genetic strategies to label specific cortico-cortical projections has provided unprecedented insight into the feedforward and feedback dynamics underlying information transmission across cortical regions [22, 26]. 1.2 The Cortex at High Resolution: Principles of In Vivo Two-Photon Imaging

Up until the end of the twentieth century, microelectrode recordings were the gold standard in neurophysiology (Chapters 15 and 16, this volume). Although electrophysiological methods guarantee precise temporal control of neuronal activity, they suffer from a spatial resolution limitation. The signals collected by extracellular recordings often correspond to neuronal activity averaged across dozens, if not hundreds, of cells within a cortical patch. Moreover, these signals are often biased toward the most robustly responding neurons [27]. Even in single neuron recordings, the spacing between cells is at least 50 μm, and typically ~100 μm or greater [28]. Consequently, variations in activity between neighboring cells are usually not examined.

376

Mariangela Panniello et al.

These issues were overcome when, in the early 1990s, 2p excitation was combined with laser scanning microscopy (LSM) [29]. This optical technique takes advantage of ultra-fast pulsed laser light to excite fluorescent molecules expressed in living tissue and to monitor the activity of local populations of neurons (and other cells) at subcellular resolution. Differently from traditional confocal microscopy, where one photon of light excites one fluorophore molecule, 2p excitation requires 2 photons to hit the fluorophore nearly simultaneously to excite it to a higher electronic state. In 2p microscopy, this temporal and spatial coincidence is obtained using lasers that, rather than emitting light continuously, generate high-density packets of light (pulses) at high frequency. As each photon’s energy is inversely proportional to the length of the electromagnetic wave, lasers used in 2p microscopy generate light of longer wavelength compared to those used in 1p confocal microscopy, typically in the near-infrared range. Such lower-energy light is conveniently less damaging for the biological tissue compared to the light used for traditional confocal imaging. Moreover, according to the Rayleigh scattering principle and to the optical properties of brain tissue, light of long wavelengths can penetrate deeper into a tissue than light of shorter wavelengths. These features allow imaging deeper in the brain (i.e., several hundred micrometers) compared to 1p confocal microscopy, which is limited to a few tens of micrometers below the cortical surface. Finally, as 2p excitation intensity drops with the square of the distance from the focal plane, light emission is spatially limited, and out-of-focus signals are substantially reduced compared to 1p microscopy [30]. For an in-depth review of the principles of two-photon excitation microscopy refer to [31, 32]. In order to scan the biological sample (e.g., a region of somatosensory cortex in a living animal), basic multiphoton microscopes direct the laser beam onto two moving galvanometric mirrors, which deflect the beam in two dimensions according to the voltage applied to the system. The slow movement of these devices allows us to scan samples at limited rates (a few Hz, depending on the size of the scanned region) and substantially limits the temporal resolution of 2p imaging. Currently, galvanometric scanners are often used at their resonant frequency, with acquisition rates of several dozens of Hz [33]. Computational methods enable selective scanning of the most informative pixels in a field of view, optimizing signal-to-noise ratio and imaging speed without need for hardware changes [34]. By combining a resonant scanner with electrically tunable lenses, it is possible to quasi-simultaneously scan two separate planes, and increase the number of recorded neurons [35]. Devices known as acousto-optic deflectors can spatially control the laser beam and are deployed for fast random-access 3D imaging [36]. Combining them with aspheric lenses for remote focusing provides a further approach for fast, multi-plane scanning

Imaging Rodents

377

[37]. While the attempt to increase the number of quasisimultaneously imaged focal planes has characterized the development of 2p technology from its early days, the lateral extension of the 2p scanned fields of view (FOVs) remained limited to individual cortical regions (maximum 1 × 1 mm) until the recent introduction of custom systems for imaging areas up to 5 × 5 mm, including up to 2000 neurons [38]. 1.3 Fluorescent Molecules for Probing Intracellular Calcium Concentrations

Typically excited with wavelengths between 900 and 1100 nm, the most common fluorophores associated with calcium indicators emit green (e.g., eGFP) or red (e.g., mRuby, mCherry) fluorescence. The first generation of calcium probes resulted from a hybridization between high-affinity calcium chelators, like EDTA and BAPTA, and a range of fluorophores (Fura2, OGB1-bapta, Fluo4; [39, 40]). These indicators lose their functionality hours after being injected into the brain tissue, therefore allowing only acute investigation of cortical function. Recent advances in protein engineering have extended the temporal window for stable neuronal expression of calcium probes, facilitating the advancement of longitudinal imaging studies. The crucial breakthrough in this respect was the design of genetically encoded calcium indicators (GECIs; [41]). Their extraordinary development in the past 10 years has allowed neurophysiologists to test probes offering multiple dynamic ranges, temporal kinetics, and calcium sensitivities. The GCaMP family includes a group of GFP-based GECIs able to detect single action potentials in specific conditions. The three versions of GCaMP6 (fast, medium, and slow) are currently the most widely used calcium indicators, but the recent introduction of GCaMP7 and GCaMP8 has improved temporal resolution and facilitated action potential detection and imaging in small structures like neurites [42, 43]. Red indicators, on the other hand, are preferred for imaging infragranular cortical layers thanks to their superior tissue-penetration ability compared to fluorophores excited at shorter wavelengths (jRCaMP1a, jRGECO—[44, 45]). They have nonetheless slower temporal kinetics and lower calcium sensitivities. Deeper brain structures, such as the rodent somatosensory thalamus or the infragranular cortical layers, can also be reached by implanting prisms [46, 47], GRIN lenses [48], or imagingwindows in direct contact with the deep region of interest [49]. These strategies require removing a volume of brain tissue above the investigated area. By contrast, three-photon (3p) excitation is emerging as a non-invasive alternative [50]. In this case, the 3 photons used to excite the fluorophore have longer wavelength and lower energy than those used in 2p microscopy. This feature makes 3p microscopy highly suited for deep tissue imaging. That said, the high number and high density of photons used for 3p excitation make photodamaging and photobleaching

378

Mariangela Panniello et al.

(see below) more likely to occur. Temporal resolution is the main limitation when measuring neural electrical activity through a cellular second messenger, calcium, probed by slow protein-based sensors (e.g., GCaMP6f half-decay time is 142 ms for 1 action potential in vivo). Genetically encoded voltage-sensitive dyes (GEVIs) can resolve spikes with sub-millisecond precision. However, their use in vivo has been limited because, due to their expression being restricted to the cellular membrane, high illumination power is necessary during imaging, often leading to inactivating the indicator molecule (an effect known as “photobleaching”) and phototoxicity. Because voltage events are faster than calcium events, GEVI imaging requires high 2p acquisition rates. Recent research has improved GEVI and imaging technologies remarkably, giving neuroscientists hope that they will soon be able to observe electrical events in neuronal populations at single-cell resolution [51]. Neuronal expression of both GECIs and GEVIs is usually achieved by stereotaxically injecting, in the investigated region, tailored viral vectors for constitutive, or Cre-LoxP regulated, expression of the indicator gene. While the most widely used are adeno-associated virus (AAV) and, to a lesser extent, lentiviral vectors, their serotypes vary in order to control cellular tropism [52] (tested specifically in somatosensory cortex, by [53]). Similarly, transgene promoters are selected to obtain expression in specific neuronal types (e.g., Synapsin 1, CaMKII, Thy1, mDlx), cortical layers (e.g., Scnn-1a, Tlx3, Cux2), or brain regions (e.g., Emx1). Despite their widespread use, stereotaxic viral injections are associated with a range of constraints. First, they are timeconsuming for the experimenter and characterized by a relatively low yield, depending on the surgical protocol. Second, the limited volume of labeled neurons prevents the investigation of large cortical areas. Finally, expression levels of GECIs increase over time and can eventually lead to aberrant activity in cells that are expressing excessive copies of the exogen protein [54]. Stable, uniform, and widespread expression of GECIs in mice is instead guaranteed in transgenic lines engineered for reporting GCaMP variants in specific cell types throughout life or after chemical induction [55]. Thanks to the rapid improvement of the recombinant technology and good commercial availability, these reporter lines are replacing the use of viral injection for GECI expression and are becoming the main choice for longitudinal imaging in awake behaving subjects. In the following sections, we will describe the experimental procedures required for performing longitudinal 2p calcium imaging of neural activity in the somatosensory cortex of awake mice over the course of days to weeks. We will focus on three main steps (Fig. 1): surgical procedures for gaining optical access to the brain

Imaging Rodents

379

Fig. 1 Example workflow of a chronic two-photon imaging experiment. Surgical procedures are indicated in yellow; the green box reports the phase of habituation to head fixation and handling by the experimenter; imaging sessions are indicated in dark blue; light blue boxes represent the temporal succession of the imaging analysis steps

and expressing the calcium probe, 2p imaging sessions, and pre-processing of the imaging data. The longitudinal procedure (referred to as “chronic imaging” from now on) differs from “acute” procedures, where imaging is typically performed only once per animal, typically under anesthesia, and immediately following implantation of an imaging window.

2

Materials

2.1 Surgical Procedures: Equipment

• Wild-type (C57Bl/6) or transgenic mice (Charles River, Jackson Laboratories). • Mouse digital stereotaxic frame inclusive of non-rupture earbars and gas anesthesia mask (e.g., KOPF, Stoerling). • Stereo microscope and cold light source (e.g., Leica, Zeiss). • DC temperature control system inclusive of heating pad, mini rectal temperature probe, and temperature controller (FHC). • Gas anesthesia induction chamber, delivery system, and charcoal-filtered scavenger (Harvard Apparatus, VetEquip). Use in a well-ventilated room to prevent accumulation of anesthetic vapors. • Dental drill and micromotor (Foredom Electric Company, Aseptico) and 0.5 mm drill bits (Fine Science Tools). • Animal warming cabinet (Datesand) Set at 30–32 °C. • Ultraviolet lamp, 365 nm (Cole-Palmer) for curing optical adhesive.

380

Mariangela Panniello et al.

2.2 Surgical Procedures: Consumables

• Artificial Cerebral Spinal Fluid (ACSF; individual compounds from Sigma Aldrich: NaCl 125 mM NaCl, 4.5 mM KCl, 26 mM NaHCO3, 1.25 mM NaH2 PO4, 2 mM CaCl2, 1 mM MgCl2, and 20 mM glucose). • Saline 0.9% solution for wound cleaning (Steripod). • Non-woven surgical sponges (Kettenbach GmbH). • Cyanoacrylate adhesive gel (Loctite). • Gel Foam (Moore Medical). • Surgical Skin Marker (Fine Science Tools). • Round coverglasses (Fisher Scientific, Harvard Biosciences). 3, 4, or 5 mm diameter depending on the configuration of choice, see Subheading 3.1). • Norland optical adhesive n. 71 (Norland Products). • C&B Metabond Quick Cement (Parkell). • Tungsten-carbide Fine Scissors (straight tip, Fine Science Tools). • Delicate bone scraper (Fine Science Tools). • Dumont #3 forceps (Fine Science Tools). • Ceramic-coated Dumont #5 forceps (Fine Science Tools, to use in case of durotomy). • Borosilicate glass capillaries (1 mm OD, 0.58 mm ID; World Precision Instruments). Pull a micropipette with needle’s taper longer than 2 cm. Break and bevel the tip. The final diameter of the tip should be 20–30 μm. Use a light microscope to check that the tip is clean and free from glass debris. • Viral vector (variable, according to experimental needs, Addgene). • Viral vector injecting system: a Hamilton Syringe connected to a micropipette and an infusion pump for slow delivery of small viral volumes (Stoelting). Alternatively, a pressure injection system (Nanoject II, connected to a micropipette). • Custom-designed headbars (preferably titanium, but stainless steel or aluminum may be used). • Kwik-Cast surgical silicone sealant (Kwik-Cast).

2.3 Pharmacological Agents

• Isoflurane (IsoFlo, Zoetis, UK). • Sodium chloride 0.9% solution injectable (for drug dilution, Moore Medical). • Dexamethasone sodium phosphate (corticosteroid antiinflammatory agent; 4 mg/mL at 2 mg/kg, Dexadreson, MSD).

Imaging Rodents

381

• Buprenorphine (opioid analgesic; 0.03 mg/mL at 0.1 mg/kg, Vetergesic). • Bupivacaine (local anesthetic; 0.25% solution at 2 mg/kg, Marcaine). • Lacrilube eye ointment (Alergan, UK). • Meloxicam (non-steroidal anti-inflammatory drug; 2 mg/mL at 5 mg/kg, Boheringer Ingelheim). 2.4 Two-Photon Imaging: Hardware

• Two-photon laser: tunable Ti:Sa with 80 MHz repetition rate (e.g., SpectraPhysics, Mai-Tai Deep See). • Multiphoton microscope: Bruker (e.g., Ultima Investigator, Ultima2Pplus), Sutter Instruments (e.g., MOM), Thorlabs (e.g., B-scope). • Objective Lens: preferably high numerical aperture and working distance between 2 and 3 mm (e.g., Nikon 16×/0.80 W; Nikon 25×/1.1 W). • Photomultiplier tubes: GaAsP are to be preferred for quantum efficiency, but are easily degradable by high-intensity incident light (e.g., Hamamatsu Photonics).

2.5 Two-Photon Imaging and Analysis: Software

• Image acquisition software (depending on the 2p microscope of choice: ScanImage; Prairie View Imaging).

2.6 Monitoring Mouse Behavior: Hardware

• Rotary encoder for recording mouse locomotion speed (e.g., HEDM-5500#B14, Broadcom).

3

• Matlab or Python, depending on the analysis toolbox of choice (e.g., Suite2P, CaImAn, OnACID).

• Camera for recording body movements (e.g., C920 or B920, Logitech; Basler AG; Motionscope M1, Redlake).

Methods All procedures must be approved by the local ethical review committees before being carried out. We will start by detailing the implantation of a chronic imaging window over vS1, followed by viral vector delivery of GECI. These two steps are preferably performed in one surgical procedure in the adult mouse. Alternatively, GECI delivery can be performed in newborn mice (postnatal day, P0–P2) by delivering the viral vector directly into the cortex [56], or in adult animals systemically into the ventricles [57] or the retroorbital vein [58].

382

Mariangela Panniello et al.

When GECIs are injected in newborn mice, implantation of the chronic cranial window for adult imaging should be performed when the skull of the experimental subject has reached its adult size (typically, at least 5 weeks of age). The chronic window implantation protocol described here can be also applied to transgenic mice constitutively expressing the GECI of interest. All implanted mice should be allowed a one-week period of recovery from the surgical procedure. If viral vector injection is also performed, 2 weeks should be added to the recovery time before any imaging can be carried out. This 3-week period is required for achieving optimal expression of the viral vector in the cells under study. During this time, mice can be prepared for the subsequent experimental phases. All animals who will undergo behavioral training should be habituated to head fixation and to handling by the experimenter in order to avoid major movement artifacts during imaging as well as to increase the probability of good performance during sensory training. At the end of 2–7 days of habituation, 2p chronic imaging procedures can start. These may last from a few days to several weeks, depending on the experimental questions and protocols. During this time, a large amount of imaging data may be recorded (several dozens of GB per hour of data collection, according to frame rate and resolution), and appropriate data processing and storage solutions should be set out in advance. 3.1 Surgical Procedures in Adult Mice: Cranial Window and Viral Vector Injection

Good quality 2p imaging recordings cannot be achieved without optimal neuronal labeling in the brain region under investigation. With this aim, and in order to reduce tissue damage, slow, minimally invasive, and stereotaxically precise viral injections must be carried out. The surgical area as well as the surgical tools should be sterilized before the procedures take place. The operator should wear sterile gloves, a sterile surgery coat, head-cap, and surgical mask. The surgical area should be equipped with a stereotaxic frame for rodents, a cold light source, a stereoscope, a heating pad connected to a rectal temperature probe, and an apparatus for delivery and scavenging of anesthetic gases. Sterile saline solution or artificial cerebrospinal fluid (ACSF) should be readily available. Weigh the mouse and prepare the appropriate volumes of pharmacological agents to be administered via subcutaneous (s.c.) or intramuscular (i.m.) injection. These should include a corticosteroid (e.g., dexamethasone) to reduce swelling and respiratory issues during surgery, an analgesic (e.g., buprenorphine), and a local anesthetic (e.g., Marcain, to be injected under the scalp). In addition, corticosteroid (e.g., dexamethasone) or non-steroid antiinflammatory agents (e.g., meloxicam) may be administered s.c. for 2–5 days after surgery to minimize inflammation and pain. Induce general anesthesia (isoflurane: 3 min at 3.5–4.0 L/min at 1% for induction; 0.8–1.5 L/min at 1% for surgery), shave the

Imaging Rodents

383

mouse head, carefully avoiding the whiskers, and stabilize the mouse in the stereotaxic frame, placing a heating pad underneath its body. Temperature, as well as breathing rate, should be monitored from now until the end of the surgical procedure to minimize post-surgical complications and guarantee speedy recovery from anesthesia. Body temperature should be between 36 and 38 °C, and breathing rate should be approximately 1 Hz. Cover the eyes with eye ointment (Lacrilube) to minimize dryness. Administer the pharmacological agents. Sterilize the scalp with a povidone-iodine solution (e.g., Betadine). Using a scalpel blade, make a small incision of the scalp, running for approximately 1 cm along the antero-posterior axis of the animal. Cut the skin flaps with surgical scissors to expose both parietal bones, the posterior half of the frontal bones, and part of the occipital bone (Fig. 2a). For chronic imaging in the mouse somatosensory cortex, both the temporal and the occipital muscles can be left untouched during the surgical procedure. Remove the periosteum by gently scraping the bone surface. This action will facilitate the attachment and increase the stability of the metallic headbar onto the skull. A solution of hydrogen peroxide may be used for more efficient removal: apply a small drop with a cotton bud and swiftly wash it out with sterile ACSF. To create enough space for the implantation of the chronic window and headbar, glue the skin edges to the skull using cyanoacrylate adhesive gel. Locate the stereotaxic coordinates corresponding to the center of vS1. In an adult mouse weighing between 22 and 28 g, this should be located 3.1 mm lateral from the midline and 1.3 mm posterior from the Bregma suture. Mark this point with a skin marker pen: it will correspond to the center of the craniotomy or to the site where the viral vector should be injected (Fig. 2a). Use a dental drill to obtain a circular craniotomy over vS1. The diameter of the incision should match the diameter of the cover slip that will subsequently be used to cover the exposed tissue (see Note 4.1). While drilling, wash regularly with cold ACSF to remove bone dust and cool down the bone as well as the tissue below. Once the bone flap is removed and the cortical surface clean, the viral vector can be injected (see Note 4.2). Injections should be performed using a glass capillary (tip diameter 20–30 μm) attached to an injection system for precise delivery of small volumes. The tip of the glass capillary should be beveled in advance to facilitate the penetration of the dura mater. Injections can be performed at multiple sites and cortical depths according to the experimental needs (e.g., imaging deep or superficial layers; imaging one or multiple barrels). The total injected volume may vary between approximately 50 nL and 1 μL, depending on the properties of the viral vector (e.g., the titer) and on the number of injection sites. Delivery should be extremely slow (50–70 nL/min), and a waiting time of 2 min between injections is recommended.

384

Mariangela Panniello et al.

Fig. 2 Surgical procedures for implantation of a chronic cranial window over vS1 and attachment of a metallic headbar. (a) The skull after scalp and periosteum removal. The skin edges are secured to the bone with cyanoacrylate glue so as to expose both parietal bones as well as part of the frontal and occipital bones. The black dot indicates the center of the barrel field (3.1 mm lateral from the midline and 1.3 mm posterior from the Bregma suture). Scale bar, 1.5 mm. (b) A picture showing a craniotomy over the right vS1. The exposed tissue is covered in ACSF. Scale bar, 3 mm. (c) The same mouse after implantation of a chronic imaging window (obtained by gluing together a 3 mm wide cover slip and a 4 mm wide one; total thickness 0.45 mm) and a titanium headbar. The headbar was first secured to the skull with a drop of cyanoacrylate gel and then covered with dental cement. (d) Two headbar designs for chronic 2p imaging. The top one was specifically designed with an asymmetric shape for imaging lateral areas of the dorsal cortex (e.g., somatosensory cortex). (e) Three mice recovering in a warm animal cabinet after viral injection and chronic window implantation

Once the viral vector has been injected, the glass cover slip (imaging implant) can be applied. In this phase, the exposed area and the surrounding bone should be dry. The implant may consist of two to three glasses stacked together using UV-cured optical glue or of an individual cover slip. The first option is recommended to maximize stability of the implant as well as to facilitate its

Imaging Rodents

385

removal if necessary [59]. However, a multi-layered glass window may lead to optical aberrations and poor image quality. Before covering the exposed brain area, the glass window should be soaked in a 75% ethanol solution for 5 min and then carefully flushed with sterile ACSF. Cover the edges of the glass window with a thin layer of gel cyanoacrylate glue and leave it to dry for 5 min. A gentle, continuous pressure may be applied to facilitate the adhesion between the glass and the bone. After the application of the optical implant, the metallic headbar for head-fixation can be implanted. The headbar should first be secured to the skull by means of a thin layer of cyanoacrylate glue. Dental cement (e.g., SuperBond) should then be generously applied to cover the whole exposed skull surface and the headbar (Fig. 2c). Care should be taken to avoid dropping dental cement onto the glass window. If necessary, the cement can immediately be removed with the help of a cotton bud soaked in acetone. Allow at least 10 min before the cement is completely dry. Make sure that the orientation of the headbar on the mouse’s head allows the animal to remain in a comfortable position while imaging (e.g., head straight and not tilted), and that it does not touch the eyes or the external ears. In the meantime, the glass cover slip should be covered with a drop of adhesive silicone elastomer (Kwik-Cast) until the first imaging session, in order to protect brain tissue from light. Remove the mouse from the stereotaxic frame and place it in an animal warming cabinet (Fig. 2e) with food pellets and water. 3.2 Imaging of the Intrinsic Optical Signal in Barrel Cortex

Intrinsic optical signal imaging (IOSI) in vS1 is used to identify the exact location of the barrel field. It relies on measuring the reflectance of local tissue, which changes depending on blood flow and oxygenation levels. It can therefore highlight local increases in neuronal activity. Thanks to the somatotopic mapping of whiskers in the barrel cortex, the deflection of a single whisker corresponds to a localized change in reflectance in vS1. By combining the observations for several whiskers, it is therefore possible to precisely map and locate imaging fields within vS1. Ideally, IOSI is performed during the surgical procedure described in Subheading 3.1 and, specifically, immediately before viral vector injection. However, this procedure may be impractical for several reasons: 1. Isoflurane anesthesia should be associated with the administration of chlorprothixene hydrochloride (i.m., 1 mg/kg), an anticholinergic agent able to inhibit whisker movements. 2. Surgical facilities may not be equipped with the IOSI apparatus.

386

Mariangela Panniello et al.

3. IOSI would substantially extend the duration of the surgical procedure, increasing the risk of postoperative complications. 4. Trimming a few whiskers may be necessary in order to optimize the quality of the recorded signal. Using the implanted head plates, mice should be head-fixed by a clamp on a heated mat. Identify a whisker that is long enough to be threaded easily through a glass capillary (usually, C2 or C3). Thread the whisker into the capillary tube attached to a ceramic piezoelectric stimulator (e.g., PB4NB2W Piezoelectric Bimorph Bending Actuator with Wires, Thorlabs). If the surrounding whiskers touch the external side of the capillary, carefully trim them using a pair of iris scissors under a dissecting microscope. Note that trimming should only be carried out if the intact whisker system is not required in any following experimental procedures. Setup a Retiga R1 camera with a 50 mm and a 135 mm lens (Nikon) attached in tandem configuration [60]. The whisker stimulation protocol consists of 1 s stimulation at 10 Hz with 20 s ITI, repeated 40 times, for a total of 400 deflections. Repeat this stimulation protocol 3–4 times per mouse on different whiskers in order to map the barrel fields in vS1. For post hoc confirmation of the imaging location in vS1, a map of vS1 barrels obtained through IOSI can be overlaid upon images of the areas investigated with 2p imaging (Fig. 3).

Fig. 3 Imaging of the intrinsic optical signal in five example mice (M1-5). For each mouse, whiskers were deflected individually and the corresponding barrels identified (yellow areas: maximal reflectance changes corresponding to increased neural activity). Several barrels were identified in each mouse. These were then mapped onto a barrel field representation (overlaid in gray). Areas targeted with two-photon imaging are shown with white frames. All panels are presented with the same orientation (A anterior, P posterior, M medial, L lateral)

Imaging Rodents

387

3.3 Habituation to Head-Fixation

Mice should be allowed 5–7 days to recover from the surgical procedure. After this period, each subject should be habituated to being handled by the experimenter as well as to being head-fixed for prolonged periods of time, the duration of which should be determined in agreement with the local ethical review committee and is generally not longer than 1.5 h. During the habituation phase, mice that will be engaged in a behavioral task during chronic 2p imaging will normally undergo a water-intake (or food-intake) regulation protocol (see [61] for further details). The experimenter should carefully handle each mouse for 10–20 min a day for 2–7 days, avoiding any potentially stressful approach with the subject, such as lifting by its tail or neck. Housing mice in groups whenever possible promotes their well-being and may facilitate interaction with the experimenter. Moreover, the use of plastic or paper tubes during the handling phase may reduce the number of sessions necessary to habituate the mouse, as these represent a safe, enclosed environment for rodents. Mice should then be habituated to the imaging apparatus for the following 2–3 days. Head-fixation should be performed for progressively increasing time periods, from a few minutes on the first day to 20–30 min on the last day. During head-fixation and imaging, mice typically rest inside a plastic tube [62] or run forward on a treadmill (usually made of Styrofoam, [7]). The first solution is preferred in order to reduce potential imaging artifacts due to mouse movement as well as to minimize the effect of locomotion of neural activity.

3.4 Chronic TwoPhoton Calcium Imaging in Barrel Cortex

Precise and minimally invasive viral vector injection together with a clear and stable head implant are crucial requirements for obtaining good-quality imaging data over the course of weeks or months. Once the chronic imaging protocol starts, nonetheless, it is as well important to limit both acute and long-term effects of photodamage on brain tissue [63]. The intensity of the two-photon laser recorded at the level of the objective should be kept low, and ideally not higher than 50 mW to avoid photobleaching and tissue heating [64, 65]. Scanning the FOV at high velocity (e.g., through resonant scanning solutions, 30 Hz or above) can also contribute to preserving the cerebral tissue, and especially small structures like dendrites, axons, or dendritic spines. Fast scanning may nonetheless require the experimenter to increase laser intensity in order to compensate for the reduction in dwell time per pixel, and these two factors should be appropriately balanced in order to obtain good quality images without damaging the tissue. Imaging quality may also drop whenever the sample (i.e., the glass window on vS1) is not perfectly parallel to the light source (i.e., the objective). Various devices are currently available to minimize the tilt angle and align the optical axis to the sample [66, 67]. These approaches are particularly useful when the

388

Mariangela Panniello et al.

experimental protocol requires imaging the activity of the same individual neurons for several subsequent sessions. For example, a laser beam can be directed to the surface of the imaging window and the reflected beam projected onto a target. By measuring the trajectory of the reflected beam, the angle between the imaging plane and the cranial window can be measured and the mismatch corrected for each day of imaging [67]. Researchers should also design robust animal stages to bolt into the imaging air-table at specific locations and couple them with headbar holders adjustable in the minimum number of possible dimensions, to maximize imaging stability (see Note 4.3). The first 2p imaging session may be dedicated to identify landmarks on the cortical surface (e.g., vasculature running parallel to the brain surface; Fig. 4a), which will serve as a reference during subsequent recordings. Alignment of FOVs across days may be performed manually or using commercial software (e.g., ScanImage, Vidrio Technologies).

Fig. 4 Chronic 2p imaging in vS1. (a) An image of the cortical surface of a transgenic mouse expressing the calcium indicator GCaMP6s under the CaMKII promoter. The image, covering the whole craniotomy, was obtained by stitching together 20 FOVs of 0.8 × 0.8 mm. This tiled image should be acquired before the start of the chronic 2p imaging protocol in order to visualize the anatomical extent of the calcium indicator expression. Moreover, assuming that the mouse will be head-fixed with the same head orientation throughout the chronic imaging protocol, the vasculature in this tiled image will serve as a reference for facilitating imaging of the same FOV across sessions. The blue square indicates the FOV represented in the inset on the right, located 250 μm below the cortical surface. Here, red arrows indicate filled neurons which express excessive levels of calcium indicator and should be excluded from analysis. (b) Example fluorescence traces extracted from 10 neurons in the FOV in a. Traces are corrected for neuropil contamination and normalized for baseline fluorescence (ΔF/F). (c) Each column shows the individual (gray) and mean (black) ΔF/F values of one neuron (blue circle in a) in response to 6 positional stimuli presented against the whiskers, from the most anterior (green bar) to the most posterior position (gray bar). The width of the bars indicates the duration of the contact between the stimulus (a metallic pole) and the whiskers. The mouse was trained to discriminate stimulus position while the same neurons were imaged throughout the learning period. The top row shows neuronal response before learning, while the bottom row shows the response of the same neuron when the mouse was expert on the discrimination task

Imaging Rodents

389

Most GCaMP calcium indicators are expressed in the cytosolic compartment and should not be present in the nucleus of labeled cells. These should therefore show, when imaged in vivo, a typical “ring” shape. Neurons appearing as full circles are likely to be expressing an excessive volume of calcium indicator; they may show aberrant activity and should be excluded from analysis (Fig. 4a). Nuclear filling should not be visible until several months from the viral injection [6]. If filled neurons are more than 5% of the total imaged cells 3 weeks after viral injection, we recommend reducing viral titer or volume during the injection procedure or expressing the transgene under a weaker promoter. 3.5 Monitoring Mouse Behavior

It is now commonly accepted that vS1 is a sensorimotor area, and stimulation of vS1 alone can drive whisker retraction in mice [68]. Whisking and locomotion are tightly related, to the extent that mice whisk whenever they run [10], although the opposite (i.e., when mice do not run, they do not whisk) is not always true [11]. Whisker movements, as well as mouse locomotion, can readily be tracked during 2p imaging experiments, allowing the relationship between whisking behavior and neural activity in vS1 to be investigated [7, 21, 69, 70]. Recording locomotion and whisker movement does not require a high-tech setup. When mice are head-fixed on a treadmill, a simple rotary encoder can track their running speed. Other setups, using computer mice, can allow for the tracking of the locomotion of mice on a spherical air-floated ball [71]. Common webcams combined with infrared lights can be sufficient to image whisker pads and correlate neural activity with overall mystacial movement [72]. In order to precisely track the dynamics of single whiskers (e.g., curvature when contacting the stimulus) higher acquisition rates are necessary and can be reached using specialized cameras (i.e., Basler AG [72]; or Redlake Motionscope M1). In this case, infrared light sources should be positioned appropriately with respect to the whiskers so as to precisely delineate their shapes and positions. In addition to whisker movements, pupil dilation can be monitored during 2p imaging in awake behaving mice. Indeed, pupil dilation is a proxy for behavioral state and correlates with cortical activity [73].

3.6

One of the disadvantages of 2p microscopy is that the acquired images require a substantial amount of pre-processing before data can be analyzed. Image processing methods are not standardized across laboratories, leading to inconsistencies among results and reproducibility issues. In an attempt to overcome this limitation, a number of open-source 2p imaging toolboxes have been released. To date, the most popular suites among the imaging community are Suite2p [74] and CaImAn [75]. Each of them offers ready-to-use scripts, in-depth documentation, and continuous maintenance by the developers.

3.6.1

Analysis Image Processing

390

Mariangela Panniello et al.

Two-photon acquisitions in vivo are often affected by displacement of the field of view due to movement of the experimental subject. These artifacts can be of two types: lateral artifacts (i.e., parallel to the imaging plane) or displacements along the vertical axis (i.e., perpendicular to the imaging plane). According to the severity of the lateral shift, several algorithms are available for image translation and displacement correction [76–81]. Movements occurring orthogonally to the focal plane, on the other hand, are difficult to correct for. Rapid, stereotyped z-movements may be due to heartbeat and breathing, while licking during imaging in headfixed awake behaving animals may produce larger and irregular displacements [82], which may cause neurons to intermittently appear and disappear from the FOV during the acquisition. Such effects inevitably lead to loss of data when single planes are imaged, and only a few corrective strategies have been described, relying on electrocardiogram signals or behavioral feedback to trigger compensatory adjustments of the focal plane [83, 84]. Nonetheless, z-plane artifacts can more easily be corrected when performing volumetric imaging, as the fluorescent transients of the ROIs in the central planes of each acquired volume can be recovered, frameby-frame, from the planes above or below [82, 85]. After correcting for movement artifacts, 2p images should be segmented, and regions of interest (ROIs) should be identified. These define the pixels from which fluorescent signals will be extracted and generally correspond to sub-cellular regions such as neuronal bodies, axons, or dendrites. ROIs can be defined manually using commercial software (e.g., ImageJ), in-house scripts, or automatically. Automatic image segmentation may be based on either anatomical [85] or functional criteria [74, 86]. Defining a ROI according to its morphological features (i.e., shape and spatial extent) may be preferable, as weakly active or silent cells may otherwise be erroneously discarded. Once image segmentation is complete, temporal fluorescence traces should be extracted by averaging, at each imaging frame, the signal recorded in the pixels defining each ROI. Because of the elongated point-spread function in 2p systems, ROIs are likely contaminated by signal generated above or below the focal plane. This signal may come from cellular bodies as well as from cellular projections such as dendrites and axons, generally denoted with the comprehensive term of “neuropil.” As the “neuropil” signal is not uniform throughout the FOV, its contamination should be corrected locally, separately for each identified ROI. Most approaches simply define a neuropil area surrounding each ROI, extract the mean signal from this area, and then subtract it frame-by-frame from the ROI signal [22]. Non-negative matrix factorization is also becoming a popular strategy for separating neuropil and ROI signal [74, 86]. Neuropil correction is essential in 2p studies as out-offocus contamination evens out signals across ROIs, leading to

Imaging Rodents

391

increased tuning similarity when measuring cellular responses to sensory stimuli (e.g., sensitivity to textures or positional stimuli against whiskers). After neuropil correction, the fluorescence time series corresponding to each ROI should be corrected for its baseline (F0) using the formula: (F(t) – F0)/F0, which is commonly denoted as ΔF/F0. This approach normalizes differences in overall fluorescence across ROIs that can occur due to differences in calcium indicator expression (Fig. 4b). Previously mentioned analysis toolboxes (e.g., Suite2P) enable tracking neuronal bodies when the same FOV is imaged throughout the chronic procedure. When sensory-guided behavioral paradigms are associated with chronic 2p imaging, this procedure allows direct comparison of stimulus responses in individual neurons before and after sensory learning has taken place (Fig. 4c). 3.6.2 Whisker Tracking Analysis

4

Image analysis to monitor whisking activity and behavioral state has been well documented, and several open-source toolboxes are available for feature extraction from movies monitoring mouse movement and/or pupil dilation (FaceMap [70]; DeepLabCut [87]). These packages allow the identification of regions of interest corresponding to disparate areas of the mouse body: whiskers, snout, tongue, pupil, paws. Motion vectors are then extracted from the ROIs, and these can be aligned with 2p imaging traces for further analysis.

Notes

4.1 Obtaining a Perfectly-Sized Craniotomy

Matching the size and shape of the craniotomy with those of the cover slip is possibly the most critical and important step in the surgical procedure to guarantee a good imaging quality for weeks or months after implantation. A craniotomy smaller than the cover slip can facilitate the formation of air bubbles between the glass and the tissue and, consequently, lead to infections and scar tissue formation. On the other hand, if a craniotomy is larger than the cover slip, the contact with the glass may damage the cortical tissue as well as reduce imaging stability. In order to ensure a good fit, consider the most appropriate cover slip size for the craniotomy. A 3 mm round cover slip will cover the whole surface of the mouse barrel cortex and will extend to adjacent areas such as other primary somatosensory regions or posterior parietal cortex. Place a spare glass cover slip on the skull and use a skin marker to draw its contour. A second strategy to optimize craniotomy shape and size is to use a biopsy punch of the same diameter as the cover slip. Gently press the punch onto the skull until a sulcus is clearly visible. Then use a dental drill along the outer edge of the sulcus (0.5 mm drill bit). Drilling should be slow and alternated with thorough ACSF

392

Mariangela Panniello et al.

washes. Applying continuous pressure to the bone and the underlying tissue may cause excessive heating, which can facilitate inflammation and, in turn, reduce the quality of the cranial window. Drill until the vasculature becomes clearly visible underneath the sulcus. Using the side of a pair of sterile tweezers, gently press down the central part of the skull area defined by the circular groove. If this moves up and down when applying and releasing pressure, it is time to remove the bone flap. First, soften the bone by covering the area with ACSF for at least 5 min. Then, use thin, sterile tweezers to very gently detach and lift the bone flap. Flush thoroughly with cold ACSF to remove bone debris and stop any active bleeding (Fig. 2b). 4.2 Performing a Durotomy to Improve Visibility

The steps outlined in Note 4.1 should allow the experimenter to obtain a clean and stable cranial window, which is crucial for longterm imaging procedures. However, applying excessive pressure while drilling may occasionally damage the cerebral tissue and, in particular, the dura mater—– the highly innervated outer layer of the meninges. Piercing the dura mater while performing a craniotomy will result in conspicuous bleeding and will negatively affect the quality of the imaging window. In such cases, it is advisable to attempt a removal of this thin protective layer, a procedure known as durotomy. After the craniotomy, wash the exposed area thoroughly with cold saline or ACSF to help bleeding stop and to avoid swelling. Administration of dexamethasone prior to the surgical procedure will also help contain the swelling. If bleeding continues, it is possible to apply a small piece of hemostatic sponge (Gelfoam) soaked in ACSF on the exposed area for 1–2 min. Gently remove the Gelfoam and cover the whole surface of the exposed tissue with cold saline or ACSF. Any dry area will easily stick to the tissue underneath and complicate the removal. Observe the craniotomy under the microscope and choose an area free from large blood vessels. The dura mater removal will start from here. The selected spot should be as distant as possible, within the craniotomy, from the barrel field, or from the area where imaging will be performed. Pierce the dura mater gently with a 20G needle. If the dura mater was already accidentally pierced while performing the craniotomy, then go to the next step and insert a ceramic-coated Dumont #5 forceps under the incision, keeping the forceps as parallel as possible to the brain surface. Make sure to avoid contact between the forceps and the brain tissue. Lift the forceps very carefully in order to pull off the dura, and keep an eye on the surrounding vessels. If bleeding starts, pause the removal and cover the area with cold ACSF or Gelfoam for a few minutes. Repeat until enough dura mater has been removed. We strongly recommend removing the dura mater from the whole exposed area, in order to minimize the imaging issues caused by the regrowth of the meninges. When finished, rinse with cold ACSF and keep the area covered until application of the chronic window.

Imaging Rodents

4.3 Designing a Headbar for 2p Imaging of vS1

393

Headbars for chronic recordings of cortical activity should be designed with the aim of maximizing stability during imaging, facilitating head-fixation by the experimenter as well as preserving the well-being of the experimental subject. Although more expensive, titanium may be preferred to stainless steel or aluminum to increase biocompatibility and reduce the weight of the bar, which should not exceed 10% of the mouse weight. In order to optimize stability, we recommend a bar that covers a large surface of the skull, possibly extending over both hemispheres; and one that is hollowed at its edges, allowing the researcher to head-fix the mouse to a support through two threaded screws rather than a clamp system. Two example designs, specifically recommended for 2p imaging of the mouse somatosensory cortex, are shown in Fig. 2d. By implanting the bar frontally to the cranial window, the distance between the bar and the window is maximized, allowing more space for the microscope lens to fit above the area of interest. Moreover, this configuration avoids obstructing the occipital bone of the mouse, which is crucial for balancing head position, particularly during locomotion.

References 1. Grienberger C, Konnerth A (2012) Imaging calcium in neurons. Neuron 73(5):862–885 2. Diamond ME, von Heimendahl M, Arabzadeh E (2008) Whisker-mediated texture discrimination. PLoS Biol 6(8):e220 3. Woolsey TA, Van der Loos H (1970) The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex: the description of a cortical field composed of discrete cytoarchitectonic units. Brain Res 17(2):205–242 4. Feldmeyer D, Brecht M, Helmchen F, Petersen CC, Poulet JF, Staiger JF et al (2013) Barrel cortex function. Prog Neurobiol 103:3–27 5. Lefort S, Tomm C, Sarria JCF, Petersen CC (2009) The excitatory neuronal network of the C2 barrel column in mouse primary somatosensory cortex. Neuron 61(2):301–316 6. Chen JL, Carta S, Soldado-Magraner J, Schneider BL, Helmchen F (2013) Behaviourdependent recruitment of long-range projection neurons in somatosensory cortex. Nature 499(7458):336–340 7. Ayaz A, St€auble A, Hamada M, Wulf MA, Saleem AB, Helmchen F (2019) Layer-specific integration of locomotion and sensory information in mouse barrel cortex. Nat Commun 10(1):1–14 8. Huber D, Gutnisky DA, Peron S, O’connor DH, Wiegert JS, Tian L et al (2012) Multiple

dynamic representations in the motor cortex during sensorimotor learning. Nature 484(7395):473–478 9. Guo ZV, Li N, Huber D, Ophir E, Gutnisky D, Ting JT et al (2014) Flow of cortical activity underlying a tactile decision in mice. Neuron 81(1):179–194 10. Sofroniew NJ, Vlasov YA, Hires SA, Freeman J, Svoboda K (2015) Neural coding in barrel cortex during whisker-guided locomotion. elife 4: e12559 11. Mayrhofer JM, Skreb V, von der Behrens W, Musall S, Weber B, Haiss F (2013) Novel two-alternative forced choice paradigm for bilateral vibrotactile whisker frequency discrimination in head-fixed mice and rats. J Neurophysiol 109(1):273–284 12. Stosiek C, Garaschuk O, Holthoff K, Konnerth A (2003) In vivo two-photon calcium imaging of neuronal networks. Proc Natl Acad Sci 100(12):7319–7324 13. Sato TR, Gray NW, Mainen ZF, Svoboda K (2007) The functional microarchitecture of the mouse barrel cortex. PLoS Biol 5(7):e189 14. Ohki K, Chung S, Ch’ng YH, Kara P, Reid RC (2005) Functional imaging with cellular resolution reveals precise micro-architecture in visual cortex. Nature 433(7026):597–603 15. Rothschild G, Nelken I, Mizrahi A (2010) Functional organization and population

394

Mariangela Panniello et al.

dynamics in the mouse primary auditory cortex. Nat Neurosci 13(3):353 16. Bonin V, Histed MH, Yurgenson S, Reid RC (2011) Local diversity and fine-scale organization of receptive fields in mouse visual cortex. J Neurosci 31(50):18506–18521 17. Panniello M, King AJ, Dahmen JC, Walker KM (2018) Local and global spatial organization of interaural level difference and frequency preferences in auditory cortex. Cereb Cortex 28(1): 350–369 18. Gaucher Q, Panniello M, Ivanov AZ, Dahmen JC, King AJ, Walker KM (2020) Complexity of frequency receptive fields predicts tonotopic variability across species. elife 9:e53462 19. Kerr JN, De Kock CP, Greenberg DS, Bruno RM, Sakmann B, Helmchen F (2007) Spatial organization of neuronal population responses in layer 2/3 of rat barrel cortex. J Neurosci 27(48):13316–13328 20. Martini FJ, Molano-Mazo´n M, Maravall M (2017) Interspersed distribution of selectivity to kinematic stimulus features in supragranular layers of mouse barrel cortex. Cereb Cortex 27(7):3782–3789 21. Peron SP, Freeman J, Iyer V, Guo C, Svoboda K (2015) A cellular resolution map of barrel cortex activity during tactile behavior. Neuron 86(3):783–799 22. Chen JL, Margolis DJ, Stankov A, Sumanovski LT, Schneider BL, Helmchen F (2015) Pathway-specific reorganization of projection neurons in somatosensory cortex during learning. Nat Neurosci 18(8):1101–1108 23. Zhang M, Kwon SE, Ben-Johny M, O’Connor DH, Issa JB (2020) Spectral hallmark of auditory-tactile interactions in the mouse somatosensory cortex. Commun Biol 3(1): 1–17 24. Bale MR, Bitzidou M, Giusto E, Kinghorn PF, Maravall M (2021) Sequence learning induces selectivity to multiple task parameters in mouse somatosensory cortex. Curr Biol 31(3): 473–485. https://doi.org/10.1016/j.cub. 2020.10.059 25. Banerjee A, Parente G, Teutsch J, Lewis C, Voigt FF, Helmchen F (2020) Value-guided remapping of sensory cortex by lateral orbitofrontal cortex. Nature 585(7824):245–250 26. Kwon SE, Yang H, Minamisawa G, O’Connor DH (2016) Sensory and decision-related activity propagate in a cortical feedback loop during touch perception. Nat Neurosci 19(9): 1243–1249 27. Brecht M, Sakmann B (2002) Dynamic representation of whisker deflection by synaptic potentials in spiny stellate and pyramidal cells

in the barrels and septa of layer 4 rat somatosensory cortex. J Physiol 543(1):49–70 28. Henze DA, Borhegyi Z, Csicsvari J, Mamiya A, Harris KD, Buzsaki G (2000) Intracellular features predicted by extracellular recordings in the hippocampus in vivo. J Neurophysiol 84(1):390–400 29. Denk W, Strickler JH, Webb WW (1990) Two-photon laser scanning fluorescence microscopy. Science 248(4951):73–76 30. Patterson GH, Piston DW (2000) Photobleaching in two-photon excitation microscopy. Biophys J 78(4):2159–2162 31. Denk W, Svoboda K (1997) Photon upmanship: why multiphoton imaging is more than a gimmick. Neuron 18(3):351–357 32. Svoboda K, Yasuda R (2006) Principles of two-photon excitation microscopy and its applications to neuroscience. Neuron 50(6): 823–839 33. Schultz SR, Copeland CS, Foust AJ, Quicke P, Schuck R (2016) Advances in two-photon scanning and scanless microscopy technologies for functional neural circuit imaging. Proc IEEE 105(1):139–157 34. Brondi M, Moroni M, Vecchia D, MolanoMazo´n M, Panzeri S, Fellin T (2020) Highaccuracy detection of neuronal ensemble activity in two-photon functional microscopy using smart line scanning. Cell Rep 30(8): 2567–2580 35. Grewe BF, Voigt FF, van’t Hoff M, Helmchen F (2011) Fast two-layer two-photon imaging of neuronal cell populations using an electrically tunable lens. Biomed Opt Express 2(7): 2035–2046 36. Katona G, Szalay G, Maa´k P, Kasza´s A, Veress M, Hillier D et al (2012) Fast two-photon in vivo imaging with threedimensional random-access scanning in large tissue volumes. Nat Methods 9(2):201–208 37. Chong EZ, Panniello M, Barreiros I, Kohl MM, Booth MJ (2019) Quasi-simultaneous multiplane calcium imaging of neuronal circuits. Biomed Opt Express 10(1):267–282 38. Sofroniew NJ, Flickinger D, King J, Svoboda K (2016) A large field of view two-photon mesoscope with subcellular resolution for in vivo imaging. elife 5:e14472 39. Tsien RY, Adams SR (1989) U.S. Patent No. 4,806,604. Washington, DC: U.S. Patent and Trademark Office 40. Paredes RM, Etzler JC, Watts LT, Zheng W, Lechleiter JD (2008) Chemical calcium indicators. Methods 46(3):143–151 41. Miyawaki A, Llopis J, Heim R, McCaffery JM, Adams JA, Ikura M, Tsien RY (1997)

Imaging Rodents Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388(6645):882–887 42. Dana H, Sun Y, Mohar B, Hulse BK, Kerlin AM, Hasseman JP et al (2019) Highperformance calcium sensors for imaging activity in neuronal populations and microcompartments. Nat Methods 16(7):649–657 43. Zhang Y, Rozsa M, Liang Y et al (2021) Fast and sensitive GCaMP calcium indicators for imaging neural populations. bioRxiv:11.08.467793 44. Dana H, Mohar B, Sun Y, Narayan S, Gordus A, Hasseman JP et al (2016) Sensitive red protein calcium indicators for imaging neural activity. elife 5:e12727 45. Tischbirek CH, Noda T, Tohmi M, Birkner A, Nelken I, Konnerth A (2019) In vivo functional mapping of a cortical column at singleneuron resolution. Cell Rep 27(5):1319–1326 46. Murayama M, Larkum ME (2009) In vivo dendritic calcium imaging with a fiberoptic periscope system. Nat Protoc 4(10):1551–1559. https://doi.org/10.1038/nprot.2009.142 47. Andermann ML, Gilfoy NB, Goldey GJ, Sachdev RN, Wo¨lfel M, McCormick DA et al (2013) Chronic cellular imaging of entire cortical columns in awake mice using microprisms. Neuron 80(4):900–913 48. Antonini A, Sattin A, Moroni M, Bovetti S, Moretti C, Succol F et al (2020) Extended field-of-view ultrathin microendoscopes for high-resolution two-photon imaging with minimal invasiveness in awake mice. elife 9: e58882. https://doi.org/10.7554/eLife. 58882 49. Dombeck DA, Harvey CD, Tian L, Looger LL, Tank DW (2010) Functional imaging of hippocampal place cells at cellular resolution during virtual navigation. Nat Neurosci 13(11):1433–1440 50. Klioutchnikov A, Wallace DJ, Frosz MH, Zeltner R, Sawinski J, Pawlak V et al (2020) Three-photon head-mounted microscope for imaging deep cortical layers in freely moving rats. Nat Methods 17(5):509–513 51. Villette V, Chavarha M, Dimov IK, Bradley J, Pradhan L, Mathieu B et al (2019) Ultrafast two-photon imaging of a high-gain voltage indicator in awake behaving mice. Cell 179(7):1590–1608 52. Watakabe A, Ohtsuka M, Kinoshita M, Takaji M, Isa K, Mizukami H et al (2015) Comparative analyses of adeno-associated viral vector serotypes 1, 2, 5, 8 and 9 in marmoset, mouse and macaque cerebral cortex. Neurosci Res 93:144–157

395

53. Nathanson JL, Yanagawa Y, Obata K, Callaway EM (2009) Preferential labeling of inhibitory and excitatory cortical neurons by endogenous tropism of adeno-associated virus and lentivirus vectors. Neuroscience 161(2):441–450 54. Chen TW, Wardill TJ, Sun Y, Pulver SR, Renninger SL, Baohan A et al (2013) Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499(7458):295–300 55. Daigle TL, Madisen L, Hage TA, Valley MT, Knoblich U, Larsen RS et al (2018) A suite of transgenic driver and reporter mouse lines with enhanced brain-cell-type targeting and functionality. Cell 174(2):465–480 56. He CX, Arroyo ED, Cantu DA, Goel A, Portera-Cailliau C (2018) A versatile method for viral transfection of calcium indicators in the neonatal mouse brain. Front Neural Circuits 12:56 57. Kim JY, Ash RT, Ceballos-Diaz C, Levites Y, Golde TE, Smirnakis SM, Jankowsky JL (2013) Viral transduction of the neonatal brain delivers controllable genetic mosaicism for visualising and manipulating neuronal circuits in vivo. Eur J Neurosci 37(8):1203–1220 58. Challis RC, Kumar SR, Chan KY, Challis C, Beadle K, Jang MJ et al (2019) Systemic AAV vectors for widespread and targeted gene delivery in rodents. Nat Protoc 14(2):379–414 59. Goldey GJ, Roumis DK, Glickfeld LL, Kerlin AM, Reid RC, Bonin V et al (2014) Removable cranial windows for long-term imaging in awake mice. Nat Protoc 9(11):2515 60. Juavinett AL, Nauhaus I, Garrett ME, Zhuang J, Callaway EM (2017) Automated identification of mouse visual areas with intrinsic signal imaging. Nat Protoc 12(1):32 61. Guo ZV, Hires SA, Li N, O’Connor DH, Komiyama T, Ophir E et al (2014) Procedures for behavioral experiments in head-fixed mice. PLoS One 9(2):e88678 62. O’Connor DH, Peron SP, Huber D, Svoboda K (2010) Neural activity in barrel cortex underlying vibrissa-based object localization in mice. Neuron 67(6):1048–1061 63. Holtmaat A, Bonhoeffer T, Chow DK, Chuckowree J, De Paola V, Hofer SB et al (2009) Long-term, high-resolution imaging in the mouse neocortex through a chronic cranial window. Nat Protoc 4(8):1128 64. Podgorski K, Ranganathan G (2016) Brain heating induced by near-infrared lasers during multiphoton microscopy. J Neurophysiol 116(3):1012–1023. https://doi.org/10. 1152/jn.00275.2016 65. Picot A, Dominguez S, Liu C, Chen IW, Tanese D, Ronzitti E, Berto P,

396

Mariangela Panniello et al.

Papagiakoumou E, Oron D, Tessier G, Forget BC, Emiliani V (2018) Temperature rise under two-photon optogenetic brain stimulation. Cell Rep 24(5):1243–1253.e5. https://doi. org/10.1016/j.celrep.2018.06.119 66. Kawakami R, Sawada K, Kusama Y, Fang YC, Kanazawa S, Kozawa Y et al (2015) In vivo two-photon imaging of mouse hippocampal neurons in dentate gyrus using a light source based on a high-peak power gain-switched laser diode. Biomed Opt Express 6(3):891–901 ˜ anes GL, Marchand PJ, Turcotte R, 67. Galin Pellat S, Ji N, Huber D (2018) Optical alignment device for two-photon microscopy. Biomed Opt Express 9(8):3624–3639 68. Matyas F, Sreenivasan V, Marbach F, Wacongne C, Barsy B, Mateo C et al (2010) Motor control by sensory cortex. Science 330(6008):1240–1243 69. Clack NG, O’Connor DH, Huber D, Petreanu L, Hires A, Peron S, Myers EW (2012) Automated tracking of whiskers in videos of head fixed rodents. PLoS Comput Biol 8(7):e1002591 70. Stringer C, Pachitariu M, Steinmetz N, Reddy CB, Carandini M, Harris KD (2019) Spontaneous behaviors drive multidimensional, brainwide activity. Science 364(6437):255 71. Erisken S, Vaiceliunaite A, Jurjut O, Fiorini M, Katzner S, Busse L (2014) Effects of locomotion extend throughout the mouse early visual system. Curr Biol 24(24):2899–2907 72. Musall S, Kaufman MT, Juavinett AL, Gluf S, Churchland AK (2019) Single-trial neural dynamics are dominated by richly varied movements. Nat Neurosci 22(10):1677–1686 73. Lee CC, Kheradpezhouh E, Diamond ME, Arabzadeh E (2020) State-dependent changes in perception and coding in the mouse somatosensory cortex. Cell Rep 32(13):108197 74. Pachitariu M, Stringer C, Dipoppa M, Schro¨der S, Rossi LF, Dalgleish H et al (2017) Suite2p: beyond 10,000 neurons with standard two-photon microscopy. Biorxiv 75. Giovannucci A, Friedrich J, Gunn P, Kalfon J, Brown BL, Koay SA et al (2019) CaImAn an open source tool for scalable calcium imaging data analysis. elife 8:e38173 76. Dombeck DA, Khabbaz AN, Collman F, Adelman TL, Tank DW (2007) Imaging large-scale

neural activity with cellular resolution in awake, mobile mice. Neuron 56(1):43–57 77. Giovannucci A, Friedrich J, Kaufman M, Churchland A, Chklovskii D, Paninski L, Pnevmatikakis EA (2017) Onacid: online analysis of calcium imaging data in real time. In: Advances in neural information processing systems 30, pp 2382–2392 78. Guizar-Sicairos M, Thurman ST, Fienup JR (2008) Efficient subpixel image registration algorithms. Opt Lett 33(2):156–158 79. Greenberg DS, Kerr JN (2009) Automated correction of fast motion artifacts for two-photon imaging of awake animals. J Neurosci Methods 176(1):1–15 80. Muir DR, Roth M, Helmchen F, Kampa B (2015) Model-based analysis of pattern motion processing in mouse primary visual cortex. Front Neural Circuits 9:38 81. Pnevmatikakis EA, Giovannucci A (2017) NoRMCorre: an online algorithm for piecewise rigid motion correction of calcium imaging data. J Neurosci Methods 291:83–94 82. Andermann ML, Kerlin AM, Reid C (2010) Chronic cellular imaging of mouse visual cortex during operant behavior and passive viewing. Front Cell Neurosci 4:3 83. Nimmerjahn A, Kirchhoff F, Helmchen F (2005) Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science 308(5726):1314–1318 84. Chen JL, Pf€a ffli OA, Voigt FF, Margolis DJ, Helmchen F (2013) Online correction of licking-induced brain motion during two-photon imaging with a tunable lens. J Physiol 591(19):4689–4698 85. Kaifosh P, Zaremba JD, Danielson NB, Losonczy A (2014) SIMA: Python software for analysis of dynamic fluorescence imaging data. Front Neuroinform 8:80. https://doi. org/10.3389/fninf.2014.00080 86. Pnevmatikakis EA, Soudry D, Gao Y, Machado TA, Merel J, Pfau D et al (2016) Simultaneous denoising, deconvolution, and demixing of calcium imaging data. Neuron 89(2):285–299 87. Mathis A, Mamidanna P, Cury KM, Abe T, Murthy VN, Mathis MW, Bethge M (2018) DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat Neurosci 21(9):1281–1289

Chapter 18 Imaging Somatosensory Cortex: Human Functional Magnetic Resonance Imaging (fMRI) Alexander M. Puckett and Rosa M. Sanchez Panchuelo Abstract Functional magnetic resonance imaging (fMRI) is a powerful tool for imaging somatosensory cortex, providing a means to non-invasively measure cortical activity in awake and behaving humans. Notably, this technique has permitted the homunculus—a hallmark of primary somatosensory cortex (S1) organization—to be examined with unprecedented detail. With the development of high-resolution fMRI (mostly at ultra-high field, 7 Tesla), it is now possible to investigate the finer topographic details of the sensory homunculus in almost any individual. Moreover, fMRI can be used to investigate other various bottom-up response properties as well as more top-down perceptual and cognitive processes (e.g., attention and prediction) across a wide range of experimental conditions. This chapter mainly focuses on tactile experiments, outlining a number of experimental paradigms and analysis techniques; practical and participant-specific difficulties are noted. Although we focus on fMRI for imaging primary somatosensory cortex, this technique can also be used to image cortical activity in other areas involved in somatosensory processing, such as secondary somatosensory cortex (S2), insular cortex, or the cerebellum. Key words Somatotopic mapping, Digits, Receptive fields

1

High-resolution,

Neuroimaging,

Hemodynamic response,

Introduction The somatosensory homunculus, put forth by Penfield and Boldrey [1], has become a mainstay in basic as well as clinical neuroscience. The homunculus serves as a pictorial representation of the localization of function in human somatosensory cortex and was born from the compilation of results from a large number of instances in which electrical stimulation was applied directly to the cortex of patients undergoing neurosurgery. This work supported previous findings in animal models, providing evidence that there is an orderly representation of the body in the brain. Not only this but the representation is somatotopically organized—there is a one-to-one correspondence between an area of the body and a specific region in the cortex.

397

398

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Although details of the original homunculus—drawn over 80 years ago—have been called into question [2, 3], the central idea conveyed by the sensory homunculus is as relevant as ever. Further detailing the particulars of this body-to-brain mapping has greatly benefited from the advent of high-resolution functional magnetic resonance imaging (fMRI). Whereas the original, invasive work by Penfield and Boldrey led to a coarse understanding of the somatotopic organization by making sparse measurements across a large number of patients, high-resolution fMRI permits similar information to be gleaned— but non-invasively in individuals. By measuring the tactile somatotopic maps in a more complete and detailed fashion in individuals, it is possible to gain a clearer understanding of the homunculus. Repeated measurements across a number of participants allow the variability across individuals to be studied, and a probabilistic atlas of functional cortical fields to be formed [4]. Importantly, the non-invasive nature of fMRI provides the opportunity to examine these maps in nearly any individual, rather than only those undergoing surgery. Note that there are exclusion criteria preventing some individuals from entering the MR environment (e.g., those with pacemakers). It also means there is no need to contend with dangers such as bleeding that are faced when performing direct electrical stimulation of the cortex, nor are there the same time constraints. With fMRI experiments, each individual can be scanned many times—allowing for a more complete examination of the somatotopic maps. Today it is possible, using fMRI, to resolve the cortical location of a number of different body part representations in human S1, such as the hands, feet, and face [5–7]. More impressive yet is the detail within these body part representations that can be resolved. For example, it is possible to individuate the digits of the hands and feet—enabling the cortical representation of each individual finger [8] or toe [9] to be mapped. It is also possible to map out different locations across the face and even around the lips [10]. Before the development of fMRI, localized cortical activity could be measured non-invasively in S1 in response to tactile stimulation using a different imaging technique, positron emission tomography (PET) [11, 12]. Early fMRI studies, performed at a magnetic field strength of 1.0–1.5 T, produced a number of relatively coarse—but convincing—examples of localized, somatotopically organized responses in the CNS elicited through stimulation of the tactile system [13–17]. Then, as fMRI continued to develop, so did the ability to resolve somatotopic maps. Using fMRI at 3–4 T it became possible to reliably identify cortical responses to stimulation of individual fingers [18–20]. And now, as illustrated in Fig. 1, it is possible to measure detailed maps of each individual finger using 7 T fMRI. Notably, the somatotopic organization can be measured not only across digits [8, 21] but also within digits

Human fMRI

399

Fig. 1 Somatotopic mapping with fMRI: (a) Schematic of somatosensory cortical areas with digit representations visible in both areas 1 and 3b. (b) Digit somatotopic map to vibrotactile stimulation of the left hand overlaid onto an axial slice of fMRI data. Note that a phase-encoded design was used, and hence the phase indicates the preferred digit as shown in the color wheel. (c) Somatotopic maps for both hands shown on inflated contralateral cortical surfaces and flattened patches of the central sulcus. Dark gray cortical regions represent sulci, while light gray regions represent gyri

[22], revealing that human S1 contains multiple orderly somatotopic maps of the fingers. These maps have been shown to correspond with different Brodmann areas and, importantly, provide a means to functionally demarcate these areas in humans. Not only has fMRI enabled human cortical somatotopic maps to be measured with far greater detail than was possible through sparse invasive measurements, but it has also enabled the maps to be examined over time and in unique populations. Using fMRI to measure somatotopic organization in the same individuals at multiple time points has revealed the somatotopic maps to be remarkably stable [23], with the maps even persisting years after loss of a limb [24]. Although the ability to non-invasively access S1 using MRI has proven valuable for mapping its somatotopic organization, it holds potential in many other areas of somatosensory study as well. Since fMRI can measure cortical activity under any number of experimental conditions, it provides a means to investigate a range of cortical processes. For example, fMRI can be used to investigate bottom-up response properties such as the effects of stimulation frequency, neuronal adaptation, or cross-digit suppression [25], as well as top-down processes such as those associated with attention [26–29] and prediction [30]. MRI can also be used to image cortical anatomy, providing a means to quantitatively compare structural and functional information at the same location in the brain, in the same individuals [6, 31]. When using fMRI to investigate cortical function, it is important to understand that the signal recorded is not a direct measure

400

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

of neuronal activity. Rather, the fMRI response reflects changes in blood properties linked to changes in neuronal activity—and hence it is referred to as a hemodynamic response. Broadly speaking, increases in neuronal activity are accompanied by increases in the metabolism of oxygen (and hence a decrease in oxygen concentration). However, a vascular response overcompensates for this consumption of oxygen by supplying more fresh oxygenated blood to the region than is needed to balance the oxygen used. This results in an increase in the concentration of oxygen in and near regions of increased cortical activity. Given that oxygenated and deoxygenated blood have different magnetic properties, this change can be detected using fMRI. More specifically, however, it is important to note that different imaging sequences (the particular set of radio frequency pulses and gradients used to acquire images) are sensitive to different hemodynamic properties [32]. Some sequences are sensitive to total blood oxygenation, whereas others are more sensitive to changes in blood flow or blood volume (see Note 4.2). The most established and widely used fMRI sequences detect blood-oxygenation-level-dependent (BOLD) changes [33, 34]. Gradient-echo (GE) based acquisitions are the preferred choice due to their easy implementation and the superior sensitivity of GE-BOLD contrast; however, spin echo (SE)-BOLD contrast offers improved spatial specificity. Improved specificity, at the cost of reduced sensitivity, can also be achieved with other techniques. Arterial spin labeling (ASL) [35, 36] can measure the direct change in cerebral blood flow (CBF) due to neuronal activation, and by imaging at shorter delays (when the tagged, or labeled, spins still primarily reside in the arteries/arterioles), ASL can also be sensitive to arterial cerebral blood volume (CBV). In contrast, the more recently developed vascular-space-occupancy (VASO) technique is sensitive to changes in total CBV (i.e., arterial and venous) [37]. Notably, changes in CBF and CBV both contribute to changes in blood oxygenation, but there are additional factors such as the aforementioned metabolic consumption of oxygen [38]. The relationship among each of these factors, how they influence the BOLD response and how they relate to underlying changes in neuronal activity, are still areas of active research. Irrespective of the specific sequence type, one of the key imaging parameters impacting what can or cannot be measured in any given experiment is the nominal spatial resolution. The spatial resolution is intrinsically limited by hemodynamic blurring (i.e., the spatial spread of the neurovascular response), and methods to suppress the contribution from draining veins are required for ultra-high spatial resolution studies. High spatial resolution is essential for mapping the finer topographic details of the body representation in S1. With increasing spatial resolution (owed in large part to increasing field strength as outlined above), it has been possible to progress from only being able to coarsely identify

Human fMRI

401

cortical activity associated with each side of the body to specific body parts (e.g., hand vs. foot) to different parts of body parts (e.g., individual digits on a hand), and even to different regions within these parts of body parts (e.g., within-digit mapping). Once in the sub-millimeter range, layer (or cortical-depth-dependent) fMRI also begins to become possible. It has been shown that layer fMRI can be used to probe the neural circuitry underlying feedforward and feedback processing of sensory information in somatosensory cortex, as these processes differentially activate cells in different cortical layers [30]. Running orthogonal to the cortical layers in area 3b of the somatosensory cortex, animal studies have revealed a finer columnar organization along the lengths of the digit representations, consisting of slowly adapting (SA) and fast adapting (FA) clusters of neurons in the middle layers of the cortex [39]. These FA/SA columns are at approximately the same spatial scale (the mesoscale) as the layers. Despite cortical columns having first been discovered in somatosensory cortex, most work to date regarding the use of fMRI to image columns has been performed in the visual system [40–42], given the difficulty in selectively stimulating a given mechanoreceptor class with standard mechanical stimulation techniques. Non-invasive imaging of cortical column activity in somatosensory cortex remains an exciting and potentially fruitful area of exploration. Temporal resolution is another key consideration when performing fMRI. The hemodynamic response measured by fMRI is much slower than the underlying neuronal responses. For instance, the BOLD response exhibits a 1–2 s delay compared to neuronal activity, has a temporal width of 4–6 s, and is marked by a post-stimulus undershoot that can last 30 s or more [43]. Given the rather slow time-course of the hemodynamic response, fMRI sequences with a temporal resolution around 2 s are typical and sufficient for most studies. With more advanced hardware and accelerated acquisition techniques [44, 45], sub-second resolutions are becoming more common—although the analysis of such data requires slightly different statistical treatment compared to more conventional temporal resolutions [46]. Given the sluggish dynamics of the BOLD response, however, it has been questioned what such an increase in the temporal resolution actually affords the researcher. For many experimental designs, this increase in temporal resolution is unnecessary; however, it has been shown that collecting fMRI data with the appropriate experimental design at a sub-second (3 T) with the use of multi-channel receive head coil arrays (e.g., 32-channel). It is possible to perform reasonably high spatial resolution (e.g., sub-millimeter) measurements at 3 T [49, 63]. However, this task becomes easier as field strength increases, given the concomitant increase in both signal strength and BOLD contrast, which is not diminished by increased physiological fluctuations, since thermal noise dominates in acquisitions with small voxel volumes [64]. The coils used during fMRI experiments to transmit and/or receive RF

Human fMRI

403

Fig. 2 Common experimental hardware. Top, left: MAGNETOM 7T whole-body research scanner (Siemens Healthcare, Erlangen, Germany). Top, right: 32-channel head-coil (Nova Medical, Wilmington, US). Bottom, left: Piezoelectric vibrotactile stimulator (Hybridmojo, San Francisco, US). Bottom, middle: MR-compatible earbuds (MR Confon, Magdeburg, Germany). Bottom, right: Button box (Current Designs, Philadelphia, US). All pictures taken at the Centre for Advanced Imaging (University of Queensland, Australia)

signals are another important hardware choice, and there is a wide range of coil types available. Not all coils are suitable for neuroimaging, but within those that are, there is still considerable variability. For example, some coils are designed for imaging large regions (e.g., arrays of coils for imaging the whole brain), while local surface receive coils can be carefully placed over the somatosensory cortex to boost the signal-to-noise ratio (SNR) locally, such that robust somatotopic maps can be obtained with improved spatial resolution [65]. Recently, there has even been a custom, two-helmet coil developed to permit the brain activity of two participants to be simultaneously imaged while facing one another inside the bore of the MRI scanner—enabling a number of specialized examinations for those interested in studying the neuroscience underlying social and interpersonal touch [66]. Highresolution fMRI typically uses multi-channel arrays of coils, as these enable the parallel imaging and image acceleration critical for achieving such high spatial resolution [67]. Apart from the scanner itself, there is a range of peripheral hardware used during somatosensory imaging experiments (e.g., Fig. 2, bottom). This hardware is typically experiment-specific and includes equipment used to apply tactile stimulation, present visual or auditory cues, and collect participant responses or physiological

404

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Table 1 Stimulating somatosensory cortex in the MR environment Type

Description

Studies

Manual

Tactile pressure manually administered by a member of the [13–15, 21, 26, 27, 157– 159] research team. Examples include applying pressure with a plastic rod, brush, or filament

Mechanical

Tactile pressure mechanically administered using a device. Examples include piezoelectric- and pneumatic-driven vibrotactile stimulation

[7, 8, 17–19, 22, 28, 29, 119, 122, 160–163]

Heat

Heat applied to body surface. Examples include the use of thermodes and lasers

[130, 142, 164, 165]

Electrical

Electricity administered to stimulate the somatosensory [14, 16, 120, 166–170] system. Examples include electrical stimulation applied at the skin surface, the median nerve, or directly into individual tactile afferents (i.e., microstimulation)

Air

[10, 14, 164, 171–173] Pneumatic stimulation of the body surface. Examples include the delivery of streams or puffs of air using various devices

Movement and Somatosensory stimulation resulting from the movement of [6, 23, 154, 173–175] Active Touch body parts with or without extereoceptive input. Examples without include simple finger flexion and extension. Examples with include pressing buttons, handling objects, and haptic exploration

recordings during the scan. Here, the most important requirement for the hardware is that it be MR-compatible. This does limit options; however, an increasing number of compatible devices are being manufactured—both by small groups of scientists as well as by specialized companies (e.g., Dancer Design in the UK and QuaeroSys in Germany). As outlined in Table 1, over the years there has been a wide range of stimulation methods safely used in the MR environment. The simplest of these involves a researcher brushing or otherwise manually stimulating some body part of a research participant lying in the scanner. More controlled, mechanical stimulation can be achieved using piezoelectric- or pneumaticdriven devices, whereas other stimuli, such as air puffs, heat (e.g., thermodes and lasers), or direct electrical current, offer other wellcontrolled means to deliver somatosensory stimulation. In contrast to these passively applied stimulation approaches, it is also possible to activate somatosensory cortex by having the participant actively stroke or explore tactile objects or textures while within the scanner. Most stimulus devices are computer-controlled and, as such, require a computer, typically located in a room adjacent to the scanner room. The computer communicates with the device inside the scanner, usually through a set of long cables. The same

Human fMRI

405

computer is often used for cueing the participant and recording behavioral responses and physiological signals. Cues provided visually can be delivered through a mirror system that reflects an image from a display outside the scanner to the participant’s field-of-view within the scanner, whereas auditory cues can be delivered via MR-compatible headphones. Behavioral responses are typically collected via MR-compatible button boxes or foot pedals, and physiological signals such as heart rate and breathing can be collected using a pulse oximeter (for a photoplethysmograph; alternatively a vectorcardiogram can be obtained using four electrodes placed over the chest) and a pneumatic respiratory belt, respectively [68]. Somatosensory fMRI experiments typically require the use of at least two different software packages: one for stimulus presentation and the other for data analysis. There are a number of available options for each, with a few outlined in Table 2. Deciding which software package(s) to use will depend on a number of factors. All the software listed in Table 2 is freely available; however, not all are compatible with every operating system. Many of the software packages offer similar functionality, but each comes with its own set of advantages and disadvantages (e.g., some are more versatile, some are integrated or compatible with others, some are more widely used or actively supported). For notes on each author’s approach to stimulus presentation and their analysis pipelines, see Note 4.3.

3

Methods When imaging somatosensory cortex using fMRI, there are a few common procedures as well as a number of important methodological considerations. Below, we aim to outline these in general, highlighting multiple specific options throughout. First, procedures for data collection (Subheading 3.1) will be reviewed before moving on to preprocessing of the data (Subheading 3.2). The methodological details outlined in these two sections apply irrespective of the particular nature of the experiment. In the final section, methodological details are outlined for a number of specific somatosensory fMRI paradigms and analyses (Subheading 3.3). Given that data collection is performed on human participants, there are a number of ethics and privacy issues that need to be considered and practices established before experiments can begin. It is recommended that you consult your local ethics committee to better understand the local and national guidelines governing ethical research practice.

3.1

Data Collection

Generally, fMRI experiments begin outside of the scanner with instruction and training. Once ready, the participant is provided hearing protection as scans tend to be uncomfortably loud (e.g.,

406

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Table 2 Software for stimulus presentation and data analysis Type

Name

Description and source

Stimulus Psychtoolbox A freely available set of Matlab and GNU Octave functions for vision and presentation neuroscience research. Aimed at making it easy to synthesize and show accurately controlled visual and auditory stimuli and interact with the observer. Supported on Linux, macOS, and Windows http://psychtoolbox.org/ MGL A freely available set of Matlab functions for displaying visual stimuli and writing experimental programs. It wraps the most commonly used graphics standard (OpenGL). Timing can be accurately kept and input from the keyboard, mouse and other devices can be finely controlled. The latest version is only supported on macOS. It is also integrated with mrTools, which is a set of Matlab tools to analyse fMRI data (see below) http://gru.stanford.edu/doku.php/mgl/overview PsychoPy A freely available, open-source software package written in Python for running a wide range of experiments in the behavioral sciences https://www.psychopy.org/ Cogent A freely available PC-based Matlab toolbox for presenting stimuli and recording responses with precise timing http://www.vislab.ucl.ac.uk/cogent.php Data analysis

AFNI/ SUMA

FreeSurfer

SPM

mrTools

FSL

AFNI is a freely available software suite of C, Python, R, and shell scripts, primarily developed for the analysis and display of anatomical and functional MRI data SUMA is a program that adds cortical surface-based functional imaging analysis to the AFNI suite AFNI/SUMA is available for macOS, Windows, and Linux/Unix https://afni.nimh.nih.gov/ A freely available, open-source software suite for processing and analyzing human brain MRI images. FreeSurfer includes tools to conduct both volume- and surface-based analyses and is often used to segment anatomical images in order to construct anatomical surface models FreeSurfer is available on macOS and Linux https://surfer.nmr.mgh.harvard.edu/ A freely available software suite of Matlab functions and subroutines with some externally compiled C routines—designed for the analysis of brain imaging data sequences. Can be used for analysis of fMRI, PET, SPECT, EEG, and MEG https://www.fil.ion.ucl.ac.uk/spm/ A free set of Matlab tools to analyze fMRI data (e.g., Fourier, eventrelated, population receptive field, and GLM analyses) and visualize results (e.g., on in-plane anatomies, flat maps, and surfaces) http://gru.stanford.edu/doku.php/mrtools/overview A freely available library of analysis tools for fMRI, MRI, and DTI brain imaging data. Runs on macOS and PCs (Linux and Windows via a Virtual Machine) https://fsl.fmrib.ox.ac.uk/fsl/fslwiki

Human fMRI

407

echo planar sequences are known to produce sound pressures in the range of 110–120 dB) and then situated on the bed of the scanner. Any somatosensory stimulators are secured in position, and response buttons (if used) and an emergency squeeze ball (for stopping the scan) are given to the participant. It is important to try to minimize participant head motion during the scans, and there are a few techniques that can be used. These range from simply padding the excess space between the participant’s ears and the surrounding head coil, to using a bite bar [69] or a custom plaster head holder [70] to restrict an individual’s movement. Once situated and comfortable (see Note 4.4 for more on participant comfort), the scan may begin. Before collecting data, B0 shimming is performed to ensure the main magnetic field is sufficiently homogeneous, as inhomogeneities can lead to susceptibility artifacts, for example, geometric distortions, signal loss, and blurring [71, 72] (see Note 4.5). It is then common to collect an anatomical dataset of the participant’s brain using one of a number of different structural imaging sequences (e.g., MPRAGE, MP2RAGE [73–75]. Although an anatomical dataset does not need to be collected first, if it is, it can then be used as a reference when positioning the acquisition slab (i.e., the volume being imaged) for the functional measurements. If a whole-head anatomical dataset is not needed (e.g., if you have already collected one during a previous scan session), the functional acquisition slab can instead be positioned using a quick set of “scout” images; however, these lack the fine structural detail present in the full, high-resolution anatomical scan. Whereas this detail is not necessary for positioning the acquisition slab for whole-brain or even many partial coverage functional studies, it can become more crucial when attempting to precisely target a small, focal patch of cortex. Which images are collected will affect the overall scan time (see Note 4.6). Once positioned, the functional imaging may begin. The following are some considerations for the optimization of fMRI protocols using echo planar imaging (EPI), which is the most widely used technique for fMRI. Multi-slice acquisitions (2D) can be performed using GE- or SE-EPI (less commonly used), but the maximum number of slices achievable within a given repetition time (TR)—which determines the temporal resolution—will be lower for SE-EPI due to tissue heating or specific absorption rate limitations. If whole-brain (cortex) coverage is required and simultaneous multi-slice (SMS) acquisition technology is available, axial slices spanning the whole brain can be collected at reasonable spatial resolution (1.25–1.50 mm isotropic) within a TR of 1–2 s—depending on the SMS factor. Axial images acquired with phase encoding in the anterior-posterior direction suffer less from geometric distortion than those with phase encoding in the rightleft direction. If multi-slice acquisitions are not available, a slightly

408

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

oblique coronal orientation can provide, for example, simultaneous S1 and S2 coverage. In this case, the acceleration factors (i.e., SENSE or GRAPPA) can be increased with respect to protocols that use SMS, hence minimizing geometric distortions [76, 77]. Optimal BOLD contrast is obtained when the TE matches the T2* (T2 in SE acquisitions) of the tissue of interest, which is determined by the magnetic field strength (e.g., TE ~ 25 ms at 7 T). Sub-millimeter acquisitions have typically been confined to partial head volumes (see [65], for methodological considerations on increasing the spatial resolution). For sub-millimeter isotropic acquisitions requiring large slice coverage, 3D GE-EPI can offer improved SNR over standard 2D GE-EPI acquisitions [78, 79]. At sub-millimeter resolutions, 3D-EPI-VASO has been shown to provide higher temporal stability and sensitivity to detect changes in CBV compared to 2D-EPI-VASO [80]. Although the temporal resolution is compromised to achieve sub-millimeter precision, recent advances such as shot-selective 2D CAIPIRINHA for 3D-EPI may provide the means to improve the temporal resolution of sub-millimeter fMRI acquisitions [81]. 3.2

3.2.1

Preprocessing

Motion Correction

Before performing experiment-specific analyses, it is necessary to pre-process the MRI data—with many of these steps being common across experiments. Examples of preprocessing include: motion correction, distortion correction, spatial alignment of the anatomical and functional datasets, surface reconstruction and mapping, spatial smoothing, and any basic manipulations of the fMRI time-courses (e.g., removing baseline periods, averaging or concatenating runs, filtering). Note that for whole-brain acquisitions with coarser spatial resolution (which are not in the thermalnoise-dominated regime), or for resting state applications, retrospective physiological noise correction may also be performed. Motion correction is performed on nearly all fMRI data, with the goal being to compensate for any participant head motion that occurred during each scan run. Motion correction is sometimes referred to as “volume registration,” since the procedure involves registering (or aligning) all the functional volumes (i.e., the functional data at each time point). This is typically done by using a rigid body transformation with six degrees-of-freedom (i.e., three rotations and three translations) applied on a volume-to-volume basis and is implemented in most of the leading fMRI analysis software packages (for a comparison of motion correction software, see [82]). This form of inter-volume motion correction (i.e., which compensates for motion occurring between time points) is, however, not able to contend with intra-volume motion (i.e., which occurs within the imaging time of a single volume). Although less commonly used, some forms of motion correction, such as prospective correction, aim to correct intra-volume inconsistencies [83, 84].

Human fMRI

409

3.2.2 Distortion Correction

Even with careful pre-scan optimization, such as shimming, distortions can still occur in the functional data. These arise from any remaining inhomogeneities in the main magnetic field. For instance, using EPI sequences, it is common to see distortion in the phase-encoding direction, which can lead to the mislocalization of functional activation as well as to difficulty aligning the functional with the anatomical data [85]. There are a number of ways to correct these distortions [86], with one of the most conventional being to collect a static map of the magnetic field, which can then be used to “unwarp” the data at all fMRI time-points. There are also more advanced, dynamic distortion correction techniques that can be used to compensate for changes to the main magnetic field that occur throughout a scan run, for example, those associated with motion and respiration [87–89]. Another option is to acquire an additional short functional run, but with the phase-encoding direction reversed. These images will also suffer local distortions, but the distortions will occur in the opposite direction, and can subsequently be used for distortion correction using a number of different algorithms [90–92]. Note that this method has been shown to work better when using SE-EPI reversed phase-encode “blip” images (rather than GE-EPI) to provide the estimate of the unwarping field [93].

3.2.3 Aligning Anatomical and Functional Data

Although some anatomical structure is apparent in functional datasets, the level of detail pales in comparison to what can be seen in structural images. Because of this, the functional and structural data are often aligned so that cortical activation can be more confidently localized to its anatomical origin [94, 95]. The two datasets can be brought into alignment using linear processes such as a 6-parameter rigid-body transformation (e.g., SPM’s COREG), similar to the motion correction described above, or a 12-parameter general affine transformation (e.g., AFNI’s 3dAllineate or FSL’s FLIRT). Alternatively, this alignment can be performed using algorithms that allow non-linear warping (e.g., AFNI’s 3dQwarp or FSL’s FNIRT), which might be useful in cases where some mismatch remains between the two datasets, for example, geometric distortions still present in the functional data. Note that FNIRT requires the functional data (usually T2*-weighted) and anatomical data (usually T1-weighted) to have the same contrast, hence, some experiments acquire an additional T1-weighted version of the functional protocol. Alternatively, a structural T2*-weighted scan can be acquired with the same slice prescription as the functional data. For a summary of the distortion correction and alignment procedures critical for further processing, see Fig. 3. Although the anatomicalfunctional alignment transform is typically computed separately to the motion correction transform, it has been shown that both these preprocessing steps can be performed simultaneously [96]. That is,

410

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Fig. 3 Correcting distortions and preparing for the projection of functional data into surface space. The conventional pipeline requires acquisition of additional scans (e.g., B0-map or EPI pairs with reversed phase encoding) for distortion correction. The distortion-corrected functional data is then co-registered with a wholehead T1-weighted anatomical volume, which may or may not be acquired in the same functional session. If not acquired within the functional session, an in-plane anatomical scan can be acquired with the same slice prescription as the functional data. This can then be used as an intermediate step in the co-registration of the functional data to the whole head T1-weighted anatomical, which is used to obtain cortical segmentations and surface models of the cortexCortices. For high spatial resolution functional data showing sufficient level of detail, it is also possible to obtain cortical segmentations in the EPI space by acquiring a distortion matched T1-weighted volume (or T1-map)

instead of registering each of the functional volumes and then aligning them all to the anatomical data with a single transform, it is possible to align each functional volume individually to the same anatomical and hence bring each functional volume into alignment at the same time. 3.2.4 Transforming Data to the Surface Domain

At this stage, a wide range of volumetric fMRI analyses can be performed (e.g., regions of interest can be defined, clusters of activation can be identified, and statistics can be performed). However, many analysis workflows go on to transform the data from the volume-domain to the surface-domain, permitting more powerful, anatomically-informed analyses, as well as more intuitive and interpretable visualization of the data [97]. This surface-based approach is particularly pertinent for examining the detailed topographic organization of somatosensory cortex. Without projecting the data onto the surface model, it is exceedingly difficult to understand how the topographic pattern is spatially organized with respect to the convoluted cortical geometry. To perform surface-based analyses, it is necessary to first construct surface models of the participant’s cortex using anatomical data (Fig. 3a, b) [98, 99]. For this, the anatomical data is segmented by classifying voxels as grey matter (GM), white matter (WM), or cerebral spinal fluid (CSF). This is typically performed

Human fMRI

411

using an automated approach (e.g., using FreeSurfer—see Table 2); however, manual intervention is often necessary to correct algorithmic misclassification. These tissue segmentations can then be used to construct models of the cortex, which are typically triangulated meshes positioned at the boundary between tissue classes. It is common to construct a surface model at the boundary between gray and white matter and another model at the pial surface between the grey matter and the CSF). Together, these two surfaces act to bind the cerebral cortex—the outer strip of cortical gray matter (Fig. 4a), which is the focus of most fMRI studies. In many workflows, especially those developed for data at conventional spatial resolutions (e.g., ~2 mm isotropic), a surface is then constructed at the midpoint between these two boundary surfaces, and the resulting “mid-grey” surface is used for further analyses (e.g., middle, red surface in Fig. 4b). The functional data can then be projected onto this surface using any number of interpolation methods (Fig. 4c). For example, it is possible to simply map the nearest neighbor voxel value onto each surface node, or alternatively, one can take the average of all voxel values that fall anywhere between the two boundary surfaces on a segment orthogonal to each mid-gray surface node. Note that although a mid-grey surface is commonly used, there may be circumstances when one is interested in using either of the boundary surfaces or some other intermediate surface. There are also studies that map data to multiple surface models. For instance, many layer fMRI studies sample data at multiple cortical depths using a set of parallel surfaces spanning the cerebral mantel (Fig. 4b) [49, 97, 100]. Once data are mapped to each desired surface, additional analyses can be performed on each surface (Fig. 4d). 3.2.5

Spatial Smoothing

Another common preprocessing step involves spatially smoothing the fMRI data. Smoothing increases the SNR and also ensures that the data more closely approximate a continuous field of random values, which is a necessary assumption for some statistical analyses [101]. However, it does come at the cost of reducing spatial specificity and can lead to an overestimation of the extent of functional activation [102] as well as the inappropriate mixing of signals [103]. As such, the degree of smoothing to be applied greatly depends on the research question and analysis approach. An important consideration here is the spatial scale of the activity of interest within somatosensory cortex—a range of organizational principles and different processing units at different scales can be examined using fMRI [104]. Decreasing in spatial scale, examples include cortical fields (e.g., the different Brodmann areas), individual fingertip representations [21, 105], and cortical layers or columns. Smoothing can be performed both volumetrically and in the surface domain, with surface-based smoothing yielding better spatial accuracy for localizing fMRI signals within the cortex [106]. One

412

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Fig. 4 Surface-based analyses. (a) First, high-resolution anatomical images are segmented by tissue type, and boundary surfaces are constructed at the pial

Human fMRI

413

of the main advantages of the surface-based approach is that by restricting the smoothing to data within the band of cortical gray matter (i.e., intracortically), it is possible to avoid noise contamination and signal dilution from the CSF and white matter, respectively [107]. Moreover, this intracortical approach is essential for smoothing when performing surface-based layer fMRI studies, as it avoids mixing signals across cortical depths (Fig. 5). 3.3 Somatosensory fMRI Paradigms and Analyses

This section is divided into two main parts: the first describes methods specifically related to imaging somatotopic maps in human S1, and the second outlines a range of other experiments that can be performed to further explore the function of somatosensory cortex using fMRI.

3.3.1 Imaging Somatotopic Maps

Somatotopic mapping experiments aim to understand the functional organization of somatosensory cortex by mapping the topographic relationship between regions of the body and their cortical representation. Many of the fMRI methods for imaging somatotopic maps were born from work performed first in visual cortex to examine retinotopic maps (i.e., the mapping between regions of the retina and their cortical representation) [108, 109]. One of the pioneering—and still frequently used—experimental paradigms for imaging retinotopic maps is the phase-encoded (or traveling wave) design [110–113]. The premise is to cyclically present stimulation across the visual field such that it progresses across some parameter that is continuously and contiguously represented across cortical space (e.g., visual field eccentricity or polar angle). In doing so, this paradigm will elicit traveling waves of activity across that cortical space, and the timing or phase can then be used to estimate the preferred eccentricity or polar angle value at each point in that space [114]. The response delay (or phase) at each voxel can be computed by simply cross-correlating its empirical time-course with a set of reference waveforms [115]; however, more computationally efficient methods have also been developed to estimate the delays using the Fourier transform—for example, from the response’s fundamental frequency alone [111] or by using all its spectral components [116]. ä Fig. 4 (continued) surface (i.e., the boundary between CSF and GM) and at the GM/WM boundary. (b) Next, intermediate surfaces are generated between the two boundary surfaces (here, nine intermediate surfaces have been generated). For studies at a more conventional resolution, a single mid-gray surface is typically constructed halfway between the two boundary surfaces. For higher resolution (i.e., layer fMRI) studies, many parallel surfaces are generated. (c) Once constructed, the functional data can be projected onto the surface models. Note that the blue ROI here denotes the hand area of S1. (d) Finally, additional analyses can be performed directly in the surface domain

414

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Fig. 5 Surface-based smoothing of somatotopic maps. Left: To avoid contaminating functional data with measurements outside of gray matter (e.g., from the CSF or white matter) as well as to avoid mixing data across cortical depths, smoothing can be applied tangential to each surface. Right: Somatotopic maps with varying degrees of surface-based smoothness. Note that the unsmoothed data was collected at a resolution of 0.8 mm (isotropic). As can be seen, smoothing greatly reduces the noisy appearance of the maps; however, legitimate somatotopic detail can also be lost (e.g., smoothing to a full-width-at-half-maximum of 6.0 mm at the pial surface results in the loss of the little finger representation for this participant)

Although developed for mapping eccentricity and polar angle values in visual cortex, this phase-encoded technique has proven useful for mapping other continuously varying parameters in the brain. For example, phase-encoded somatotopic stimulation have revealed the presence of multiple topographically organized body part representations in primary somatosensory cortex, such as the arm [117], the lips and face [10], and the digits [8, 22]. Note that phase-encoded designs can also be applied to study top-down processes like attention [118], and, pertinently, it has been shown that somatotopic maps constructed by sweeping attention across the fingertips in a phase-encoded fashion are nearly indistinguishable from those constructed using sensory stimulation [29]. Although the rationale and basic analysis for phase-encoded somatotopic mapping are fundamentally the same as that for its retinotopic counterpart, there is one considerable difference: the somatosensory sensory space is not continuously represented in the cortex. In primary visual cortex (V1), the entire eccentricity range is continuously mapped from the fovea to the periphery along the posterior-to-anterior axis of the cortex, and the entire polar angle range is continuously mapped orthogonal to the eccentricity bands—with the only discontinuity being between the two hemifields of visual space, as they are represented in opposite hemispheres of the brain. In primary somatosensory cortex, the two halves of the body are likewise represented in contralateral hemispheres. However, unlike in V1, there are discontinuities between

Human fMRI

415

some body parts within each half. For example, the face is represented adjacent to the fingers, away from the head and neck. Practically, this prevents the entire somatosensory space from being easily mapped with a single, continuously applied trail of stimulation. Another difficulty in the somatosensory domain arises from the nature of most somatosensory stimulation. With the exception of manual stroking or brushing, nearly all methods are only capable of stimulating a few discrete sites on the body surface. Moreover, most applied stimulation (e.g., vibration) results in considerable propagations of physical traveling waves on the skin— limiting the resolution of the applied stimulation. Somatotopic mapping data can be acquired using a number of experimental approaches other than the phase-encoded paradigm, including block- and event-related designs [119, 120]. For these, data are typically analyzed using a generalized linear model (GLM) in which the stimulation at each somatotopic location is treated as a separate experimental condition. In this way, an estimate of each voxel’s response to each stimulated location is obtained. This is different, and more informative, than the results yielded by a conventional delay analysis—which only estimates each voxel’s preferred somatotopic location. This distinction is important, as voxels are not responsive to stimulation of only a single point on the body but of a region. For example, although a voxel response may be greatest for stimulation at a particular fingertip, it may still show significant activation in response to stimulation of the neighboring digits. Thus, one of the advantages of using block- or eventrelated designs, along with a GLM-based analysis, is that they allow for this overlap to be estimated [121]. Note, however, that this comes at a cost. In a comparison study between phase-encoded and event-related designs, the phase-encoded approach was demonstrated to be a factor of 3–4 times more efficient for somatotopic mapping [122]. Furthermore, two different Bayesian analysis techniques have recently been developed, both of which are capable of estimating the degree of somatotopic overlap from phase-encoded data (iMGP analysis [123]; pRF analysis [124]). This illustrates an important distinction: the limitation of being able to extract only the preferred somatotopic location via a conventional delay analysis reflects a drawback in the analysis approach rather than the experimental paradigm per se. See Fig. 6 for a summary of common somatotopic mapping paradigms and analysis approaches. Regardless of the specific experimental paradigm and analysis approach, it is usually necessary to threshold somatotopic mapping data to identify significant responses. There are, however, no hard and fast rules for determining what threshold to use. For example, at conventional resolutions, it has been commonplace for researchers to simply threshold phase-encoded topographic maps based on the degree of correlation between the empirical time-course and a

416

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Fig. 6 Somatotopic mapping paradigms and analyses. Top: Common experimental paradigms for mapping somatotopic organization using fMRI. Bottom: Various analysis methods that have successfully been applied to data acquired via the above paradigms

reference waveform (e.g., typically at a correlation coefficient around 0.3). At sub-millimeter resolutions, however, application of this convention could lead to unintended consequences, as the correlation coefficient varies as a function of cortical depth. As such, the choice should be as principled as possible and consistently applied across participants. Another, more general approach is to threshold based on statistical evidence of significance (e.g., p-values associated with the delay or GLM analysis). When appropriate, the statistics should be corrected for multiple comparisons given the number of voxels and hence statistical tests being performed. This is commonly achieved through false discovery rate (FDR) [125] or family-wise error rate (FWER) [126] corrections. These approaches alone, however, ignore the known spatial coherence of brain activations, treating each voxel as if it is independent of its neighbors. A promising, and more recently developed, approach to multiple comparison correction based on Local Indicators of Spatial Association (LISA) incorporates spatial context via a non-linear filter and has been shown to produce more spatially specific somatotopic maps compared to more conventional approaches [127]. 3.3.2 Imaging Somatosensory Function

Beyond mapping somatotopic organization, fMRI can be used to image cortical function associated with a range of other stimulus characteristics such as texture, shape, and hardness [128, 129] or intensity [130]. Given that vibrotactile stimuli can elicit two qualitatively distinct cutaneous sensations in a frequency-specific manner—flutter (5–60 Hz) and vibration (>60 Hz) [131]— much research has investigated how somatosensory cortical activity changes as a function of vibrotactile stimulation frequency [132, 133]. Other phenomena include tactile adaptation, whereby activation decreases over time during sustained or repeated stimuli [134], tactile discrimination (such as frequency discrimination of vibrotactile stimuli [135]), and somatosensory activation during

Human fMRI

417

the observation of touch [136]. fMRI has also been used to study cognitive processes such as the modulation effects of attention to tactile stimulation [29], mental imagery of tactile sensations [137, 138], tactile learning [139], and prediction of touch [30]. Some of these studies also collect a somatotopic mapping dataset or a quick functional block-design localizer and use these data to define ROIs for further analysis, such as to examine activity within these ROIs associated with the independently mapped body part (s). If it is not possible to collect this additional dataset (e.g., because of time constraints), there are a number of other ways to select ROIs, such as cytoarchitectonic probabilistic atlases [137], anatomical landmarks based on cytoarchitectonic studies [21], or a probabilistic atlas of functional domains [4]. When delineating ROIs, a surface view is advantageous, as these can be drawn on a single view rather than having to inspect multiple slices in a volume view. Note that surface ROIs can easily be transformed to volume space and vice versa. The different somatosensory cutaneous modalities (touch, temperature, itch, and pain) are encoded by specialized sensory receptor types that innervate the skin. Different forms of stimulation are used to engage different sub-modalities of somatosensory afferents, which can target distinct cortical areas. A variety of skin stimulation methods, including vibrotactile, pneumatic, stroking, human touch, and electrical stimulation of the skin and median nerve, can be used to engage low-threshold mechanoreceptors (Aβ) responsible for discriminative touch. This kind of stimulation can elicit robust activations within areas 3b and 1 as well as, to a lesser extent, area 2 (more robust activations within area 2 can be evoked using a salient stimulus, such as human touch [21, 140]. Piezoelectric devices are popular as they can generate a wider range of frequencies than pneumatic devices (1–20 Hz), and hence can be used to study both flutter and vibration. Vibration primarily stimulates fast-adapting afferents innervating Meissner’s corpuscles (FA1) and Pacinian receptors (FA2). FA1 afferents are preferentially activated at frequencies up to 60 Hz, whereas FA2 afferents are most sensitive to frequencies >100 Hz, hence, the choice of stimulation frequency can preferentially stimulate one kind of receptor over the others. Slowly adapting afferents can be preferentially stimulated using pressure [134]. Note, however, that any simulation of the skin surface engages multiple receptors of different kinds, and the only way to select one individual type is by using intraneural microstimulation. Stroking of the non-glabrous skin can also engage low-threshold unmyelinated mechanoreceptors (C-tactile) afferents in addition to the thickly myelinated Aβ afferents. These CT afferents are tuned to respond to velocities of ~5 cm/s, which are perceived as the most pleasant and hence play an important role in the affective touch system [141] (Chapters 6, 10, and 15, this volume).

418

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

Noxious thermal heat can be delivered using contact thermode or laser pulses (Chapters 7 and 9, this volume). Brief stimuli can produce clear pinprick painful sensations related to Aδ afferents, which innervate areas 3b and 1 of the S1. This is an early, sharp, and well-localized percept designated as “fast or discriminative pain.” Notably, Mancini et al., (2012)], showed that nociceptive (Aδ) digit somatotopic maps (generated with laser stimulation) were highly aligned with non-painful tactile (Aβ) maps. It has also been shown that thermonoxious stimulation can evoke a separate neural response, mediated by C-afferent drive and associated with the second or burning pain, in the depth of the central sulcus—in Brodmann’s area 3a [142]. To evoke such a response, sufficiently long (multi-second) stimulation that builds up the slow, second, or burning pain percept is required.

4

Notes Although fMRI can be applied to study somatosensory cortex in a wide variety of ways, there are some issues and considerations common to most experiments. The most important of these are outlined below, and practical advice is given where possible.

4.1 How Many Participants Are Needed?

This is highly dependent on the research question and aims. Since it is possible to construct detailed somatotopic maps at the individual level, interesting and impactful work can be performed even with a very low number of participants. Averaging across a large group can actually obfuscate meaningful details in this type of work. By contrast, other research aims necessitate large sample sizes. For example, work investigating individual differences in somatosensory function or work aiming to construct a probabilistic atlas of human somatotopic organization would, by their nature, benefit from large groups of participants.

4.2 What MR Sequences and Parameters Should be Used?

As mentioned in Subheading 1: Introduction, imaging sequences vary significantly, and careful consideration should be given to the sequence type and specific parameters before data collection begins [143]. First, it is important to identify any requirements you have based on your research question, for example, whether a specific spatial or temporal resolution is needed. It is then best to consult with a local MR physicist or radiographer to determine if it is possible to meet your needs given your specific scanner setup. Not only do capabilities vary across scanner types, but they also vary across the same scanner at different institutions—reflecting local expertise and research priorities. Those working on developing and implementing the various sequences will likely have the best understanding of your local MR scanner’s capabilities.

Human fMRI

419

4.3 Any Additional Advice on Which Software to Use?

There are situations when a specific package might be required, but overall there is a great deal of overlapping capability in the software packages outlined in Table 2. Below is a brief account of the software the authors typically use in their somatosensory fMRI experiments. Author AMP mostly uses Matlab to present visual and auditory cues, collect button response data, and interact with piezoelectric vibrotactile stimulators. This approach relies heavily on Psychtoolbox. Most analyses are performed using FreeSurfer and AFNI/ SUMA. FreeSurfer is used to segment the anatomical images and subsequently generate cortical surface models. AFNI is used for preprocessing and analyzing the volumetric functional data. The volumetric functional data is then mapped onto the surface models from FreeSurfer and further analyzed and visualized using SUMA. Author RSP uses Matlab (MGL) or Python to drive the piezoelectric stimulators and to present visual cues, and mrTools (Matlab) for analysis. Cortical segmentations are generated with FreeSurfer, and output surfaces are imported to mrTools. Once a participant’s functional data set has been aligned to the whole head T1-weighted anatomical (using mrAlign), it is trivial to swap between EPI (functional) view, whole head volume anatomical, and surface views, or a selected patch of the cortical surface. It is not at all uncommon to use a combination of multiple software packages during the course of an fMRI study. These workflows have traditionally been customized for each study by each research team, leading to undesirable variability. However, there is an increasing effort to develop more consistent workflows that can be shared across researchers in the field. For example, fMRIPrep is an analysis-agnostic tool designed to automatically adapt best practices for the preprocessing of a diverse range of fMRI data, which can help ensure consistent results [144].

4.4 What Makes a Good Participant?

Three aspects are important here: expertise, comfort, and alertness.

4.4.1

The best participants are often those that have had previous experience as a research participant in a MR scanner. To most people, the scanner environment is foreign at first and not terribly inviting. It is dark, noisy, cramped, and often cold. This can lead to restlessness and hence, participant movement or distraction. Any movement, particularly that of the head, can negatively affect data quality. Fortunately, a great deal of this can be corrected retroactively using the modern post-processing algorithms described in Subheading 3: Methods; however, it is undoubtedly best to avoid this when possible—especially for high-resolution work such as laminar or columnar imaging. With experience, most participants grow more tolerant of the scanner environment and hence can remain

Expertise

420

Alexander M. Puckett and Rosa M. Sanchez Panchuelo

more still throughout the session. This is particularly important if the experimental task is cognitively demanding (e.g., an attention task), as participant restlessness can also lead to distraction or an inability to perform. 4.4.2

Comfort

Besides expertise, there are a number of factors that influence participant comfort, and as a researcher, there are a few actions that can be taken to maximize this for the participant. 1. Familiarize the participant with the experimental stimuli and task before the scan session, going so far as to practice the experiment in a mock scanner set-up if available. This can help reduce participant uncertainty or confusion during the actual scan. 2. Do not rush when setting up the experiment at the scanner. Schedules are often tight at scanner facilities, and scan costs are often high, leading to a natural tendency to maximize the amount of time spent actively collecting data within a scheduled scan slot. Be aware that a few extra minutes spent on participant comfort may mean the difference between a participant completing a scan or not. 3. Attempt to address any participant discomfort, even if only slight, before the actual scanning begins. This is particularly important for long scans, as a slight discomfort can become a troubling pain by the end of a 90 min scan session. Common discomforts are related to poor placement of hearing protection or headphones, too much pressure associated with head stabilization, and insufficient padding under the back of the head or elbows. A leg rest is also important to avoid back discomfort, especially during long scan sessions.

4.4.3

Alertness

Lying down in a supine position in a dark, confined space is not the most conducive environment for high alertness. However, many experiments require the participant to be vigilant and engaged—or at the very least, to be awake. There are a few methods that can be used to monitor alertness: eye-tracking can be used to detect if the participant’s eyes close for an extended period of time; the experiment can be designed such that the participant must give behavioral responses throughout; and subjective alertness levels can be checked verbally after each scan run. Note that there can be a trade-off between maximizing comfort and alertness. That is, if a participant is “too comfortable,” it may be easier for them to lose focus or fall asleep. This is the reason why both authors of this chapter usually refuse a blanket for themselves during scans. Yes, the room is a bit colder than preferred, however, a blanket is simply too comfortable!

Human fMRI

421

4.5 Are There Any Other Ways to Deal with Distortions?

As mentioned in Subheading 3.2.2, distortions in the functional data can not only lead to mislocalization of activity, but they can also lead to difficulties in aligning the functional data with the anatomical data (since the anatomical images do not typically suffer from the same distortions as the functional ones). Rather than correcting the distortion and then aligning with the anatomical, some researchers opt to collect and then work with a distortionmatched anatomical instead [145]. To avoid the difficulty in aligning a distorted functional to an anatomical scan, another option is to forego the anatomical image entirely. At conventional resolutions, this is ill-advised, as the functional datasets lack sufficient anatomical detail. However, with high-resolution measurements, there is sufficient detail to identify various sulci and gyri directly in the functional images, which may be an adequate level of structure for some studies.

4.6 How Long Should the Scan Session Last?

Whenever possible, it is recommended to limit scans to around 60 min, particularly for less experienced participants. This is mainly so the participant can maintain comfort and alertness. Relatedly, it is advantageous to test all experimental conditions in a single scan session to avoid between-session differences in the fMRI measures [146, 147]. However, this approach can lead to long scan times depending on the number of conditions. More experienced research participants may be useful here, as they can often endure longer sessions. Otherwise, if it is necessary to collect the data across multiple sessions, it is important that the researcher understands and accounts for intersession variability as appropriate. In planning the scan session, one should allocate a few minutes for the magnetic shimming (1000 ms). This has made the process of finding specific cortical responses difficult. It is not yet clear what the source of this activity is, but it is hypothesized that the activity is from evaluative processes, determining the affective valuation of the stimuli. Despite the latency of these evoked responses, Ackerley and colleagues [39] showed not only an evoked response late into the ERP epoch, but also that this response changed the time course of its peak based on minor changes (±1 cm/s) in the velocity of the stroking touch. This reveals the link between the ULP and this specific type of touch input, and further suggests that this response is linked to the offset and subsequent evaluation of the stimuli (Fig. 3).

Fig. 3 The time course of an ULP is atypical, considering past research studies. The black line is the ERP response and the blue line is the force measured by the RTS (stroking robot). The afferents transmitting responses to affect touch are slow conducting, and so do not reach the cortex until a time when most sensory stimuli typically have already been processed, particularly the earliest stages of primary sensory processing. (a) shows the ULP measured by Ackerley and colleagues [39] and (b) that measured by Haggarty and colleagues [40]. (Reproduced with permission from Elsevier and Wiley)

Somatosensory EEG

1.4 Emotions and SEPs

437

Current models of emotion understanding propose that initial visual processing is followed by activation of sensorimotor and somatosensory areas that are key to the experience of the emotion. Among these areas, the motor and sensory cortices, including the somatosensory cortices S1 and S2, are critical areas for action representation, and they are highly interconnected to the limbic system [43–45]. Ample evidence supports the contribution of the somatosensory cortex to emotion understanding. A novel body of work has allowed for investigation of the selective involvement of the right S1 and S2 cortices in visual emotional processing [46, 47]. The core principle of this approach relies on the ERP subtraction method to isolate somatosensory responses from visual processing. Specifically, this method comprises the presentation of an emotional facial expression in two experimental conditions. In one condition, the visual emotional expression is presented alone, and EEG activity is recorded from the visual cortex with scalp electrodes (visual-only condition); in another, the visual emotional expression is followed shortly after by tactile stimulation on a body part such as the right finger, while EEG activity is recorded from both somatosensory and visual regions (visual-tactile condition, Fig. 4). This experimental setup allows the researcher to subsequently isolate the responses of S1 and S2 over and above the effects induced by other processing regions. It is therefore possible to subtract purely visually-evoked potentials (VEPs; visual-only condition) from tactually-evoked SEPs (visual-tactile condition) during facial processing, obtaining “VEP-free SEPs.”

Fig. 4 Typical experimental setup comprising two visual-tactile conditions, visual-tactile face condition (VTFAC), visual-tactile finger condition (VTFIC), and visual only condition (VOC) in an emotion task. In VTFAC and VTFIC, tactile probes were delivered 105 ms after the face onset to the face and the finger, respectively. Participants were instructed to observe the emotions, and in 20% of trials, they were asked to indicate the emotional content of the stimulus after presentation of the face. (Reproduced from Sel, Forster, and CalvoMerino, 2014 under Creative Commons license)

438

Jonas Vibell et al.

The efficacy of the ERP subtraction method to isolate somatosensory from visual responses has not only been proven effective in emotion understanding but also in other domains such as visual perception and memory of bodies and actions [48–51] (for an extensive review of the ERP subtraction method, see [50]). The ERP subtraction method can be readily applied to study the involvement of somatosensory cortices in emotional processing in other domains, such as music or emotional sounds, as well as for body-related information in both healthy and clinical populations. 1.5 Self-Other Processing and SEPs

Human touch in the real world is frequently a multisensory experience, which is felt more than merely through the receptors in our skin. Somatosensation, like other bodily senses, thus provides one critical basis for our ability to distinguish experiences that pertain to the self from other sensory impressions in the environment (for reviews, see [52–55]). This section showcases the use of SEPs in somatosensory resonance paradigms (also known as visual remapping of touch, VRT, or mirror touch paradigms) to investigate the neural basis of embodiment of one’s self and its distinction from the embodiment of others within the wider somatosensory system. Like the emotional paradigms mentioned in the previous section, somatosensory resonance paradigms exploit the propensity of the somatosensory system to activate vicariously. Critically, such activations reflect the self-relatedness of the observed stimuli at different ERP components, which researchers may exploit to investigate the bodily self and its socio-cognitive basis across the lifespan, as well as in clinical disorders. In most variants of this paradigm, observers receive tactile stimuli on their own body, which is hidden from view, while viewing a body being touched or not touched at the same time on a computer screen. Touches are, typically, a brief tap or stroke with a finger, pencil, cotton bud, or brush, shown via video or a series of still images. In behavioral studies, participants are then asked to report on the presence, location, or intensity of the touch on their own body. In ERP versions of the paradigm, SEPs are obtained from tactile stimuli in each trial. SEP studies have shown that resonance effects are stronger and/or occur earlier in cortical processing for more self-related stimuli, such as for touch on human hands compared to rubber objects [56], for touch on one’s own face compared to that of another person [57], and for touch on hands shown in first-person compared to third-person perspective [58] (for more on illusions, see Chapter 13, this volume). Furthermore, clinical conditions marked by detachment from one’s bodily self (e.g., depersonalization or derealization disorder) are also marked by alterations in somatosensory resonance for tactile events on one’s own body [57, 59] (Fig. 5).

Somatosensory EEG

439

Fig. 5 Example trial design for embodiment and self-other processing studies. A trial starts and ends with the visual presentation of a body and a tactile stimulation device (in this example, an image of the self-face and a pencil). This is replaced by a brief presentation of a visual touch (centre top left) or, in separate trials, of a visual no touch (center bottom left) together with an actual tactile stimulus felt on the observer’s own body (center top and bottom right). As embodiment entails the internal simulation of observed bodily events, SEPs in touch-viewed trials should be enhanced relative to no-touch-viewed trials (somatosensory resonance)

2

Materials

2.1 Tactile Stimulators

Most of the devices used to produce mechanical tactile stimulation for behavioral studies can be applied in the EEG environment, though solenoid, electrical, or piezoelectric stimulators are most common (see Chapters 1, 2, 3, 4, and 5, this volume). Solenoids can have a mechanical delay, and electrical buzzers have a minimum duration, which need to be taken into consideration, if relevant for the paradigm (see Note 4.1). EEG recordings will quickly show if any of the stimulators produce artifacts, which then need to be addressed. For affective touch, a rotary tactile stimulator (RTS, Chapter 6, this volume) is commonly used, which will send a signal once it starts to move. Force records can be used to deduce the moment the skin is touched, to time lock the onset of the stimulus. This is more difficult with manual stroking, however, so the equipment needs to be capable of delivering accurate recordings at the beginning of a stimulus. For example, in Haggarty and colleagues’ apparatus [40], the breaking of a laser beam over the participant’s arm signals the beginning of the stroking protocol (Fig. 6a).

440

Jonas Vibell et al.

Fig. 6 Time-sensitive measures of touch stimulation. (a) Haggarty and colleagues [40] used the breaking of a laser beam over the arm to signal when the stroking has begun, this signal is sent directly to the EEG acquisition computer as a unique stimulus trigger, to distinguish time 0 from the trigger sent by the onset of trials on the experimental computer. (b) Hauser and colleagues [60] used a series of motion sensors to detect the onset and relative position of the stroking touch throughout. NB: Only Haggarty and colleagues [40] used EEG, but both methods are appropriate for time locking manual stroking to the onset of the stimulus

Furthermore, Hauser and colleagues [60] used a series of sensory and infrared measurements (Fig. 6b) to determine the location of the stroking hand in relation to the stroked surface. 2.2 EEG Acquisition Systems

3

Commercial systems that record EEG come from a number of companies, including BrainSystems, GTec, Biosemi, Compumedics Neuroscan, Brain Products, and Magstim EGI. EEG is very sensitive to timing, so it is important that the stimulators are reliable and instant. All of these products can have accurate timing, but it is important to verify your set-up. The systems have different set-up times, with gel-free systems being the fastest. The EEG commonly relies on a trigger to indicate when a stimulus was presented, and it is important that this trigger occurs simultaneously (or reliably) with the tactile stimulus. Most modern presentation software (e.g., Presentation, MatLab, ePrime, PsychoPy) have accurate timing [61], though this should always be evaluated for each system in situ, especially for multisensory presentations (see Note 4.2).

Methods Methods for recording SEPs are mostly similar to those for recording VEPs and auditory evoked potentials (AEPs) and have been detailed elsewhere [62]. Here we focus on those aspects of recording and analysis that are specific to SEPs.

Somatosensory EEG

3.1

Timing of SEPs

441

The timings of the characteristic potentials are approximate and vary slightly depending on the experimental paradigm, although they remain quite stable between individuals. In addition, the density and type of mechanoreceptor found in a specific skin site can influence the responses found and their timing (Fig. 2). Signal strength and latency can also vary based on factors such as the age, gender, height, skin temperature, sleepiness, drugs or medication, and attentional state of the participant [63]. It is worth noting that, especially with mechanical stimulation, the delay between the electrical signal for the mechanical parts to move and the activation of the touch receptors can be several milliseconds, which needs to be considered in study design and also the placing of triggers or markers in the continuous EEG data record. Information on mechanical delay, force, and other stimulus parameters is normally provided by the device manufacturer. Skin contact from mechanical tappers (solenoids) relative to other stimuli and triggers may also be measured accurately via acoustic sensors (e.g., Cedrus StimTracker) with tappers attached to a resonant surface (e.g., table top) or a force sensor in line with the actuator. When comparing multisensory interactions, it is important to keep in mind the different processing times of the constituent stimuli. Vibell and colleagues [14] found that touch had to lead vision by 38 ms for them to be perceived as simultaneous, but that this could increase to 94 ms depending on how attention was directed. These findings are consistent with those of behavioral studies of the visuotactile and audiovisual prior-entry effects [27, 64]. Therefore, it is important to calibrate stimulus presentation times between the senses and to know if the study requires objective or subjective (perceived) simultaneity.

3.2 Vision of the Body and SEPs

Non-informative vision of the body can increase tactile acuity [65] and enhance early SEPs [66]. Similarly, attentional modulations of SEPs are usually enhanced when viewing the body parts [67, 68], except when selecting between different fingers of the same hand [69, 70]. At least on a behavioral level, such effects of vision on tactile perception have been found even when location was taskirrelevant [71], suggesting that simply viewing the body profoundly changes tactile perception, which is also reflected in enhancements of SEPs.

3.3 Multisensory Temporal and Spatial Proximity

The response times to different sensory stimuli can differ significantly within and between participants, and a decision will have to be made on how best to equate these, particularly since non-simultaneous stimuli can have a distracting effect. Some examples of setting the presentation timing of stimuli from different sensory modalities include no adjustment, individual adjustment, and staircase procedures [72]. Spatial attention effects between modalities depend on the spatial coincidence of stimuli, and if

442

Jonas Vibell et al.

stimuli are moved further apart, it can reduce attentional distraction [73]. In addition, the somatotopic reference frame for touch allows researchers to create conflicting stimulus lateralizations when the arms are crossed. This means that a stimulus presented on the left side and viewed in left visual space can be presented to the right arm. The contralateral projection to the brain results in the multisensory stimulus activating the right hemisphere for the visual (and auditory) component, but the left hemisphere for the somatosensory component. 3.4 Body Posture, Spatial Congruency, and SEPs

Body posture can have a profound effect on tactile processing, as shown in delayed and less accurate behavioral responses when the hands are crossed compared to uncrossed [74, 75]. This has been attributed to a rapid recoding of tactile stimuli from a somatotopic to an external reference frame [76]. The influence of body posture has also been documented on SEPs. In particular, attentional modulations are stronger when the hands are close together [77, 78], and even within-hand posture effects have been reported [69]. Therefore, the relative location of tactile stimuli to each other in both somatotopic and external space need to be considered in experimental designs [75]. The spatial congruency between the body parts that are seen and felt to be touched may also play a role in somatosensory resonance paradigms.

3.5 Temporal Window of Emotion SEPs

Hierarchical models of visual processing propose that emotional face processing and other types of face processing require a series of interactions between brain areas, starting from the visual cortex (occipital face area, fusiform face area, and superior temporal sulcus), that feed-forward to central and frontal regions [79, 80]. In this vein, Pitcher [81] demonstrated that facial emotion recognition comprises a hierarchical cascade of activations starting in the visual cortex from about 60–100 ms after face onset, followed by activation in the somatosensory cortex between 100 and 170 ms after face onset. Therefore, it is crucial that we probe the state of S1 and S2 with tactile stimulation at the time that somatosensory cortices are maximally involved—that is, between 100 and 170 ms after the onset of the emotional visual stimulus. Given that tactile information transduced by sensory fibers in the periphery takes around 20 ms to reach S1, an ideal tactile stimulation onset may lie between 100 and 140 ms after the emotional visual stimulus onset.

3.6 Emotion and Self-Other SEPs

The majority of studies demonstrate that the somatosensory involvement in emotion processing is lateralized to the right hemisphere [47, 81, 82]. Similarly, self-related processing is more strongly lateralized to the right hemisphere across wider cortical networks [83]. Therefore, locations on the left side of the body (e.g., the left finger) are likely to be good candidates to study somatosensory processing related to visual emotions and self-

Somatosensory EEG

443

other processes, especially if the duration of an experiment necessitates a choice between stimulation or left or right sides. In addition, a recent study [84] has demonstrated that right S1 and S2 do not exhibit a general excitatory response to emotional stimuli, but they do contribute in a discrete somatotopic fashion to specific emotions. As an example, activations in the finger S1 and S2 recorded during the processing of angry faces show a significantly different pattern of response than the finger S1 and S2 activity recorded in response to sad face processing. Relatedly, the experiential match between seen and felt touch may amplify resonant responses in S1, which also encodes the sensory qualities of touch [84]. Rigato and colleagues [58] recently argued that vicarious enhancement of the P45 component may be opposite in direction, depending on whether the seen and felt sensory qualities were more or less similar in different SEP studies (e.g., soft touch or brushes, pointed taps, vibrations). Researchers may therefore wish to maximize, or purposefully manipulate, visual and tactile matches. 3.7 Vicarious Touch Parameters

There are two further critical aspects of somatosensory resonance paradigms. One is that the viewed no-touch stimulus should contain similar visual information as the touch stimulus (e.g., similar motion paths but without the final contact) in order to avoid SEP enhancements being confounded by stronger activation of the visual system or by greater attention in more visually exciting viewed touch trials. The other is to ensure that participants pay attention to the visual stimulation (specifically, the touch event), which may be necessary to obtain an optimal resonant response from the primary somatosensory system. An fMRI study by Chan & Baker [85], where participants responded to infrequent visual events unrelated to observed touch, found that vicarious activations were restricted to posterior parietal cortex and absent from S1 and S2. To increase the relevance of seen touch events in SEP studies, participants could silently count and eventually report infrequently seen “double taps” [57].

3.8 Affective Touch Parameters

It is important to consider previous psychophysical work when designing an affective touch study (see Chapter 6, this volume), but it is especially important to take into account the “gold-standard” microneurography studies (see Chapter 15, this volume), which show the kind of stimuli that these nerve afferents respond to optimally. Studies may use “affective touch” as a descriptor for their stimuli, but this is not always the case. Two important points to consider are first that stroking touch delivered at 3 cm/s is optimal for CTs, and that this is not greatly affected by the manner of stroking (i.e., RTS, brush, hand, or glove), so this should be considered the basis of affective touch probing. However, if you stroke a rough surface over the skin, it is not at all pleasant, but CTs would still respond similarly. This is the difference between (positive) affective touch and CTs. CTs are just encoding the touch and

444

Jonas Vibell et al.

the Aβ afferents would be very much involved in the conscious pleasantness perception too (even if it is not directly encoded in their firing). Second, the onset of the touch should be controlled to ensure that the epoch is precisely time locked, which requires the use of creative methods to ensure the beginning of the stroking is accurate. A- and C-type fibers have very different conduction velocities, so these need to be considered during the planning of the study and extraction of epochs (see Note 4.3).

4

Notes

4.1 Tactile Interference in EEG Recording

The driving pulse for the tactile stimulators from the computer is commonly a square wave, consisting of all frequencies, which may introduce electrical noise in the EEG recording. The electrical discharge from the tactile stimulators can sometimes be detected in the EEG trace and should be eliminated through, for example, grounding of the device. A sinusoidal pulse with a gradual incline and decline usually stimulates equally well but without potential noise. Any stimulator-related noise usually happens before SEPs and can be removed during offline processing if needed. (For further advice on noise and artifacts, see Chapters 15, 16, 17, 18, and 20, this volume.)

4.2 Acoustic Interference

Tactile stimulators typically make a noise that is perceivable by participants and that can confound the processing of tactile stimulation. To mask sounds made by the tactile stimulators, it is recommended to play white noise (~65 dB, measured from the participants’ head) with loudspeakers at a distance of about 90 cm from the participants’ head. Headphones (e.g., in-ear) may also be used to play white noise, but care must be taken to avoid interference with EEG recording. If earlobe (reference) electrodes are used with in-ear headphones, it can help to place these at the back rather than the front of the earlobe.

4.3 Peripheral Conduction Velocity

When comparing the cortical responses to affective touch, the conduction velocity of CTs means that the signal does not reach the cortex as quickly as other sensory modalities. For example, Aβ touch afferents have a conduction velocity of around 50 m/s, while CTs conduct at a velocity of less than 1 m/s, therefore reaching the cortex much later. In Ackerley and colleagues [39] and Haggarty and colleagues’ work [40], this was overcome by being aware that a 700 ms delay was added to the epochs to account for the approximate distance from the stroked surface to the cortex and the conduction velocity of these afferents. Furthermore, this distinction between conduction velocities has been considered in facial electromyography research, allowing for conduction velocity to account for delays in affective arousal [86].

Somatosensory EEG

445

References 1. Caton R (1877) Interim report on investigation of the electric currents of the brain. Br Med J 3(Suppl 62) 2. Eason RG, Harter R, White CT (1969) Effects of attention and arousal on visually evoked cortical potentials and reaction time in man. Physiol Behav 4:283–289 3. Hillyard SA, Munte TF (1984) Selective attention to color and location: an analysis with event-related brain potentials. Percept Psychophys 36(2):185–198 4. Nuwer MR (1998) Fundamentals of evoked potentials and common clinical applications today. Electroencephalogr Clin Neurophysiol 106(2):142–148. https://doi.org/10.1016/ s0013-4694(97)00117-x 5. Keil A, Debener S, Gratton G, Jungho¨fer M, Kappenman ES, Luck SJ, Luu P, Miller GA, Yee CM (2014) Committee report: publication guidelines and recommendations for studies using electroencephalography and magnetoencephalography. Psychophysiology 51:1–21. https://doi.org/10.1111/psyp.12147 6. Paul M, Govaart GH, Schettino A (2021) Making ERP research more transparent: guidelines for preregistration. Int J Psychophysiol 164:52–63. https://doi.org/10.1016/j. ijpsycho.2021.02.016 7. Iwamura Y (1998) Hierarchical somatosensory processing. Curr Opin Neurobiol 8(4): 522–528. https://doi.org/10.1016/s09594388(98)80041-x 8. Allison T, McCarthy G, Wood CC, Jones SJ (1991) Potentials evoked in human and monkey cerebral cortex by stimulation of the median nerve: a review of scalp and intracranial recordings. Brain 114:2465–2503. https:// doi.org/10.1093/brain/114.6.2465 9. Schubert R, Ritter P, Wu¨stenberg T, Preuschhof C, Curio G, Sommer W, Villringer A (2008) Spatial attention related SEP amplitude modulations covary with BOLD signal in S1--a simultaneous EEG--fMRI study. Cereb Cortex 18(11):2686–2700. https://doi.org/ 10.1093/cercor/bhn029 10. Allison T, McCarthy G, Wood CC (1992) The relationship between human longlatency somatosensory evoked potentials recorded from the cortical surface and from the scalp. Electroencephalogr Clin Neurophysiol 84: 301–314. https://doi.org/10.1016/01685597(92)90082-m 11. Hari R, Reinikainen K, Kaukoranta E, H€am€al€ainen M, Ilmoniemi R, Penttinen A et al (1984) Somatosensory evoked cerebral

magnetic fields from SI and SII in man. Electroencephalogr Clin Neurophysiol 57(3): 254–263. https://doi.org/10.1016/00134694(84)90126-3 12. Mima T, Nagamine T, Nakamura K, Shibasaki H (1998) Attention modulates both primary and second somatosensory cortical activities in humans: a magnetoencephalographic study. J Neurophysiol 80(4):2215–2221. https://doi. org/10.1152/jn.1998.80.4.2215 13. Zhu Z, Disbrow EA, Zumer JM, McGonigle DJ, Nagarajan SS (2007) Spatiotemporal integration of tactile information in human somatosensory cortex. BMC Neurosci 8(1): 21. https://doi.org/10.1186/1471-22028-21 14. Vibell J, Klinge C, Zampini M, Spence C, Nobre AC (2007) Temporal order is coded temporally in the brain: early ERP latency shifts underlying prior entry in a crossmodal temporal order judgment task. J Cogn Neurosci 19: 109–120. https://doi.org/10.1162/jocn. 2007.19.1.109 15. Vibell J, Klinge C, Zampini M, Spence C, Nobre AC (2017) Differences between endogenous attention to spatial locations and sensory modalities. Exp Brain Res 235(10): 2983–2996. https://doi.org/10.1007/ s00221-017-5030-4 16. Gomez-Ramirez M, Hysaj K, Niebur E (2016) Neural mechanisms of selective attention in the somatosensory system. J Neurophysiol 116(3): 1218–1231. https://doi.org/10.1152/jn. 00637.2015 17. Hawkins HL, Shafto MG, Richardson K (1988) Effects of target luminance and cue validity on the latency of visual detection. Percept Psychophys 44(5):484–492. https://doi. org/10.3758/bf03210434 NJ, 18. Giabbiconi CM, Trujillo-Barreto Gruber T, Mu¨ller MM (2007) Sustained spatial attention to vibration is mediated in primary somatosensory cortex. NeuroImage 35(1): 255–262. https://doi.org/10.1016/j. neuroimage.2006.11.022 19. Michie PT, Bearpark HM, Crawford JM, Glue LC (1987) The effects of spatial selective attention on the somatosensory event-related potential. Psychophysiology 24(4):449–463. https://doi.org/10.1111/j.1469-8986.1987. tb00316.x 20. Garcia-Larrea L, Lukaszewicz AC, Mauguiere F (1995) Somatosensory responses during selective spatial attention: the N120-to-N140 transition. Psychophysiology 32(6):526–537.

446

Jonas Vibell et al.

https://doi.org/10.1111/j.1469-8986.1995. tb01229.x 21. Eimer M, Forster B (2003) Modulations of early somatosensory ERP components by transient and sustained spatial attention. Exp Brain Res 151(1):24–31. https://doi.org/10.1007/ s00221-003-1437-1 22. Eimer M, Forster B (2003) The spatial distribution of attentional selectivity in touch: evidence from somatosensory ERP components. Neurophysiol 114(7):1298–1306. Clin https://doi.org/10.1016/s1388-2457(03) 00107-x 23. Michie PT (1984) Selective attention effects on somatosensory event-related potentials. Ann N Y Acad Sci (425):250–255. https://doi.org/ 10.1111/j.1749-6632.1984.tb23542.x 24. Gherri E, Forster B (2012) Crossing the hands disrupts tactile spatial attention but not motor attention: evidence from event-related potentials. Neuropsychologia 50(9):2303–2316. h t t p s : // d o i . o r g / 1 0 . 1 0 1 6 / j . neuropsychologia.2012.05.034 25. Gherri E, Gooray E, Forster B (2016) Cue-locked lateralized components in a tactile spatial attention task: evidence for a functional dissociation between ADAN and LSN. Psychophysiology 53(4):507–517. https://doi.org/ 10.1111/psyp.12596 26. Karlinski M, Jones A, Forster B (2019) Electrophysiological evidence for changes in attentional orienting and selection in functional somatic symptoms. Clin Neurophysiol 130(1):85–92. https://doi.org/10.1016/j. clinph.2018.09.027 27. Spence C, Shore DI, Klein RM (2001) Multisensory prior entry. J Exp Psychol 130:799– 832. https://doi.org/10.1037/0096-3445. 130.4.799 28. Macaluso E, Frith C, Driver J (2000) Selective spatial attention in vision and touch: unimodal and multimodal mechanisms revealed by PET. J Neurophysiol 83(5):3062–3075. https:// doi.org/10.1152/jn.2000.83.5.3062 29. Spence C, Driver J (2004) Crossmodal Space and Crossmodal Attention. Oxford University Press, Oxford, UK 30. Eimer M, Van Velzen JL (2005) Spatial tuning of tactile attention modulates visual processing within hemifields: an ERP investigation of crossmodal attention. Exp Brain Res 166(3–4):402–410. https://doi.org/10. 1007/s00221-005-2380-0 31. Gondan M, Vorberg D, Greenlee MW (2007) Modality shift effects mimic multisensory interactions: an event-related potential study. Exp

Brain Res 182(2):199–214. https://doi.org/ 10.1007/s00221-007-0982-4 32. Lloyd DM, Shore DI, Spence C, Calvert GA (2003) Multisensory representation of limb position in human premotor cortex. Nat Neurosci 6:17–18. https://doi.org/10.1038/ nn991 33. Longo MR, Musil JJ, Haggard P (2012) Visuo-tactile integration in personal space. J Cogn Neurosci 24(3):543–552. https://doi. org/10.1162/jocn_a_00158 34. Rigato S, Bremner AJ, Mason L, Pickering A, Davis R, Van Velzen JL (2013) The electrophysiological time course of somatosensory spatial remapping: vision of the hands modulates effects of posture on somatosensory evoked potentials. Eur J Neurosci 38(6): 2884–2892. https://doi.org/10.1111/ejn. 12292 35. Schepers RJ, Ringkamp M (2010) Thermoreceptors and thermosensitive afferents. Neurosci Biobehav Rev 34(2):177–184. https:// doi.org/10.1016/j.neubiorev.2009.10.003. Epub 2009 Oct 12. PMID: 19822171 36. Watkins RH, Dione M, Ackerley R, Backlund Wasling H, Wessberg J, Lo¨ken LS (2021) Evidence for sparse C-tactile afferent innervation of glabrous human hand skin. J Neurophysiol 125(1):232–237 37. Ackerley R, Watkins RH (2018) Microneurography as a tool to study the function of individual C-fiber afferents in humans: responses from nociceptors, thermoreceptors, and mechanoreceptors. J Neurophysiol 120(6): 2834–2846. https://doi.org/10.1152/jn. 00109.2018 38. Morrison I, Lo¨ken LS, Olausson H (2010) The skin as a social organ. Exp Brain Res 204(3): 305–314. https://doi.org/10.1007/s00221009-2007-y 39. Ackerley R, Eriksson E, Wessberg J (2013) Ultra-late EEG potential evoked by preferential activation of unmyelinated tactile afferents in human hairy skin. Neurosci Lett 535:62–66. https://doi.org/10.1016/j.neulet.2013. 01.004 40. Haggarty C, Malinowski P, McGlone F, Walker S (2020) Autistic traits modulate cortical responses to affective but not discriminative touch. Eur J Neurosci 51(8):1844–1855. https://doi.org/10.1111/ejn.14637 41. Bromm B, Neitzel H, Tecklenburg A, Treede RD (1983) Evoked cerebral potential correlates of C-fibre activity in man. Neurosci Lett 43(1):109–114. https://doi.org/10.1016/ 0304-3940(83)90137-4

Somatosensory EEG 42. Bromm B, Treede R (1987) Reliability and validity of C-fibre evoked cerebral potentials in man. Pain 30:S179. https://doi.org/10. 1016/0304-3959(87)91430-8 43. Carr L, Iacoboni M, Dubeau MC, Mazziotta JC, Lenzi GL (2003) Neural mechanisms of empathy in humans: a relay from neural systems for imitation to limbic areas. Proc Nat Acad Sci 100(9):5497–5502. https://doi.org/10. 1073/pnas.0935845100 44. Nakamura K, Kawashima R, Ito K, Sugiura M, Kato T, Nakamura A, Hatano K, Nagumo S, Kubota K, Fukuda H, Kojima S (1999) Activation of the right inferior frontal cortex during assessment of facial emotion. J Neurophysiol 82(3):1610–1614. https://doi.org/10.1152/ jn.1999.82.3.1610 45. Vuilleumier P, Pourtois G (2007) Distributed and interactive brain mechanisms during emotion face perception: evidence from functional neuroimaging. Neuropsychologia 45(1): 174–194. https://doi.org/10.1016/j. neuropsychologia.2006.06.003 46. Sel A, Forster B, Calvo-Merino B (2014) The emotional homunculus: ERP evidence for independent somatosensory responses during facial emotional processing. J Neurosci 34(9): 3263–3267. https://doi.org/10.1523/ JNEUROSCI.0106-13.2014 47. Haxby JV, Hoffman EA, Gobbini MI (2002) Human neural systems for face recognition and social communication. Biol Psychiatry 51(1): 59–67. https://doi.org/10.1016/s00063223(01)01330-0 48. Galvez-Pol A, Calvo-Merino B, Capilla A, Forster B (2018) Persistent recruitment of somatosensory cortex during active maintenance of hand images in working memory. NeuroImage 174:153–163. https://doi.org/ 10.1016/j.neuroimage.2018.03.024 49. Galvez-Pol A, Calvo-Merino B, Forster B (2020) Revealing the body in the brain: an ERP method to examine sensorimotor activity during visual perception of body-related information. Cortex 125:332–344. https://doi. org/10.1016/j.cortex.2020.01.017 50. Galvez-Pol A, Forster B, Calvo-Merino B (2018) Modulation of motor cortex activity in a visual working memory task of hand images. Neuropsychologia 117:75–83. https://doi. org/10.1016/j.neuropsychologia.2018. 05.005 51. Tame` L, Longo MR (2020) Probing the neural representations of body-related stimuli: comment on “revealing the body in the brain: an ERP method to examine sensorimotor activity during visual perception of the body-related information” by Alejandro Galvez-Pol, Beatriz

447

Calvo-Merino and Bettina Forster. Cortex 134:358–361. https://doi.org/10.1016/j.cor tex.2020.08.019 52. Bufalari I, Ionta S (2013) The social and personality neuroscience of empathy for pain and touch. Front Hum Neurosci. https://doi.org/ 10.3389/fnhum.2013.00393 53. Gillmeister H, Bowling N, Rigato S, Banissy MJ (2017) Inter-individual differences in vicarious tactile perception: a view across the lifespan in typical and atypical populations. Multisens Res 30(6):485–508. https://doi. org/10.1163/22134808-00002543 54. Keysers C, Kaas JH, Gazzola V (2010) Somatosensation in social perception. Nat Rev Neurosci 11(6):417–428. https://doi.org/10. 1038/nrn2833 55. Lamm C, Silani G, Singer T (2015) Distinct neural networks underlying empathy for pleasant and unpleasant touch. Cortex 70:79–89. https://doi.org/10.1016/j.cortex.2015. 01.021 56. Adler J, Gillmeister H (2019) Bodily selfrelatedness in vicarious touch is reflected at early cortical processing stages. Psychophysiology 56(12):e13465. https://doi.org/10. 1111/psyp.13465 57. Adler J, Schabinger N, Michal M, Beutel ME, Gillmeister H (2016) Is that me in the mirror? Depersonalisation modulates tactile mirroring mechanisms. Neuropsychologia 85:148–158. h t t p s : // d o i . o r g / 1 0 . 1 0 1 6 / j . neuropsychologia.2016.03.009 58. Rigato S, Bremner A, Gillmeister H, Banissy MJ (2019) Interpersonal representations of touch in somatosensory cortex are modulated by perspective. Biol Psychol 146:107719. https://doi.org/10.1016/j.biopsycho.2019. 107719 59. Farmer H, Cataldo A, Adel N, Wignall E, Gallese V, Deroy O, Hamilton A, Ciaunica A (2020) The detached self: investigating the effect of depersonalisation on self-bias in the visual remapping of touch. Multisens Res:1–22. h t t p s : // d o i . o r g / 1 0 . 1 1 6 3 / 2 2 1 3 4 8 0 8 bja10038 60. Hauser SC, McIntyre S, Israr A, Olausson H, Gerling GJ (2019) Uncovering human-tohuman physical interactions that underlie emotional and affective touch communications. In: IEEE World Haptics Conference, 9–12 July 2019. https://doi.org/10.1109/whc.2019. 8816169 61. Bridges D, Pitiot A, MacAskill MR, Peirce JW (2020) The timing mega-study: comparing a range of experiment generators, both lab-based

448

Jonas Vibell et al.

and online. PeerJ 8:e9414. https://doi.org/ 10.7717/peerj.9414 62. Luck S (2014) An introduction to the eventrelated potential technique. MIT Press 63. Mauguiere F (1999) Somatosensory evoked potentials. In: Niedermeyer E, Lopes da Silva F (eds) Electroencephalography: basic principles, clinical applications and related fields. Williams and Wilkins 64. Zampini M, Brown T, Shore DI, Maravita A, Roder B, Spence C (2005) Audiotactile temporal order judgments. Acta Psychol 118(3): 277–291. https://doi.org/10.1016/j.actpsy. 2004.10.017 65. Eads J, Lorimer Moseley G, Hillier S (2015) Non-informative vision enhances tactile acuity: a systematic review and meta-analysis. Neuropsychologia 75:179–185. https://doi.org/10. 1016/j.neuropsychologia.2015.06.006 66. Taylor-Clarke M, Kennett S, Haggard P (2002) Vision modulates somatosensory cortical processing. Curr Biol 12:233–236. https:// doi.org/10.1016/S0960-9822(01)00681-9 67. Gillmeister H, Forster B (2010) Vision enhances selective attention to body-related information. Neurosci Lett 483:184–188. https://doi.org/10.1016/j.neulet.2010. 08.004 68. Sambo CF, Gillmeister H, Forster B (2009) Viewing the body modulates neural mechanisms underlying sustained spatial attention in touch. Eur J Neurosci 30(1):143–150. https://doi.org/10.1111/j.1460-9568 69. Gillmeister H, Forster B (2012) Adverse effects of viewing the hand on tactile-spatial selection between fingers depend on finger posture. Exp Brain Res 221(3):269–278. https://doi.org/ 10.1007/s00221-012-3171-z 70. Gillmeister H, Sambo CF, Forster B (2010) Which finger? Early effects of attentional selection within the hand are absent when the hand is viewed. Eur J Neurosci 31(10):1874–1881. https://doi.org/10.1111/j.1460-9568.2010. 07195.x 71. Wesslein AK, Spence C, Frings C (2014) Vision affects tactile target and distractor processing even when space is task-irrelevant. Front Psychol 6(5):84. https://doi.org/10.3389/ fpsyg.2014.00084 72. Wahn B, Murali S, Sinnett S, Ko¨nig P (2017) Auditory stimulus detection partially depends on visuospatial attentional resources. i-Perception 8(1):2041669516688026 73. Driver J, Grossenbacher PG (1996) Multimodal spatial constraints on tactile selective attention. In: Innui T, McClelland J (eds)

Attention & performance, vol XVI. MIT Press, Cambridge, MA 74. Shore DI, Spry E, Spence C (2002) Confusing the mind by crossing the hands. Cogn Brain Res 14(1):153–163. https://doi.org/10. 1016/s0926-6410(02)00070-8 75. Heed T, Buchholz VN, Engel AK, Ro¨der B (2015) Tactile remapping: from coordinate transformation to integration in sensorimotor processing. Trends Cogn Sci 19(5):251–258. https://doi.org/10.1016/j.tics.2015.03.001 ˜ o´n E, Soto-Faraco S (2008) Changing 76. Azan reference frames during the encoding of tactile events. Curr Biol 18(14):1044–1049. https:// doi.org/10.1016/j.cub.2008.06.045 77. Gillmeister H, Adler J, Forster B (2010) Object-guided spatial attention in touch: holding the same object with both hands delays attentional selection. J Cogn Neurosci 22(5): 931–942. https://doi.org/10.1162/jocn. 2009.21265 78. Eimer M, Forster B, Van Velzen J, Prabhu G (2005) Covert manual response preparation triggers attentional shifts: ERP evidence for the premotor theory of attention. Neuropsychologia 43(6):957–966. https://doi.org/10. 1016/j.neuropsychologia.2004.08.011 79. Calder AJ, Young AW (2005) Understanding the recognition of facial identity and facial expression. Nat Rev Neurosci 6(8):641–651. https://doi.org/10.1038/nrn1724 80. Fairhall SL, Ishai A (2007) Effective connectivity within the distributed cortical network for face perception. Cereb Cortex 17(10): 2400–2406. https://doi.org/10.1093/ cercor/bhl148 81. Pitcher D (2014) Facial expression recognition takes longer in the posterior superior temporal sulcus than in the occipital face area. J Neurosci 34(27):9173–9177. https://doi.org/10. 1523/JNEUROSCI.5038-13.2014 82. Adolphs R (2002) Neural systems for recognizing emotion. Curr Opin Neurobiol 12(2): 169–177. https://doi.org/10.1016/s09594388(02)00301-x 83. Hu C, Di X, Eickhoff SB, Zhang M, Peng K, Guo H, Sui J (2016) Distinct and common aspects of physical and psychological selfrepresentation in the brain: a meta-analysis of self-bias in facial and self-referential judgements. Neurosci Biobehav Rev 61:197–207. https://doi.org/10.1016/j.neubiorev.2015. 12.003 84. Ploner M, Schmitz F, Freund HJ, Schnitzler A (2000) Differential organization of touch and pain in human primary somatosensory cortex. J

Somatosensory EEG Neurophysiol 83:1770–1776. https://doi. org/10.1152/jn.2000.83.3.1770 85. Chan AW, Baker CI (2015) Seeing is not feeling: posterior parietal but not somatosensory cortex engagement during touch observation. J Neurosci 35:1468–1480. https://doi.org/ 10.1523/JNEUROSCI.3621-14.2015

449

86. Ree A, Mayo L, Leknes S, Sailer U (2019) Touch targeting C-tactile afferent fibers has a unique physiological pattern: a combined electrodermal and facial electromyography study. Biol Psychol 140:55–63. https://doi.org/10. 1016/j.biopsycho.2018.11.006

Chapter 20 Neurostimulation in Tactile Perception Luigi Tame` and Nicholas Paul Holmes Abstract Neurostimulation techniques are used to study the healthy human central and peripheral nervous system non-invasively by stimulating neural tissue magnetically or electrically. Such approaches have been successfully applied to study the motor system as well as several other brain systems. This chapter will focus on stimulation of the somatosensory system. Typically, neurostimulation is applied to a certain brain area by positioning a coil (e.g., in transcranial magnetic stimulation, TMS) or an electrode (e.g., transcranial electrical stimulation, TES) on the scalp location over the brain area of interest. When primary motor cortex (M1) is stimulated with TMS, motor-evoked potentials (MEPs) and twitches are observed in the targeted muscles of the body. However, unlike over M1, stimulation to somatosensory and other cortices does not produce immediately observable outputs. This introduces problems of localization and other challenges, such as the optimal experimental designs and behavioral tasks, when using neurostimulation to study tactile perception. This chapter will describe and evaluate these approaches. Practical and participantspecific difficulties will be noted. Neurostimulation methods offer relatively cheap and reliable means of modulating somatosensation, yet care is required to ensure that the experimental design is adequate, that the optimal location is stimulated, and that the data are able to answer your theoretical question. Key words Transcranial magnetic stimulation, Transcranial direct current stimulation, Transcranial alternating current stimulation, Detection, Discrimination, Vibrotactile, Fingers, Somatosensory cortex, Localization, Neuronavigation

1

Introduction

1.1 A Brief Outline and History of Brain Stimulation

Non-invasive brain stimulation (NIBS) techniques can be used to study sensory, cognitive, and motor processes in the brain and peripheral nervous system [1]. NIBS methods include primarily transcranial electrical stimulation (TES) and transcranial magnetic stimulation (TMS). These techniques are typically used to establish the excitability of a certain brain region and of corticocortical and corticospinal pathways. Moreover, by interference with ongoing processing, it can be used to establish the role played by a brain region in a certain process [2, 3] and to monitor the magnitude and timing of physiological responses in healthy people and individuals with neurological disorders. Recently, such approaches have also

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6_20, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

451

452

Luigi Tame` and Nicholas Paul Holmes

proved useful as a treatment for patients suffering from neurological and psychiatric conditions. There is evidence that TES can be effective, for instance, in reducing impairment after stroke [4], improving symptoms of neglect [5], and reducing symptoms of depression [6]. What is considered the major advantage of these techniques, differently from other neuroimaging methods such as functional magnetic resonance (fMRI, see Chapter 18, this volume), electroencephalography (EEG, see Chapter 19, this volume), and magnetoencephalography (MEG), is the possibility to establish a causal role of a particular brain area in a certain sensory, motor, or cognitive process. Therefore, these approaches are potentially very powerful. However, they also come with limitations and require a series of important precautions that must be taken for appropriate and effective use. In this chapter, we will describe the most common neurostimulation protocols adopted in the context of stimulating the somatosensory system. We will discuss the strengths and limitations of these techniques and provide advice on good practice that may help to improve the reliability and effectiveness of studies in this domain. We will focus on the practical use of TES and TMS techniques in studying the somatosensory system rather than discussing models and mechanisms that could explain the stimulation effects, a topic that is still unclear and a matter of debate [7, 8]. Although these techniques are considered safe, it is critical to follow some basic safety procedures for routine clinical and research applications, to reduce the risks of adverse effects [9– 11]. Such procedures will be briefly detailed later in the chapter. Historically, the first successful controlled direct electrical stimulation of the mammalian cerebral cortex can be traced back to the study performed by Fritsch & Hitzig in 1870 [12], in which they delivered galvanic current through bipolar electrodes to the anterior half of the dog’s hemisphere. As a result of this stimulation, they found movements of muscle groups in the opposite half of the dog’s body. The first human brain stimulation study, as credited by Beevor and Horsley (1890), was performed in 1874 by Roberts Bartholow, an American surgeon. Another surgeon, Harvey Cushing, stimulated the somatosensory cortex of an awake human patient in 1909 [13]. But the first systematic attempt to electrically stimulate the human somatosensory cortex was the pioneering work of Penfield and Boldrey [14], in which they mapped the somatosensory and motor cortices in humans using electrical stimulation. They described the functional anatomy of these brain areas in human, emphasizing the somatotopic organization of the hemibody [15, 16]. These pioneering works lead to the development of trans-cranial electrical stimulation (TES) methodology, whereby applying an electrical current at the scalp surface can directly affect brain activity. Originally, TES was applied with high intensity (i.e., 3–60 mA), however, more recently, lower intensities are most

Neurostimulation

453

commonly used (i.e., 1–2 mA) [17] (note that we will refer to the latest approach in all the subsequent parts of this chapter). Low-intensity TES is typically applied using different approaches (e.g., transcranial direct current stimulation, TDCS; transcranial alternating current stimulation, TACS; transcranial random noise stimulation, TRNS), which will be discussed later in relation to the somatosensory system. A comprehensive discussion on these different methods is beyond the scope of the present chapter (for a review on the topic see [18]). The first attempt to apply magnetic stimulation to the head came a bit later, starting with the work of d’Arsonval in 1896 [19], in which the author applied an alternating current to a coil surrounding the head of an individual, successfully inducing phosphenes (i.e., luminous floating stars, zigzags, swirls, spirals, and squiggles seen in the visual field). Years later, in 1959, Kolin and colleagues [20] for the first time applied magnetic stimulation to the sciatic nerve of a frog, inducing muscle contraction. Bickford and Freming extended this work in 1959 by stimulating peripheral nerves in animals and humans [21]. The birth of modern TMS methodology was 1985, with the work of Barker and colleagues [22] in which the motor cortex was stimulated using a coil placed over the scalp. One of the early attempts of stimulating primary somatosensory cortex to disrupt tactile perception in humans was provided by Cohen and colleagues in 1991 [23]. They reported that detection of electrical stimuli delivered on the index finger of a subject was attenuated or completely abolished when single pulse (sp) TMS was applied over contralateral sensorimotor cortex between 200 milliseconds (ms) before and 20 ms after the occurrence of the tactile stimulus. Figure 1 depicts a stimulation coil positioned to target the somatosensory cortex. 1.2 Stimulating the Somatosensory System

The somatosensory system, and in particular the primary somatosensory cortex (S1), which was central to the studies conducted by Penfield and colleagues, can be divided into four distinct cytoarchitectonic areas [24]. These areas are the Brodmann areas (BA) 3a, 3b, 1, and 2 [25] (see Chapters 16, 17, and 18, this volume). S1 covers a large territory along the central sulcus and postcentral gyrus, where there are several topographically organized maps of the body [26]. Although the precise location of each body part representation may vary between people, there is a within- and between-person consistency in the locations of different body part representations in S1. This allows neuroscientists to aggregate and map the results from different people in studies of cortical somatosensory functions [27]. For instance, the different fingers are organized (i.e., from little finger to the thumb) following a medio-to-lateral distribution symmetrically in the two hemispheres [26]. The anatomical location of S1, in the context of

454

Luigi Tame` and Nicholas Paul Holmes

Fig. 1 A figure-of-8 coil (black) positioned over the primary somatosensory cortex (orange) in the left hemisphere: note that in the image the scalp has been removed to better appreciate the TMS coil location relative to the brain area

neurostimulation, has some advantages, for instance being very superficial—close to the scalp—and therefore easy to stimulate. However, there are also some disadvantages, for instance being anatomically adjacent to the primary motor cortex. The implications of these anatomical and other physiological characteristics of the somatosensory cortex will be discussed in detail later in the chapter. When planning neurostimulation studies which aim to affect the somatosensory system while measuring participants’ performance in a perceptual or cognitive task, the researcher must take several decisions. Such choices will later determine the quality of the data and, in turn, the reliability and effectiveness of the study. In subsequent sections, we will first describe the materials that are typically necessary to perform TES and TMS studies (e.g., hardware, consumables, and software). Then, we will review the most common parameters and protocols used to perform neurostimulation of the somatosensory system, primarily while participants perform a tactile task. We will continue by describing the different options available to identify an optimal scalp location as a target site for stimulation. The main experimental designs, type of tasks, and the dependent variables that can be measured will also be considered. In the last section, we will provide a series of recommendations that can be useful to overcome some limitations common to neurostimulation of the somatosensory cortex.

Neurostimulation

2

455

Materials Here, we will describe some currently available TES and TMS devices that can be used to stimulate the somatosensory system both at cortical and peripheral levels. Then, we will present current neuronavigation systems (NNS) devices, which are now considered key tools to allow precise stimulation of the brain area of interest during testing. This logic applies well beyond stimulation of the somatosensory cortex (i.e., any other brain area), with the exception of the primary motor cortex (M1), from which TMS provides direct readouts of changes in the cortical activity by means of the motor evoked potential (MEP) response.

2.1 TES and TMS Hardware

For both TES and TMS approaches, there are several available devices (Fig. 2). Some examples of companies providing TES equipment are MagStim, NeuroConn, and BrainStim. Such TES devices currently cost several thousand pounds, depending on the features and model, and are much less expensive than TMS devices. In terms of the available TMS devices, there are several options. The main manufacturers include, but are not limited to: Magstim, Mag&More, MagVenture, Deymed, NeuroStart, Nexstim,

Fig. 2 Neurostimulation hardware: (a) Magstim BiStim system, with two Magstim-200 magnetic stimulators, (b) MagVenture system, with one Butterfly figure-eight coil, (c) NeuroConn tES device with a pair of rubber electrodes and sponge pads applied on a mannequin’s head using elastic rubber straps, (d) BrainStim tES device with a pair of rubber electrodes and sponge pads

456

Luigi Tame` and Nicholas Paul Holmes

Neurosoft, and Neuronetics. Such devices cost perhaps ten times as much as TES systems, again depending on the features and model. They can be used both in research and clinical settings, although specific approvals may be required for clinical use. There is some debate about whether such devices provided by different manufacturers are equivalent in terms of their effectiveness. A few studies have directly compared them, though, in the clinical context [28] and in the research setting [29, 30]. Such comparisons show there may be slight differences in the strength and sound produced by different devices. It is therefore suggested that the intensity of TMS should be adjusted according to the individual participant’s motor threshold to obtain comparable responses with different devices [29]. Establishing the resting (RMT) or active (AMT) motor threshold is an almost universal standard practice to tailor the stimulation intensity of TMS to each participant. Such thresholds are usually defined, according to Rossini et al. [31], as the minimum stimulator intensity that produces MEPs with a peak-to-peak amplitude of 50μV or higher on 5 or more out of 10 trials in which TMS is applied over M1 [31]. Even though this threshold is computed for M1 stimulation, the stimulator intensity defined is often used for stimulation of all other brain areas. This reference intensity is based only on the motor cortex. More recently, 10 out of 20 trials has been recommended [32], while more theoretically-motivated and potentially more efficient approaches have been developed, often called “threshold-hunting,” but involving parameter estimation by sequential testing (PEST or QUEST, see Chapter 1, this volume). 2.2 Neuronavigation Systems

Neuronavigation systems in the context of TES and TMS are very valuable tools which allow more accurate stimulation of specific brain areas. Typically, a neuronavigator system consists of a means of locating the participant’s head (3D motion capture) and a means of registering the subject’s head to a 3D MRI scan. Anatomical and/or functional images from an fMRI scan can be imported into the neuronavigation software. If only anatomical images are available, the locus of stimulation can be visualized and targeted using the subject space or in a normalized brain template space (e.g., MNI). If functional images are also available an fMRI contrast for the relevant condition can be used to visualize and target the functional activated brain area of interest [33]. Neuronavigation (Fig. 3) is critical for accurately stimulating all brain regions, with probably the only exception being M1, from which we have relatively direct readout of excitability from MEPs. The way in which the TMS coil is positioned on the scalp, in terms of location (2 or 3 degrees of freedom) and orientation (3 degrees of freedom), can significantly affect the brain’s physiological response [34]. Critically, neuronavigation devices can effectively monitor these parameters when the coil is initially positioned and later while TMS is delivered

Neurostimulation

457

Fig. 3 Typical neuronavigation system used to assist with TMS positioning. The subject is sitting down, wearing a cap to flatten hair and allow easy access to and marking on the head (a). The experimenter holds a TMS coil (b) in the left hand, and supports the subject’s head in the right hand. Behind the experimenter is a Polaris infrared camera, which monitors the locations of reflective markers placed on the subject’s head and the TMS coil (c). On the right is a workstation (d) incorporating the TMS hardware (bottom) and the neuronavigation computer monitor (top). As the experimenter moves the TMS coil around on the subject’s head, the MRI scan of the subject’s brain is rotated accordingly, showing the intended target and/or trajectory of the TMS pulse (e). See also Note 4.6

(online) during the experimental phase. Some neuronavigation systems are sold including both hardware and software components. One widely used system is Brainsight, which also provides MEP recording options. Similar systems are the Soterix, BrainVoyager, or Brainlab. These systems may cost as much as the TMS system itself, which is currently more than £20 k. An interesting freely available neuronavigator software system is the InVesalius Navigator (IN) [35], which is an implementation of the InVesalius software program [36]. This system provides image-guided coil placement for TMS. It comes with software that can communicate with different motion tracking devices (i.e., MicronTracker, Patriot, Fasttrack, and Isotrack II). Note that, differently from the systems described above, in this case the researcher will need to have available such a motion tracking system to interface with the neuronavigation software. Likewise, this software will not provide MEP recordings, and this needs separate hardware and software.

458

Luigi Tame` and Nicholas Paul Holmes

The “gold standard” for locating brain areas of interest is to have a well-calibrated neuronavigation system along with a recent functional and/or structural MRI scan for each participant. When this is not possible or available, several different methods are typically used to locate brain areas on the scalp based, primarily, on the 10–20 system for locating electroencephalography (EEG) electrodes. To facilitate measurement of participants’ heads, it is common practice to use a cap (e.g., swimming cap) on which researchers can mark anatomical landmarks such as the inion or vertex or apply geometric grids of locations with the origin, for example, on the optimal location for eliciting MEPs from over M1. From these standard and relatively unambiguous landmarks, it is possible to move the TMS to different locations as a proxy for neuronavigation. In general, the further away from M1 the less precise the coil localization. At least for the S1 hand area, with careful measurements of the head and M1 locations, we can now be quite confident about accurate coil positioning [27, 37]. 2.3 Vibrotactile Devices and Environmental Conditions

Several types of tactile stimulation devices can be used in behavioral and neuroimaging tactile tasks. A first main distinction can be made between electrical and mechanical tactile stimulators. The most commonly used electrical stimulator is the Digitimer, a constantcurrent stimulator device. Bipolar or ring electrodes are used to stimulate certain body parts, such as the fingers. In particular, bipolar adhesive electrodes can be placed on the distal and middle phalanges of one finger (e.g., left index finger), with the anode approximately 2 cm proximal to the cathode [38]. Such stimulation is very effective and produces a good neural response that can easily be recorded using EEG and the peak somatosensory evoked potential (SEP) location and time used in subsequent neurostimulation experiments. However, the stimulation, despite being applied locally, tends to spread across nearby regions of the skin. Mechanical stimulators may instead be better in terms of targeting a certain skin location and reducing the spread of the signal towards neighboring skin regions. Some of these devices work better when the stimulation needs to be delivered at low frequencies (e.g., 0–40 Hz) and others work better when the stimulation is delivered at higher frequencies (e.g., 100 Hz and above). The former are the air-puff stimulators, pneumatic, and/or hydraulic stimulators that are also typically used in a MEG and fMRI settings given their compatibility with these techniques (see Chapter 18, this volume). Solenoids, loudspeakers, and Oticon (i.e., bone conductor) stimulators are also widely used for higher frequency stimulation. Depending on the parameters set, it is also possible to create a tapped or a more vibratory stimulation. MEG- and MRI-compatible piezoelectric stimulators are also provided by Quaerosys and Dancer Design. These stimulators can consist of a matrix of 10 or more rods (1 mm in diameter), protruding from a

Neurostimulation

459

flat surface of 4 × 8 mm. The rod can protrude and retract at 20 Hz for 1000 ms, producing clearly perceivable skin indentations [39]. If the matrix is large enough, it is possible to create different spatial patterns of stimulation that can, for instance, resemble alphabet letters (see Chapter 4, this volume). Temperature and humidity are known to be important factors that can affect tactile performance (see Chapter 9, this volume). It is important that these parameters are monitored in the experiments [40]. This is critical when, for instance, a researcher wants to perform a behavioral tactile study that has been previously done in a laboratory room (about 24 °C) or within the fMRI scanner room (18 °C) [41]. Remember, too, that electromagnetic interference between your neurostimulation and your somatosensory hardware is always a strong possibility (see Note 4.1).

3

Methods In this section, we will describe the typical experimental designs and protocols used to perform a neurostimulation study in the context of tactile perception. As a general remark, despite TES and TMS being brain stimulation techniques, the main outcomes of these experiments are typically behavioral data. The general rules that apply to this type of studies are therefore very similar to those used for behavioral experiments (see Chapter 1, this volume). We will discuss three main aspects: experimental design, the type of neurostimulation to use, and how test and control sites for stimulation can be localized on the scalp.

3.1 Experimental Design

There are different types of designs that are currently used to assess participants’ performance in tactile perception tasks under neurostimulation (i.e., TES and TMS). Note that, as these are similar for tactile experiments not using neurostimulation, such designs won’t be covered in detail and extended description is provided in Chapter 1 of the present volume.

3.1.1 Single-Interval Designs

One type of design used for TES and TMS studies investigating tactile perception is a one-interval forced choice (1-IFC) design (i.e., also known as “Yes”/“No” task) [23, 42] in which participants have to report on each trial if a target stimulus (e.g., a tactile stimulus on a finger of a certain intensity or frequency) is present or not. Different designs could also require a one-interval discrimination or classification (e.g., was it a strong or a weak stimulus?) The target stimulus in a detection task is typically delivered randomly in 50% of the trials. Depending on the study specific purpose, feedback can be provided to participants about their performance. One major advantage of single-interval designs is that the total amount

460

Luigi Tame` and Nicholas Paul Holmes

of neurostimulation can be minimized. With only one interval, trials and blocks can be shorter, and if TMS is given on every interval, this halves the total number of TMS pulses given. A limitation of the one-interval design is that it can be affected by response bias—the tendency to respond more often with one option, such as “no” or “weak.” Therefore, the outcome of participants’ performance may well reflect a combination of both perceptual sensitivity and the decision criteria used to respond. To overcome such a limitation, it is recommended to adopt an analytic approach that allows you to disentangle these two components. In this regard, the “d-prime” index can be used to estimate the perceptual component and “criterion” the decisional component of the task [33]. Indeed, in a 1IFC tactile detection task in which TMS was applied in three different conditions (on the S1 target site, on a parietal control site, supramarginal gyrus, SMG, and with no TMS), we found that participants adopted a more conservative criterion when TMS was delivered over both S1 and SMG compared to no TMS [33]. Therefore, when the TMS was present, participants were more likely to report that the stimulus was absent, even though their ability to detect the tactile stimulus was unaffected (comparison between S1 and SMG vs. no TMS). 3.1.2 Multiple Interval Designs

A second type of approach is a 2 or 3-I/AFC design that, differently from the one just discussed, is a design that can be considered bias free or less prone to bias [43, 44]. In a typical experiment, participants are instructed to indicate in which of two temporal intervals a tactile stimulus was delivered on the body, or when a certain body part (e.g., the index finger) was stimulated. Critically, in each trial, the tactile stimulation is always present, though it is unknown to the participant in which of the two intervals (i.e., first or second). The occurrence of the tactile stimulus in the two intervals is randomly assigned. As for the 1-I/AFC design explained above, depending on the study purpose, feedback can be provided. We also used flashing light-emitting diodes (LED) to indicate to the participant the beginning of each interval and the potential imminent arrival of the tactile stimulus. We have found these indications are particularly useful when tactile threshold has to be established— the indicators allow participants to focus their attention in time and space to the potential arrival of the target. As mentioned elsewhere, these two types of design are not equivalent, even though they may look quite similar. One-interval designs are known to have a greater memory component and also to be more prone to bias compared to two or three interval designs [45]. Designs with two or more intervals will take much longer to perform and require more neurostimulation. In the case of an on-line neurostimulation protocol, participants should receive TES or TMS on every trial and during every interval, regardless of the presence of the tactile stimulus [80, 131–

Neurostimulation

461

134]. Specifically, and against our expectations, we found subjectively that the cognitive demands of the task were greatly increased when TMS was delivered only on some trials—not knowing when a (distracting, slightly annoying) TMS pulse was going to be presented made it harder to focus attention on the target stimulus. When an off-line neurostimulation protocol is used, participants receive a period of TES or TMS before the behavioral task [135, 136]. In addition, there are also cases in which both online and offline neurostimulation (i.e., TES) is used as part of the same protocol [137]. 3.1.3 Detection and Discrimination Tasks

The two main experimental designs just described can be used with different tactile tasks. A very commonly used task in somatosensory neurostimulation is simple tactile detection, in which participants are instructed to report the presence of a stimulus (e.g., electrical, mechanical, or air puff) on the body [23, 33]. A more complex task is tactile localization, in which participants need to localize a tactile stimulus on their body, typically by naming the body part stimulated or by estimating the portion of the skin that received the stimulus [46]. Another level of complexity is provided by discrimination tasks [47], in which the participant has to establish whether a target tactile stimulus varies from a particular comparison stimulus, across different domains such as intensity [37], frequency [33, 48], spatial distribution [49, 50], temporal occurrence [51, 52], shape [53] or stimulus orientation [54, 55]. A widely used discrimination task is one that compares different frequencies of vibration. In a recent study, we adopted this paradigm in a 2-I/ AFC design in which we applied single or double TMS pulses over the scalp in three separate conditions (i.e., TMS over S1, inferior parietal lobe, SMG, or with the active coil held away from the head) while asking participants to perform a tactile frequency discrimination task [33]. In this experiment, participants received a tactile stimulation on their index and middle fingers, one in the first interval and another in the second interval. The tactile stimulation delivered in the first interval was the standard (i.e., a vibrotactile stimulus delivered always at 200 Hz on the index finger), whereas the tactile stimulus delivered in the second interval on the middle finger was lower or higher in frequency than that in the first interval (i.e., the standard). Participants were asked to indicate whether the second stimulus was higher or lower in frequency compared to the first stimulus (see Note 4.2). Responses were given by lifting a pedal placed beneath the left or right foot. TMS was applied at 25 and 75 ms after stimulus onset in both of the intervals [33]. We found that TMS over S1 significantly disrupted tactile discrimination performance, compared to when TMS was not present or was applied over the control site. The results of this experiment and one which involved a different type of tactile task is shown in Fig. 4.

462

Luigi Tame` and Nicholas Paul Holmes

Fig. 4 TMS over primary somatosensory cortex affects tactile detection in a taskdependent manner: (a) effect of S1 TMS on tactile detection in a one-interval forced choice (“yes”/“no”) design, expressed in d-prime, (b) tactile detection thresholds from a two-interval forced-choice design, expressed in arbitrary units. Both experiments were done with dual-pulse (dp) TMS over SI and SMG. Results are redrawn from Tame` and Holmes 2016, Experiments 1 and 2, respectively [33]. TMS over S1 was effective only in the one-interval design, implicating different task demands. P values are from a paired-sample t-test between SMG (control) and S1 (experimental) TMS conditions 3.1.4 Responses and Measures

In terms of participants’ responses, in tactile neurostimulation experiments, as in other cognitive studies, these can be given vocally or using the hands or feet. Vocal responses are useful when it is necessary to control for a lateralized response that is typically performed using hands or feet [56]. Vocal responses can be recorded using a microphone to allow off-line estimation of the

Neurostimulation

463

participant’s reaction times [57], particularly in the case in which speeded responses are required. When a hand or a foot is used for the response, inter-hemispheric interactions between sensory, motor, and sensorimotor brain areas between the two hemispheres [58] should also be considered to avoid potential confounds [59]. This problem can be overcome by using different effectors, for example by using one finger for the stimulation, and a different finger for the response [39]. The outcomes of somatosensory TES and TMS experiments are primarily behavioral data, except for motor evoked potentials (MEP) when TMS is near motor cortex, or when neurostimulation is coupled with other neuroimaging techniques such as EEG, MEG, fMRI (discussed in Subheadings 3.2.3 below). The dependent variables typically considered are accuracy, reaction times, d-prime (i.e., sensitivity), or criterion (i.e., decision processes). We describe some of these measures here, but see Chapters 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10, this volume, for a wide range of somatosensory variables. Here we just want to point out the critical role played by the decision criterion in tactile neurostimulation experiments (this logic certainly applies beyond the tactile task context). We have found that neurostimulation directly affects participants’ responses, changing their decisional criterion. The mere presence of neurostimulation (e.g., TMS over the scalp, regardless of the site) increased participants’ inclination to say that a tactile stimulus was not present. In theory, the decision criterion is independent from participants’ sensitivity [33], although in practice they often co-vary. If signal detection theory (SDT) approaches are not adopted during certain tactile neurostimulation tasks, it would be very difficult to interpret the results where both the decisional and perceptual components are blended. Therefore, it would be a beneficial practice to include such measurements, whenever possible, in all relevant TES and TMS studies, particularly, when a one-interval design is used (see Note 4.3). 3.1.5 Output from Motor Cortex Stimulation

Motor responses (MEPs) are very commonly recorded during TMS experiments when the motor cortex is stimulated. However, it is less common practice to record them when other brain areas, such as the somatosensory cortices, are stimulated. We argue that such measurements are very important in the context of somatosensory TMS, given the anatomical and physiological proximity of somatosensory and motor cortices, and the significant interactions between the two systems. Indeed, MEP recordings can be used as a tool to monitor online M1 activity while the somatosensory cortex is stimulated. This has two main advantages: first, it allows researchers to correlate these physiological data with behavioral performance in the tactile task, potentially on a trial-by-trial basis,

464

Luigi Tame` and Nicholas Paul Holmes

to rule out possible confounds arising from accidental stimulation of M1 while targeting S1. Second, the presence of MEPs while stimulating S1 may provide an indirect indication of the accuracy and sufficiency of the targeted S1 stimulation. Indeed, given that these two brain regions are very close anatomically, the lack of any MEP responses while stimulating S1 may suggest that the site selected may not be ideal (e.g., far from both M1 and S1) or that the TMS intensity is too low (see Note 4.4). Several other methods can be used to generate and study the indirect effects of neurostimulation on somatosensation. Specifically, these approaches include short- (SAI) and long- (LAI) afferent inhibition. In a typical afferent inhibition (AI) protocol, a single [60] or multiple [38] cutaneous stimulation (typically electrical) is followed at a certain delay by TMS of the motor cortex. These cutaneous stimulation designs have been shown to modulate the amplitude of the MEPs evoked by TMS over M1 [61–65]. SAI and LAI are two dependent phases of inhibition that occur when there is a short or long interval between the afferent input and the TMS pulse. That is, activity of neurons in M1 changes in response to peripheral stimulation [66] with a decreased [67] corticospinal excitability in an interval included between about 20 ms and 50 ms (i.e., SAI) and 200 ms and 1000 ms (i.e., LAI) following median nerve (MN) or digit stimulation [68]. Note that the specific timing within these ranges at which the inhibition emerges and the magnitude of the response depends on the stimulated nerve or digit. Somatosensory neurostimulation paradigms have been used as a tool to study sensorimotor control in healthy humans and sensorimotor function in disease and following neurological injury. This is possible because specific sensory and motor pathways are involved in the generation of the phenomena (for a comprehensive review on the topic, see [68]). The emergence of SAI and LAI is affected by factors such as the intensity of the cutaneous stimulation and the TMS, and the particular muscle recorded. A novel protocol called dual-site TMS (ds-TMS) has been developed relatively recently. In this method, ds-TMS is used to investigate intra- [69] and inter-hemispheric [70, 71] interactions between the somatosensory areas and the primary motor cortex (M1). When designing this type of experiments, it is also recommended to possibly test participants at similar day time as circadian rhythms in hormones and neuromodulators can affect neuroplasticity [72]. 3.1.6 Participant Debriefing

It is always good and informative practice, after the testing phase, to debrief the participants. This is particularly critical in the context of neurostimulation, in at least three different respects. First, establishing whether the stimulation caused more than a reasonable temporary discomfort, if present. Second, it is beneficial to establish

Neurostimulation

465

through a series of open questions about the participant’s level of awareness of the target and control neurostimulation sites (i.e., for TMS studies) or montage condition (TES) during the experiment. Third, to assess whether there was a certain condition in which they experienced a greater discomfort or any general problems in paying attention [73] to the task compared to another condition or phase of the experiment. Ideally, the different neurostimulation conditions should be comparable in terms of their side effects, obviously excluding the no-TMS condition. Paying attention to what your participants experience is likely to improve recruitment and statistical power (see Note 4.5). 3.2 Neurostimulation Type 3.2.1 Transcranial Electrical Stimulation (TES)

As mentioned above, there are three main types of low-intensity TES: transcranial direct current (TDCS), alternating (TACS), and random noise stimulation (TRNS). In TCDS, a current is applied between two electrodes and has been argued to induce long-lasting changes in the brain. This technique may work by modulating brain excitability via membrane polarization. In TACS, a current is applied across the scalp which reverses polarity, for example, sinusoidally, at a certain frequency. TACS is supposed to interact with and affect ongoing electrical rhythms in the cortex. Finally, TRNS is an alternating current oscillating at random frequencies delivered through the scalp. For a comprehensive description of these methods, see [74]. These techniques are capable of inducing changes in the electrical activity of neurons by altering their membrane potentials and, in turn, changing their synaptic efficiency [7, 75]. There are several standard protocols used to interfere with brain activity using TES, however, its effect is the result of different interacting factors that go beyond the protocol, for example, the physiological status of the participant [76] and the type of task used. In general, and similar to TMS, there are two ways in which TES can be applied: before the participant performs the task (offline) or while the participant is performing the task (online). Several different protocols can be applied, and comprehensive guidelines, including safety and ethical considerations, is provided by Antal and colleagues [77]. Several studies have used TES with the aim of modulating neural activity of the somatosensory system, and, in turn, participants’ performance in tactile tasks [47, 78–82]. However, results regarding the effectiveness of such neurostimulation interventions on the somatosensory system are quite mixed. For instance, Saito and collaborators [47] compared the effects of TDCS, TACS, and TRNS protocols on inhibitory circuit activity in the primary somatosensory cortex while participants performed tactile spatial discrimination tasks. They found that, with the neurostimulation parameters they used (e.g., stimulation frequency at 140 Hz) anodal TDCS decreased the N20 component of the SEP and TRNS increased the N20 component. No effects were reported

466

Luigi Tame` and Nicholas Paul Holmes

after TACS stimulation. More studies that systematically investigate the effects of these types of neurostimulation, using different protocols and parameters, are needed to clarify the effectiveness of such techniques. 3.2.2 Transcranial Magnetic Stimulation (TMS)

Just as in TES, several different types of TMS protocols have been used to stimulate the somatosensory cortex [83]. Here, we will briefly describe the most popular ones. TMS can be applied with one pulse at a time, known as single-pulse TMS (spTMS); in pairs of pulses separated by a certain (short, e.g., 1–100 ms) temporal interval, known as paired- or dual- pulse TMS (ppTMS/dpTMS) [33]; or in trains of three or more TMS pulses, known as repetitive TMS (rTMS). Repetitive TMS can be given at a constant frequency (e.g., “slow” at 1 Hz, “rapid” at 10 Hz), or in more complex temporal patterns, such as in theta-burst TMS [84]. TMS can also be applied either online or offline. In online protocols, TMS is applied while participants are performing the task, for example just before or after stimulus presentation and/or during response generation [85], whereas offline TMS protocols apply TMS before, but not while, participants perform the task [86]. Typically, rTMS is used in offline protocols in which the effect caused by TMS is assumed to last after the end of stimulation, perhaps as long again as the duration of the TMS (e.g., TMS applied for 20 min may have 20 min after-effect). Such protocols are widely applied for studying the somatosensory system, however, a comprehensive list of available protocols can be found in the international TMS safety guidelines [32, 83]. The current developed and used protocols are considered both effective and safe in terms of the stimulation parameters, therefore, these guidelines should be carefully followed. In Fig. 5, we depict some of the most common TMS protocols used to stimulate the somatosensory system.

3.2.3 TES/TMS Combined with Other Neuroimaging Techniques

Neurostimulation of the somatosensory system has been used in combination with other neuroimaging techniques such as EEG, MEG, and fMRI. Such neuroimaging techniques won’t be discussed here in detail (see Chapters 17, 18, and 19, this volume). A few studies have attempted the concurrent use of TMS and fMRI techniques [87], however, such approaches, probably due to technical challenges, are not widely used. More common is to use TMS offline (e.g., rTMS) before participants undergo fMRI scanning [88]. Such an approach allows researchers to estimate the effects of TMS on blood-oxygenation level dependent (BOLD) signal changes across brain areas, including and beyond the site (s) stimulated. Combining TMS with EEG is becoming more popular to evaluate the effects of TMS, and to study the mechanisms behind the modulatory effects of TMS on cognitive and sensorimotor processing [89, 90]. These types of concurrent

Neurostimulation

467

Fig. 5 Some commonly used TMS protocols while stimulating the somatosensory system with TMS. SAI, short afferent inhibition; LAI, long afferent inhibition. Single pulse TMS typically requires 5–10 s intervals between stimulation. Dual pulse TMS may increase the effectiveness of TMS; the pulses could be 1 ms to 1 s apart (e.g., using the MagStim BiStim system). Repetitive TMS can be done “offline” at low frequencies (1 Hz) for up to 30 minutes before performing a task or with high-frequency patterned TMS, or “online,” with higher frequency (say, 5–20 Hz) trains of pulses during task performance. TMS over the motor cortex can be combined with prior somatosensory stimulation in the short- and longlatency afferent inhibition methods (SAI, LAI)—prior tactile stimulation inhibits the subsequent output from motor cortex TMS

TMS-EEG studies have been also growing in the tactile domain, and it seems promising for understanding the effects of TMS on the somatosensory system [91, 92]. TES has also been used successfully in combination with EEG and MEG [93]. 3.2.4 Peripheral Nerve Magnetic Stimulation

The somatosensory system is different from all other sensory systems in that the afferent nerves are distributed throughout the peripheral nervous system and the body [94]. This gives the somatosensory researcher unique access to the peripheral nerves, which is taken advantage of most spectacularly by microneurography (see Chapter 15, this volume). While the peripheral nerves have very often been stimulated to study the responses in the brain (i.e.,

468

Luigi Tame` and Nicholas Paul Holmes

the somatosensory evoked potential, SEP, see Chapter 19, this volume), the peripheral nervous system has not been much exploited as a site of electrical and magnetic interference with somatosensory perception. It is remarkable that there are so many studies of the effects of brain stimulation on somatosensory perception [27], but only two that we know of using magnetic stimulation of peripheral nerves [33, 95]. Stimulation of the peripheral nerve provides several advantages over stimulation of the brain, which we describe here in the context of interfering with tactile perception on the fingertip. First, between the index fingertip and the brain, there are perhaps 100 cm of peripheral nerves that can be targeted—along the arm, in the brachial plexus, even in the cervical spine and midbrain. Second, when the median nerve passes over the wrist and elbow joints, it is pushed very close to the surface. This allows precise electrical and magnetic stimulation at the lowest-possible intensity to affect the nerves. Third, successful nerve stimulation can be reported directly on a trial-by-trial basis by the participant. Nerve stimulation, even at low intensities, evokes a prominent tingling sensation up and down the arm; at higher intensities, stimulation of motor nerves also evokes objective responses in the EMG as well as visible twitches. The tingling side effect of stimulation can be used by the participant, inadvertently or on purpose, to monitor and control coil position. Fourth, and perhaps most importantly, stimulation of the peripheral nerve is, in our experience, much more tolerable than brain stimulation and poses fewer risks. Together, these advantages make peripheral nerve magnetic stimulation (PNMS) an ideal method for pilot-testing all potential brain stimulation studies of somatosensory perception. More than being just better in principle or in participant comfort and safety, PNMS works extremely well. Unlike several of our attempts at interfering with tactile perception using TMS, every experiment we’ve done with PNMS has worked as expected— stimulation over the median nerve impairs all aspects of task performance for tactile stimuli presented at the index or middle fingertip [33]. Stimulation impairs performance on tactile detection and discrimination in both one- and two-interval experimental designs and with both single and double-pulses of TMS. Increasing stimulation intensity worsens tactile detection thresholds in a predictable manner. Stimulation during presentation of the tactile stimulus has a greater effect on perceptual thresholds than stimulation before or after stimulus presentation. Stimulation of the median nerve has a greater effect on index finger perception than stimulation of the ulnar nerve, despite similar side effects. These results will be reported elsewhere. Based on our experience of PNMS, we strongly encourage and advise all somatosensory brain stimulation researchers to test and develop their protocols first on the peripheral nervous system—if you can’t make your protocol work well there, what hope do you have when stimulating the brain?

Neurostimulation

3.3 Scalp Test and Control Site Localization

469

Here, we will discuss how to localize the test site and an appropriate control site, with a focus on stimulation of the somatosensory cortex. Several approaches have been used to identify the location on the scalp to target a specific brain area, both for TES and TMS. When the location is an area beyond M1, including for S1, the task is quite complex because of the lack of direct and objective consequences of S1 stimulation on a trial-by-trial basis. However, although some studies have reported that TMS over S1 may elicit sensations, such phenomena have not received systematic investigation [49, 96–99]. It is therefore difficult for the researcher to establish whether the coil is correctly positioned to appropriately affect S1 activity as desired. Identifying the correct site of stimulation on the scalp is a fundamental prerequisite that will affect all subsequent steps of the experimental procedure. We will briefly describe the three most common approaches used to overcome this problem and identify S1 on the scalp (for a systematic review on this topic, see [27]). A widely used approach to target the hand area of S1 is a heuristic that consists in finding the location of a particular representation in M1 (estimated through MEP responses in a hand muscle) and then moving the coil posteriorly by about 2 cm (i.e., in a range between 1 and 4 cm) and/or until a motor response (i.e., MEP) is no longer detected [27]. This strategy is problematic for several reasons. First, given that S1 and M1 are anatomically contiguous in the brain, moving away from M1 implies that researchers are likely also moving away from S1. That said, it is generally very difficult to stimulate S1 and M1 independently. Activity of S1 directly affects M1 and vice versa [38, 68] given that the two systems communicate via a network of extensive connections [100–105] and motor cortex cells respond directly to sensory stimuli [106–109] as sensory cortex cells can control motor behavior [110]. In order to mitigate such circumstances, it may be beneficial to monitor the MEP responses and to test whether it correlates with the effects of TMS on tactile perception [33]. You may also be able to use different coil orientations to stimulate S1 as compared to M1 [111, 112]. Indeed, avoiding M1 stimulation may significantly reduce the probability of directly stimulating S1. Second, and critically, the hand area of S1 is not posterior to the hand area of M1, it is approximately 2 cm lateral [27, 33, 37]. A second localization approach, especially in the earliest TMS studies that stimulated S1, consists in positioning the coil over the C3/C4 electrode in the 10:20 system coordinates [23, 113–115], or moving posteriorly halfway between C3/C4 and P3/P4, to CP3/CP4, or moving ~2 cm posterior to C3’/C4’. The rationale behind this specific positioning can be traced back to earlier evidence suggesting that the C3/C4 location lay approximately over the central sulcus, or slightly posteriorly [113, 116]. However, different studies have attributed a different anatomical location to the C3/C4 position, either that of the S1 or M1 hand area [42, 96,

470

Luigi Tame` and Nicholas Paul Holmes

Fig. 6 Mean scalp locations for M1-FDI (red) and S1-index (magenta) locations, as used by Tame`, Holmes, and colleagues [23, 33, 37]. Also shown are scalp landmarks used in the 10:20 electrode positioning system. White circle: vertex, Cz; green circles: other scalp landmarks and electrode positions, including C3’, often positioned as indicated, at 2 cm posterior to C3, although C3’ is also often described as halfway between C3 and P3, which is likely ~3.6 cm posterior to C3

117, 118]. Holmes and Tame` [27] systematically reviewed this literature, and it seems that there was no general agreement among researchers regarding the actual location of S1 and M1 hand area relative to the C3/C4 electrodes on the scalp. Therefore, this strategy also seems not to be an optimal approach to correctly identify S1 at the scalp level. Systematic review and meta-analysis of all the available brain imaging and neuronavigation evidence, however, is much less ambiguous [37]. A third strategy that can be used to localize S1 is to take advantage of anatomical and functional data coming from structural and functional magnetic resonance (fMRI) imaging. In particular, some studies have used a single, standard head and brain template scan and registered each participant’s head to the template [119]. Others have used structural scans of each participant (this reduces the risk of misplacing the coil due to inter-subject variability) [120], and yet other studies used individual structural scans in combination with individual functional MRI data [33, 121]. This last approach arguably provides the best estimation of the S1 location (see Fig. 6). Overall, we strongly recommend using—and reporting—all available sources of evidence to target the S1 location optimally: accurate head and scalp measurements to find the vertex and C3/C4 locations; functional localizers using TMS to locate the optimal M1-hand area; MRI data where available, ideally both individual functional and anatomical scans. If neuronavigation is available, then a carefully coregistered scan of a head template may be sufficient. At the very least, the C3/C4 electrode location is, on average, very close to the representation of the index finger in S1 [27, 37].

Neurostimulation

471

Once the target brain area has been established, the next important step is to identify a control area to stimulate which is not involved in the processes we are aiming to study, yet at the same time is comparable in terms of the peripheral auditory, cutaneous, and deep sensations and movements induced by the stimulation (TMS or TES). Ideally, the participant will not be able to discriminate the different stimulation sites by how they feel on the head. For an appropriate choice, there are a few critical aspects that should be considered. We will start by discussing this in the context of TMS. First, despite the fact that newer TMS devices are becoming more and more precise in terms of localization and capability to selectively stimulate a certain brain area [32], TMS procedures can still be quite uncomfortable and even painful. Importantly, the level of annoyance and/or discomfort is not the same across different locations on the scalp. Instead, it varies strongly and systematically as a function of scalp location [122], as well as the intensity [123] and type of protocol with which TMS is applied. Typically, superior and posterior scalp locations are not associated with significant discomfort, whereas, frontal and temporal locations can cause moderate to high discomfort and pain [124]. It has been shown that the TMS side effects affect some aspects of task performance, such as participants’ accuracy at the task [125] as well as their reaction times [124]. When selecting a control site, it is therefore critical to find a location that, in addition to being suitable for the process that is studied, is also comparable in terms of side effects with the target site (see Note 4.6). In particular, when participants are performing a tactile task, it is critical to monitor and balance the tactile sensations generated by the TMS or TES on the scalp surface of different sites (see Note 4.7). Recently, Meteyard and Holmes [122] developed a “scalp mapping of annoyance ratings and twitches” caused by TMS in humans (TMS-SMART, [122]). They systematically mapped the degree of disturbance caused by TMS at 43 locations across the scalp. Participants were asked to perform a choice reaction time task while TMS was applied on different locations. After each of the five trials of the task per TMS coil location and orientation, they provided ratings of annoyance, pain, and muscle twitches. The resulting maps can be a very useful tool to assess the best matching location of a control site for any given target site in a study. The authors provided the data as an online resource that can be freely accessed and used (https://tms-smart.info). To ensure that the experimental results derive from modulation in the brain active caused by the TMS or TES and not simply because of peripheral side-effects [3, 126, 127], additional types of control are also available. For instance, the stimulation location could stay the same, but additional behavioral tasks and conditions can provide a task-based control. The least useful or powerful control condition is

472

Luigi Tame` and Nicholas Paul Holmes

one in which TMS is either not present, or a “sham” form of TMS is given. Such sham TMS needs to be proven to be effective on a location-by-location basis before it can be taken as a strong control condition. TMS over the vertex is commonly used, but this can only be a good control for other midline coil locations that have almost no side-effects (i.e., for very few locations). One final factor that should be considered when stimulating the somatosensory cortex with TMS or TES is the feasibility of reaching the desired body part’s representation in the somatosensory cortex. As we have seen from the mapping studies described in the first section of this chapter, the representation of body parts in the homunculus follows a topographic distribution over primary somatosensory cortex. Such organization makes some body parts easily accessible in terms of the capability of TMS to interfere with the relevant cortical activity, and some others more difficult to access. For instance, body parts such as the genitals, feet, and toes, at least according to the traditional homunculus, may almost be out of reach, at least with the current TMS devices due to their anatomical location in S1. Moreover, other sensory areas, such as the secondary somatosensory cortex (S2), may also not be easy to stimulate, though there are some examples in the literature in which stimulation of this area has been done [85, 128–130].

4

Notes In this section, we will describe some methodological advice that can be useful to refer to when conducting somatosensory TES and TMS experiments.

4.1 Interactions Between Neurostimulators and Tactile Stimulators

Neurostimulation devices produce large transient electromagnetic impulses (TMS) or constant or time-varying electrical currents (TES). These electromagnetic disturbances can travel relatively unimpeded around the lab. Just this week, we set up a MagStim BiStim TMS system about 30 cm from a National Instruments Card, which was passing an analogue signal to an Oticon vibrotactile stimulator via a small car stereo amplifier. Two TMS pulses were presented on every trial of the experiment, along with a vibrotactile target in half of the trials. The participant (NH) could not feel any of the target stimuli and instead only felt the vibrator click with every TMS pulse (confirmed by holding the vibrator to his ear). The solution was to move the TMS system to the other side of the laboratory (~2 m away), away from whichever sensitive electrical point of our setup (likely in the NI Card connector interface) was generating the artifact. We strongly advise researchers to assume that nothing in your setup works correctly until you can prove it! Critically, comparing a vibrotactile task with TMS or TES to a task that has no TMS or TES

Neurostimulation

473

(or when the neurostimulation system is away from the body and less likely to produce electrical artifacts on the body) is a very poor way to ensure that your experiment is working correctly. Make sure to rule out all potential artifacts as explanations for any apparent interference in tactile perception. It is not your reviewers’ or readers’ responsibility to control your experiments—it is yours. 4.2 Choosing a Tactile Task

In the experiment referred to, we deliberately made the task complex—the standard frequency was 200 Hz and presented on the index finger, and the comparison was higher or lower and presented on the middle finger. This task required participants to attend to two fingers, in sequence, and to focus on vibration frequency differences. We chose a complex design because two earlier experiments in this series, both using a two-interval task, had failed to find any effects of TMS over S1 on tactile detection. By adding both frequency and location components to the discrimination task, we hoped that S1 would be more involved in the task, and that TMS would then be effective. This hope was successful, although we do not know which task component was critical. In our experience, neurostimulation studies need to be designed to be as powerful as possible on the first attempt (e.g., high intensity TMS, difficult task, strong control conditions, accurate neuronavigation). First, because neurostimulation can be uncomfortable and comes with some risk for our participants—it is our duty to make the experiments as good as they can be. Second, because neurostimulation studies often just don’t work. By attempting to maximize the involvement of a brain area, and by stimulating that brain area precisely, both in location, and at the time it is likely to be involved in the task, we can improve our chances of learning something— anything—from the study.

4.3 Neurostimulation Likely Changes Participants’ Decision Criterion

When estimating the effect of the TMS applied on the somatosensory cortex, the main measure is participants’ task performance and possibly reaction time. When participants undergo neurostimulation, in addition to the well-known perceptual changes that the intervention can generate, there may be also changes in the criterion used to respond. Indeed, it may be that the simple presence on the scalp of the TMS or TES is altering the decision criterion that participants decide to adopt in responding to the task—e.g., becoming more conservative or liberal in deciding whether a stimulus has been felt or not. Therefore, the outcome of participants’ performance may well reflect a combination of both their perceptual sensitivity and their decision criteria used to respond. To overcome this limitation, we suggest whenever possible to adopt an approach that considers such effects. A useful approach is to apply signal detection theory to analyse the data that allows an estimation of both potential changes in perceptual sensitivity and decision criterion.

474

Luigi Tame` and Nicholas Paul Holmes

4.4 Neurostimulation in Relation with Motor Evoked Responses

Determining the motor threshold is an important component in any TMS experiment, as this parameter is often used to determine the intensity of the neurostimulation. Therefore, care should be taken to assess threshold in an efficient and effective manner. The excitability of the corticospinal tract is often taken as an index of the excitability also for stimulating other brain areas and pathways. This is a sensible thing to do and a universally accepted protocol, however, excitability of other brain areas may not be the same as the motor cortex. At least, it may provide a good general estimate of scalp-to-brain distance for your participant. Such an approach is probably primarily dictated by the fact that the motor cortex is the only brain region in which it is possible to have a direct and relatively simple physiological read out of excitability in the MEP. As specified elsewhere in the chapter, when stimulating S1 it is recommended to monitor the output from motor cortex responses during TMS sessions, as S1 stimulation can cause direct and/or indirect changes in motor excitability. This is particularly critical when a brain region away from S1 and/or M1 is used as a control site (e.g., SMG), as such regions will likely affect motor cortex excitability to a lesser extent. MEP variability should also be monitored across sessions if the experiment is performed in several sessions across days, weeks, or months. When stimulating M1, it is generally recommended to orient the TMS coil so that the induced current runs posterior-to-anterior at about 45 degrees to the midline (i.e., perpendicular to the central sulcus). However, for S1 stimulation of the fingers (i.e., index and middle fingers), we found that an orientation of 90 degrees seems to be better in terms of producing fewer MEPs and fewer side effects (e.g., jaw twitching) whilst also being effective in interfering with task performance, and by inference, S1 activity. Note that this orientation (90 degrees) differs from what is typically recommended for the stimulation of other locations (brain areas) on the scalp.

4.5 Neurostimulation and the Participant Sample Size

Similar to other psychophysical experiments, it is critical to have a comparable number of participants tested for the control and experimental group/conditions, also enough to get good counter-balancing, but we have found that sometimes more is worse. As participants are recruited for neurostimulation studies, we find—anecdotally—that well-motivated, or “good” participants often turn up early during recruitment. As the study continues and the recruitment pool of participants willing to undergo neurostimulation decreases, it becomes harder to find experienced, motivated participants who are comfortable with neurostimulation techniques. There may be, therefore, a trade-off between sample size and data quality—uncomfortable participants do not perform somatosensory tasks well. In theory, statistical power increases proportional to the square-root of the sample size (i.e., to double power, you need to quadruple the sample). We suspect that this

Neurostimulation

475

theoretical law of diminishing returns is even worse in practice because better participants tend to be recruited earlier. In general, we recommend improving the statistical power of your study by focusing on increasing the effect size—more trials, better controls, stronger stimulation—than on the sample size. Of course, you won’t know if this strategy has worked until you run a wellpowered replication of any of the effects that you find! Counterbalancing is particularly important for short experiments, or those with small sample sizes. It is also a good practice to always consider the participants’ perspective on the task. Indeed, we should be aware that they can be anxious and not familiar with neuroscience methods, and that we are stimulating their brains directly. We have found several participants arriving at the experiment without having read the Information Sheet, seeming surprised that their brains will be stimulated. These participants did not complete the study. Particular care must be taken in making participants comfortable before, during, and after neurostimulation. This can be done by describing to them the different phases of the experiment as well as the potential sensations and discomfort that these techniques may cause, in a calm and sensible manner. The experimenter may need to stop an experiment for “technical reasons” when their participant is clearly not comfortable with the procedures, even if they state verbally that they are happy to continue. 4.6 Neurostimulation Control Locations for Somatosensory Stimulation

A great challenge to apply neurostimulation on the somatosensory cortex (e.g., S1 and S2) is to correctly identify the target site (e.g., S1) and a comparable control site on another region of the brain. The target location is challenging as there is no direct physiological read out when stimulating the somatosensory cortex. The best practice is to use a neuronavigation device that can provide anatomical and possibly functional information of the correct location to reduce the probability of site errors [137]. As a control site, an area that is not involved in the perceptual and/or cognitive process under study should be identified, also assuring a similar grade of side effects for the target and control site [121]. The vertex should be avoided as a potential control site given the absence of major side effects or cutaneous sensations generated from stimulating that location. Remember that neuronavigation is only as good as the images and human operators that are used with it. We find that the only way to minimize error in coil positioning is to take multiple head measurements, to record everything, and always to be wary of human error. Measure twice, stimulate once!

4.7 Neurostimulation as a Somatosensory Stimulus

Remember that TES and TMS are themselves multisensory stimuli—the electrical and magnetic field can induce current in muscles, nerves, and receptors [138]. They may contract muscles, stimulate free nerve endings and peripheral nerves, contract blood

476

Luigi Tame` and Nicholas Paul Holmes

vessels. TMS also creates a loud sound. When close to the eyes, retinal stimulation may directly produce flashes of light; when close to the ears, muscular contractions may move the eardrums and the Eustachian tube (personal experience!). If all of these side-effects are matched by location- or task-based control conditions, a clean and simple experimental design may be possible. But, specifically in somatosensation, the peripheral side effects of stimulation may act as masking stimuli, interfering with touch perception not due to stimulation of the brain but to stimulation of the skin, nerves, and muscles. Note that there have been attempts to reduce scalp discomfort and tactile sensations by treating the locus of stimulation before the TMS was applied, using topical lidocaine or prilocaine cream [10], although this is not a common practice. Did TMS interfere with somatosensory perception only because of its effect on S1, or also due to its effect on the scalp? It is each researcher’s task to answer this question and rule out trivial confounds.

References 1. Jahanshahi M, Rothwell J (2000) Transcranial magnetic stimulation studies of cognition: an emerging field. Exp Brain Res 131:1–9. https://doi.org/10.1007/s002219900224 2. Kuo M-F, Nitsche MA (2012) Effects of transcranial electrical stimulation on cognition. Clin EEG Neurosci 43:192–199. https:// doi.org/10.1177/1550059412444975 3. Sandrini M, Umilta` C, Rusconi E (2011) The use of transcranial magnetic stimulation in cognitive neuroscience: a new synthesis of methodological issues. Neurosci Biobehav Rev 35:516–536. https://doi.org/10.1016/ j.neubiorev.2010.06.005 4. Allman C, Amadi U, Winkler AM et al (2016) Ipsilesional anodal tDCS enhances the functional benefits of rehabilitation in patients after stroke. Sci Transl Med 8. https://doi. org/10.1126/scitranslmed.aad5651 5. O’Shea J, Revol P, Cousijn H et al (2017) Induced sensorimotor cortex plasticity remediates chronic treatment-resistant visual neglect. elife 6. https://doi.org/10.7554/ eLife.26602 ˜ or IM et al 6. Shiozawa P, Fregni F, Bensen (2014) Transcranial direct current stimulation for major depression: an updated systematic review and meta-analysis. Int J Neuropsychopharmacol 17:1443–1452. https://doi.org/ 10.1017/S1461145714000418 7. Fertonani A, Miniussi C (2017) Transcranial Electrical Stimulation. Neurosci 23:109–123.

h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 1073858416631966 8. Miniussi C, Ruzzoli M, Walsh V (2010) The mechanism of transcranial magnetic stimulation in cognition. Cortex 46:128–130. https://doi.org/10.1016/j.cortex.2009. 03.004 9. Rossini PM, Burke D, Chen R et al (2015) Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots and peripheral nerves: basic principles and procedures for routine clinical and research application. An updated report from an I.F.C.N. Committee. Clin Neurophysiol 126:1071–1107. https://doi.org/10.1016/ j.clinph.2015.02.001 10. Lenoir C, Algoet M, Vanderclausen C et al (2018) Report of one confirmed generalized seizure and one suspected partial seizure induced by deep continuous theta burst stimulation of the right operculo-insular cortex. Brain Stimul 11:1187–1188. https://doi. org/10.1016/j.brs.2018.05.004 11. Reader A, Tame` L, Holmes NP (2017) Two instances of presyncope during magnetic stimulation of the median nerve, and evaluation of resting motor threshold with transcranial magnetic stimulation (TMS) 12. Fritsch G, Hitzig D (1870) On the electrical excitability of the cerebrum. In: Some papers on the cerebral cortex. Charles C. Thomas, Springfield

Neurostimulation 13. Cushing H (1909) A note upon the faradic stimulation of the postcentral gyrus in conscious patients. Brain 32:44–53 14. Penfield W, Boldrey E (1937) Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 60:389–443. https://doi.org/10. 1093/brain/60.4.389 15. Penfield W, Rasmussen T (1950) The cerebral cortex of man; a clinical study of localization of function. Macmillan, New York 16. Penfield W, Jasper HH (1954) Epilepsy and the functional anatomy of the human brain. Little, Brown & Co, Boston 17. Liu A, Vo¨ro¨slakos M, Kronberg G et al (2018) Immediate neurophysiological effects of transcranial electrical stimulation. Nat Commun 9:5092. https://doi.org/10.1038/s41467018-07233-7 18. Guleyupoglu B, Schestatsky P, Edwards D et al (2013) Classification of methods in transcranial Electrical Stimulation (tES) and evolving strategy from historical approaches to contemporary innovations. J Neurosci Methods 219:297–311. https://doi.org/10. 1016/j.jneumeth.2013.07.016 19. D’Arsonval A (1896) Dispositifs pour la mesure des courants alternatifs de toutes frequences. Comput Rend Soc Biol 3:430–451 20. Kolin A, Brill NQ, Broberg PJ (1959) Stimulation of irritable tissues by means of an alternating magnetic field. Exp Biol Med 102: 251–253. https://doi.org/10.3181/ 00379727-102-25209 21. Bickford R, Freming B (1965) Neural stimulation by pulsed magnetic fields in animals and man. Dig 6th Int Conf Med Electron Biol Eng Japan. pp 7–6 22. Barker AT, Jalinous R, Freeston IL (1985) Non-invasive magnetic stimulation of human motor cortex. Lancet 325:1106–1107. https://doi.org/10.1016/S0140-6736(85) 92413-4 23. Cohen LG, Bandinelli S, Sato S et al (1991) Attenuation in detection of somatosensory stimuli by transcranial magnetic stimulation. Electroencephalogr Clin Neurophysiol 81: 366–376 24. Powell TP, Mountcastle VB (1959) Some aspects of the functional organization of the cortex of the postcentral gyrus of the monkey: a correlation of findings obtained in a single unit analysis with cytoarchitecture. Bull Johns Hopkins Hosp 105:133–162 25. Brodmann K (1909) Vergleichende lokalisationslehre der grosshirnrinde in ihrem prinzipien dargestellt auf grund des zellenbaues

477

[comparative theory of localization of the cerebral cortex presented in its principles on the basis of the cell structure]. Barth JA, Leipzig, Germany 26. Roux F, Djidjeli I, Durand J-B (2018) Functional architecture of the somatosensory homunculus detected by electrostimulation. J Physiol 596:941–956. https://doi.org/10. 1113/JP275243 27. Holmes NP, Tame` L (2019) Locating primary somatosensory cortex in human brain stimulation studies: systematic review and metaanalytic evidence. J Neurophysiol 121:152– 162. https://doi.org/10.1152/jn.00614. 2018 28. Novey W (2019) Are all rTMS machines equal? New research suggests there may be clinically significant differences. Ment Illn 11. https://doi.org/10.4081/mi.2019.8125 29. Van Doren J, Langguth B, Schecklmann M (2015) TMS-related potentials and artifacts in combined TMS-EEG measurements: comparison of three different TMS devices. Neurophysiol Clin Neurophysiol 45:159–166. https://doi.org/10.1016/j.neucli.2015. 02.002 30. Koponen LM, Goetz SM, Tucci DL, Peterchev AV (2020) Sound comparison of seven TMS coils at matched stimulation strength. Brain Stimul 13:873–880. https:// doi.org/10.1016/j.brs.2020.03.004 31. Rossini PM, Barker AT, Berardelli A et al (1994) Non-invasive electrical and magnetic stimulation of the brain, spinal cord and roots: basic principles and procedures for routine clinical application. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol 91:79–92 32. Rossi S, Antal A, Bestmann S et al (2021) Safety and recommendations for TMS use in healthy subjects and patient populations, with updates on training, ethical and regulatory issues: expert guidelines. Clin Neurophysiol 132:269–306. https://doi.org/10.1016/j. clinph.2020.10.003 33. Tame` L, Holmes NP (2016) Involvement of human primary somatosensory cortex in vibrotactile detection depends on task demand. NeuroImage 138:184–196. https://doi.org/10.1016/j.neuroimage. 2016.05.056 34. Ruohonen J, Karhu J (2010) Navigated transcranial magnetic stimulation. Neurophysiol Clin Neurophysiol 40:7–17. https://doi. org/10.1016/j.neucli.2010.01.006 35. Souza VH, Matsuda RH, Peres ASC et al (2018) Development and characterization of

478

Luigi Tame` and Nicholas Paul Holmes

the InVesalius navigator software for navigated transcranial magnetic stimulation. J Neurosci Methods 309:109–120. https:// doi.org/10.1016/j.jneumeth.2018.08.023 36. Amorim P, Moraes T, Silva J, Pedrini H (2015) InVesalius: an interactive rendering framework for health care support, p 45–54 37. Holmes NP, Tame` L, Beeching P et al (2019) Locating primary somatosensory cortex in human brain stimulation studies: experimental evidence. J Neurophysiol 121:336–344. https://doi.org/10.1152/jn.00641.2018 38. Tame` L, Pavani F, Braun C et al (2015) Somatotopy and temporal dynamics of sensorimotor interactions: evidence from double afferent inhibition. Eur J Neurosci 41:1459– 1465. https://doi.org/10.1111/ejn.12890 39. Tame` L, Pavani F, Papadelis C et al (2015) Early integration of bilateral touch in the primary somatosensory cortex. Hum Brain Mapp 36:1506–1523. https://doi.org/10. 1002/hbm.22719 40. von Be´ke´sy G (1967) Sensory inhibition. Princepton University Press 41. Rusconi E, Tame` L, Furlan M et al (2014) Neural correlates of finger gnosis. J Neurosci 34. https://doi.org/10.1523/JNEUROSCI. 3119-13.2014 42. McKay DR, Ridding MC, Miles TS (2003) Magnetic stimulation of motor and somatosensory cortices suppresses perception of ulnar nerve stimuli. Int J Psychophysiol 48: 25–33 43. Vroomen J, Stekelenburg JJ (2014) A biasfree two-alternative forced choice procedure to examine intersensory illusions applied to the ventriloquist effect by flashes and averted eye-gazes. Eur J Neurosci. https://doi.org/ 10.1111/ejn.12525 44. Yeshurun Y, Carrasco M, Maloney LT (2008) Bias and sensitivity in two-interval forced choice procedures: tests of the difference model. Vis Res 48:1837–1851. https://doi. org/10.1016/j.visres.2008.05.008 45. Harris JA, Karlov L, Clifford CWG (2006) Localization of tactile stimuli depends on conscious detection. J Neurosci 26:948–952 46. Seyal M, Siddiqui I, Hundal NS (1997) Suppression of spatial localization of a cutaneous stimulus following transcranial magnetic pulse stimulation of the sensorimotor cortex. Electroencephalogr Clin Neurophysiol 105:24– 28 47. Saito K, Otsuru N, Inukai Y et al (2019) Comparison of transcranial electrical stimulation regimens for effects on inhibitory circuit activity in primary somatosensory cortex and

tactile spatial discrimination performance. Behav Brain Res 375:112168. https://doi. org/10.1016/j.bbr.2019.112168 48. Knecht S, Ellger T, Breitenstein C et al (2003) Changing cortical excitability with low-frequency transcranial magnetic stimulation can induce sustained disruption of tactile perception. Biol Psychiatry 53:175–179 49. Tegenthoff M, Ragert P, Pleger B et al (2005) Improvement of tactile discrimination performance and enlargement of cortical somatosensory maps after 5 Hz rTMS. PLoS Biol 3: e362. https://doi.org/10.1371/journal. pbio.0030362 50. Porro CA, Martinig M, Facchin P et al (2007) Parietal cortex involvement in the localization of tactile and noxious mechanical stimuli: a transcranial magnetic stimulation study. Behav Brain Res 178:183–189. https://doi. org/10.1016/j.bbr.2006.11.011 51. Hannula H, Neuvonen T, Savolainen P et al (2008) Navigated transcranial magnetic stimulation of the primary somatosensory cortex impairs perceptual processing of tactile temporal discrimination. Neurosci Lett 437:144– 147. https://doi.org/10.1016/j.neulet. 2008.03.093 52. Bolognini N, Cecchetto C, Geraci C et al (2012) Hearing shapes our perception of time: temporal discrimination of tactile stimuli in deaf people. J Cogn Neurosci 24:276– 286. https://doi.org/10.1162/jocn_a_ 00135 53. Master S, Tremblay F (2009) Task-specific increase in corticomotor excitability during tactile discrimination. Exp Brain Res 194: 163–172. https://doi.org/10.1007/ s00221-008-1679-z 54. Pascual-Leone A, Walsh V, Rothwell J (2000) Transcranial magnetic stimulation in cognitive neuroscience--virtual lesion, chronometry, and functional connectivity. Curr Opin Neurobiol 10:232–237 55. Konen CS, Haggard P (2014) Multisensory parietal cortex contributes to visual enhancement of touch in humans: a single-pulse TMS study. Cereb Cortex 24:501–507. https:// doi.org/10.1093/cercor/bhs331 56. Poliakoff E, Spence C, O’Boyle DJ et al (2002) Tactile inhibition of return: non-ocular response inhibition and mode of response. Exp Brain Res 146:54–59. https:// doi.org/10.1007/s00221-002-1153-2 57. Tame` L, Dransfield E, Quettier T, Longo MR (2017) Finger posture modulates structural body representations. Sci Rep 7:43019. https://doi.org/10.1038/srep43019

Neurostimulation 58. Tame` L, Braun C, Holmes NP et al (2016) Bilateral representations of touch in the primary somatosensory cortex. Cogn Neuropsychol 1–19. https://doi.org/10.1080/ 02643294.2016.1159547 59. O’Shea J, Taylor PCJ, Rushworth MFS (2008) Imaging causal interactions during sensorimotor processing. Cortex 44:598– 608. https://doi.org/10.1016/j.cortex. 2007.08.012 60. Tamburin S, Fiaschi A, Andreoli A et al (2005) Sensorimotor integration to cutaneous afferents in humans: the effect of the size of the receptive field. Exp Brain Res 167:362– 369. https://doi.org/10.1007/s00221-0050041-y 61. Miniussi C, Paulus W, Rossini PM (2013) Transcranial brain stimulation. CRC Press, Boca Raton 62. Delwaide PJ, Olivier E (1990) Conditioning transcranial cortical stimulation (TCCS) by exteroceptive stimulation in parkinsonian patients. Adv Neurol 53:175–181 63. Chen R, Corwell B, Hallett M (1999) Modulation of motor cortex excitability by median nerve and digit stimulation. Exp Brain Res 129:77–86 64. Tokimura H, Di Lazzaro V, Tokimura Y et al (2000) Short latency inhibition of human hand motor cortex by somatosensory input from the hand. J Physiol 523(Pt 2):503–513 65. Abbruzzese G, Marchese R, Buccolieri A et al (2001) Abnormalities of sensorimotor integration in focal dystonia: a transcranial magnetic stimulation study. Brain 124:537–545 66. Evarts EV (1973) Motor cortex reflexes associated with learned movement. Science (80- ) 179:501–503 67. Mariorenzi R, Zarola F, Caramia MD et al (1991) Non-invasive evaluation of central motor tract excitability changes following peripheral nerve stimulation in healthy humans. Electroencephalogr Clin Neurophysiol 81:90–101 68. Turco CV, El-Sayes J, Savoie MJ et al (2018) Short- and long-latency afferent inhibition; uses, mechanisms and influencing factors. Brain Stimul 11:59–74. https://doi.org/10. 1016/j.brs.2017.09.009 69. Brown MJN, Weissbach A, Pauly MG et al (2019) Somatosensory-motor cortex interactions measured using dual-site transcranial magnetic stimulation. Brain Stimul 12: 1229–1243. https://doi.org/10.1016/j.brs. 2019.04.009 70. Ni Z, Gunraj C, Nelson AJ et al (2009) Two phases of interhemispheric inhibition between

479

motor related cortical areas and the primary motor cortex in human. Cereb Cortex 19: 1654–1665. https://doi.org/10.1093/ cercor/bhn201 71. Iwata Y, Jono Y, Mizusawa H et al (2016) Interhemispheric inhibition induced by transcranial magnetic stimulation over primary sensory cortex. Front Hum Neurosci 10. https://doi.org/10.3389/fnhum.2016. 00438 72. Sale MV, Ridding MC, Nordstrom MA (2007) Factors influencing the magnitude and reproducibility of corticomotor excitability changes induced by paired associative stimulation. Exp Brain Res 181:615–626. https://doi.org/10.1007/s00221-0070960-x 73. Holmes NP, Tame` L (2018) Multisensory perception: magnetic disruption of attention in human parietal lobe. Curr Biol 28:R259– R261. https://doi.org/10.1016/j.cub.2018. 01.078 74. Paulus W (2011) Transcranial electrical stimulation (tES – tDCS; tRNS, tACS) methods. Neuropsychol Rehabil 21:602–617. https:// doi.org/10.1080/09602011.2011.557292 75. Liebetanz D (2002) Pharmacological approach to the mechanisms of transcranial DC-stimulation-induced after-effects of human motor cortex excitability. Brain 125: 2238–2247. https://doi.org/10.1093/ brain/awf238 76. Li LM, Uehara K, Hanakawa T (2015) The contribution of interindividual factors to variability of response in transcranial direct current stimulation studies. Front Cell Neurosci 9. https://doi.org/10.3389/fncel. 2015.00181 77. Antal A, Alekseichuk I, Bikson M et al (2017) Low intensity transcranial electric stimulation: safety, ethical, legal regulatory and application guidelines. Clin Neurophysiol 128:1774– 1809. https://doi.org/10.1016/j.clinph. 2017.06.001 78. Fujimoto S, Yamaguchi T, Otaka Y et al (2014) Dual-hemisphere transcranial direct current stimulation improves performance in a tactile spatial discrimination task. Clin Neurophysiol 125:1669–1674. https://doi.org/ 10.1016/j.clinph.2013.12.100 79. Fujimoto S, Kon N, Otaka Y et al (2016) Transcranial direct current stimulation over the primary and secondary somatosensory cortices transiently improves tactile spatial discrimination in stroke patients. Front Neurosci 10. https://doi.org/10.3389/fnins.2016. 00128

480

Luigi Tame` and Nicholas Paul Holmes

80. Rogalewski A, Breitenstein C, Nitsche MA et al (2004) Transcranial direct current stimulation disrupts tactile perception. Eur J Neurosci 20:313–316. https://doi.org/10. 1111/j.0953-816X.2004.03450.x 81. Rehmann R, Sczesny-Kaiser M, Lenz M et al (2016) Polarity-specific cortical effects of transcranial direct current stimulation in primary somatosensory cortex of healthy humans. Front Hum Neurosci 10. https:// doi.org/10.3389/fnhum.2016.00208 82. Manzo N, Guerra A, Giangrosso M et al (2020) Investigating the effects of transcranial alternating current stimulation on primary somatosensory cortex. Sci Rep 10:17129. https://doi.org/10.1038/s41598-02074072-2 83. Rossi S, Hallett M, Rossini PM et al (2009) Safety, ethical considerations, and application guidelines for the use of transcranial magnetic stimulation in clinical practice and research. Clin Neurophysiol 120:2008–2039. https:// doi.org/10.1016/j.clinph.2009.08.016 84. Huang Y-Z, Edwards MJ, Rounis E et al (2005) Theta burst stimulation of the human motor cortex. Neuron 45:201–206. https://doi.org/10.1016/j.neuron.2004. 12.033 85. Bolognini N, Rossetti A, Maravita A, Miniussi C (2011) Seeing touch in the somatosensory cortex: a TMS study of the visual perception of touch. Hum Brain Mapp 32:2104–2114. https://doi.org/10.1002/hbm.21172 86. Vidoni ED, Acerra NE, Dao E et al (2010) Neurobiology of learning and memory role of the primary somatosensory cortex in motor learning : an rTMS study. Neurobiol Learn Mem 93:532–539. https://doi.org/10. 1016/j.nlm.2010.01.011 87. Blankenburg F, Ruff CC, Bestmann S et al (2008) Interhemispheric effect of parietal TMS on somatosensory response confirmed directly with concurrent TMS-fMRI. J Neurosci 28:13202–13208. https://doi.org/10. 1523/JNEUROSCI.3043-08.2008 88. Schneider SA, Pleger B, Draganski B et al (2010) Modulatory effects of 5Hz rTMS over the primary somatosensory cortex in focal dystonia-an fMRI-TMS study. Mov Disord 25:76–83. https://doi.org/10.1002/ mds.22825 89. Rogasch NC, Fitzgerald PB (2013) Assessing cortical network properties using TMS-EEG. Hum Brain Mapp 34:1652–1669. https:// doi.org/10.1002/hbm.22016 90. Tremblay S, Rogasch NC, Premoli I et al (2019) Clinical utility and prospective of

TMS–EEG. Clin Neurophysiol 130:802– 844. https://doi.org/10.1016/j.clinph. 2019.01.001 91. Pisoni A, Romero Lauro LJ, Vergallito A et al (2018) Cortical dynamics underpinning the self-other distinction of touch: a TMS-EEG study. NeuroImage 178:475–484. https:// doi.org/10.1016/j.neuroimage.2018. 05.078 92. Schu¨rmann M, Nikouline VV, Soljanlahti S et al (2001) EEG responses to combined somatosensory and transcranial magnetic stimulation. Clin Neurophysiol 112:19–24. https://doi.org/10.1016/S1388-2457(00) 00509-5 93. Noury N, Siegel M (2018) Analyzing EEG and MEG signals recorded during tES, a reply. NeuroImage 167:53–61. https://doi. org/10.1016/j.neuroimage.2017.11.023 94. Corniani G, Saal HP (2020) Tactile innervation densities across the whole body, J Neurophysiol 124:1229–1240 95. Olney RK, So YT, Goodin DS, Aminoff MJ (1990) A comparison of magnetic and electrical stimulation of peripheral nerves. Muscle Nerve 13:957–963. https://doi.org/10. 1002/mus.880131012 96. Feurra M, Paulus W, Walsh V, Kanai R (2011) Frequency specific modulation of human somatosensory cortex. Front Psychol 2. https://doi.org/10.3389/fpsyg.2011. 00013 97. Ragert P, Dinse HR, Pleger B et al (2003) Combination of 5 Hz repetitive transcranial magnetic stimulation (rTMS) and tactile coactivation boosts tactile discrimination in humans. Neurosci Lett 348:105–108. https://doi.org/10.1016/S0304-3940(03) 00745-6 98. Sugishita M, Takayama Y (1993) Paraesthesia elicited by repetitive magnetic stimulation of the postcentral gyrus. Neuroreport 4:569– 570. https://doi.org/10.1097/00001756199305000-00027 99. Amassian VE, Rothwell JC, Somasundaram M et al (1991) Paraesthesias are elicited by single pulse, magnetic coil stimulation of motor cortex in susceptible humans. Brain 114:2505–2520. https://doi.org/10.1093/ brain/114.6.2505 100. Hikosaka O, Tanaka M, Sakamoto M, Iwamura Y (1985) Deficits in manipulative behaviors induced by local injections of muscimol in the first somatosensory cortex of the conscious monkey. Brain Res 325:375–380 101. Asanuma H, Stoney SD, Abzug C (1968) Relationship between afferent input and

Neurostimulation motor outflow in cat motorsensory cortex. J Neurophysiol 31:670–681. https://doi.org/ 10.1152/jn.1968.31.5.670 102. Makris N, Kennedy DN, McInerney S et al (2005) Segmentation of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cereb Cortex 15:854–869. https:// doi.org/10.1093/cercor/bhh186 103. Huffman KJ (2001) Area 3a: Topographic organization and cortical connections in marmoset monkeys. Cereb Cortex 11:849–867. https://doi.org/10.1093/cercor/11.9.849 104. Catani M, Dellacqua F, Vergani F et al (2012) Short frontal lobe connections of the human brain. Cortex 48:273–291. https://doi.org/ 10.1016/j.cortex.2011.12.001 105. Mao T, Kusefoglu D, Hooks BM et al (2011) Long-range neuronal circuits underlying the interaction between sensory and motor cortex. Neuron 72:111–123. https://doi.org/ 10.1016/j.neuron.2011.07.029 106. Albe-Fessard D, Liebeskind J (1966) Origine des messages somato-sensitifs activant les cellules du cortex moteur chez le singe. Exp Brain Res 1:127–146. https://doi.org/10. 1007/BF00236866 107. Goldring S, Ratcheson R (1972) Human motor cortex: sensory input data from single neuron recordings. Science (80- ) 175:1493– 1495. https://doi.org/10.1126/science. 175.4029.1493 108. Fetz EE, Finocchio DV, Baker MA, Soso MJ (1980) Sensory and motor responses of precentral cortex cells during comparable passive and active joint movements. J Neurophysiol 43:1070–1089. https://doi.org/10.1152/ jn.1980.43.4.1070 109. Fromm C, Wise SP, Evarts EV (1984) Sensory response properties of pyramidal tract neurons in the precentral motor cortex and postcentral gyrus of the rhesus monkey. Exp Brain Res 54:177–185. https://doi.org/10. 1007/BF00235829 110. Matyas F, Sreenivasan V, Marbach F et al (2010) Motor control by sensory cortex. Science (80- ) 330:1240–1243. https://doi. org/10.1126/science.1195797 111. Pascual-Leone A, Cohen LG, Brasil-Neto JP et al (1994) Differentiation of sensorimotor neuronal structures responsible for induction of motor evoked potentials, attenuation in detection of somatosensory stimuli, and induction of sensation of movement by mapping of optimal current directions. Electroencephalogr Clin Neurophysiol 93:230– 236

481

112. Raffin E, Pellegrino G, Di Lazzaro V et al (2015) Bringing transcranial mapping into shape: Sulcus-aligned mapping captures motor somatotopy in human primary motor hand area. NeuroImage 120:164–175. https://doi.org/10.1016/j.neuroimage. 2015.07.024 113. Jasper HH (1958) Report of the committee on methods of clinical examination in electroencephalography. Electroencephalogr Clin Neurophysiol 10:370–375. https://doi.org/ 10.1016/0013-4694(58)90053-1 114. Seyal M, Masuoka LK, Browne JK (1992) Suppression of cutaneous perception by magnetic pulse stimulation of the human brain. Electroencephalogr Clin Neurophysiol 85: 397–401 115. Seyal M, Browne JK, Masuoka LK, Gabor AJ (1993) Enhancement of the amplitude of somatosensory evoked potentials following magnetic pulse stimulation of the human brain. Electroencephalogr Clin Neurophysiol Potentials Sect 88:20–27. https://doi.org/ 10.1016/0168-5597(93)90024-J ˜ os J, Suarez D et al (1993) 116. Towle VL, Bolan The spatial location of EEG electrodes: locating the best-fitting sphere relative to cortical anatomy. Electroencephalogr Clin Neurophysiol 86:1–6. https://doi.org/10.1016/ 0013-4694(93)90061-Y 117. Restuccia D, Ulivelli M, De Capua A et al (2007) Modulation of high-frequency (600 Hz) somatosensory-evoked potentials after rTMS of the primary sensory cortex. Eur J Neurosci 26:2349–2358. https://doi. org/10.1111/j.1460-9568.2007.05828.x 118. Fiorio M, Haggard P (2005) Viewing the body prepares the brain for touch: effects of TMS over somatosensory cortex. Eur J Neurosci 22:773–777. https://doi.org/10. 1111/j.1460-9568.2005.04267.x 119. Ruzzoli M, Soto-Faraco S (2014) Alpha stimulation of the human parietal cortex attunes tactile perception to external space. Curr Biol 24:329–332. https://doi.org/10.1016/j. cub.2013.12.029 120. Romaigue`re P, Calvin S, Roll J-P (2005) Transcranial magnetic stimulation of the sensorimotor cortex alters kinaesthesia. Neuroreport 16:693–697. https://doi.org/10. 1097/00001756-200505120-00008 ´ urcˇic´-Blake B, Renken RJ et al 121. Valchev N, C (2015) cTBS delivered to the left somatosensory cortex changes its functional connectivity during rest. NeuroImage 114:386–397. https://doi.org/10.1016/j.neuroimage. 2015.04.017

482

Luigi Tame` and Nicholas Paul Holmes

122. Meteyard L, Holmes NP (2018) TMS SMART – scalp mapping of annoyance ratings and twitches caused by transcranial magnetic stimulation. J Neurosci Methods 299:34–44. https://doi.org/10.1016/j.jneumeth.2018. 02.008 123. Bestmann S, Ruff CC, Blankenburg F et al (2008) Mapping causal interregional influences with concurrent TMS–fMRI. Exp Brain Res 191:383–402. https://doi.org/ 10.1007/s00221-008-1601-8 124. Holmes NP, Meteyard L (2018) Subjective discomfort of TMS predicts reaction times differences in published studies. Front Psychol 9. https://doi.org/10.3389/fpsyg. 2018.01989 125. Abler B, Walter H, Wunderlich A et al (2005) Side effects of transcranial magnetic stimulation biased task performance in a cognitive neuroscience study. Brain Topogr 17:193– 196. https://doi.org/10.1007/s10548005-6028-y 126. Arana AB, Borckardt JJ, Ricci R et al (2008) Focal electrical stimulation as a sham control for repetitive transcranial magnetic stimulation: does it truly mimic the cutaneous sensation and pain of active prefrontal repetitive transcranial magnetic stimulation? Brain Stimul 1:44–51. https://doi.org/10.1016/j. brs.2007.08.006 127. Duecker F, Sack AT (2013) Pre-stimulus sham TMS facilitates target detection. PLoS One 8:e57765. https://doi.org/10.1371/ journal.pone.0057765 128. Lockwood PL, Iannetti GD, Haggard P (2013) Transcranial magnetic stimulation over human secondary somatosensory cortex disrupts perception of pain intensity. Cortex 49:2201–2209. https://doi.org/10.1016/j. cortex.2012.10.006 129. Fabri M, Del Pesce M, Paggi A et al (2005) Contribution of posterior corpus callosum to the interhemispheric transfer of tactile information. Brain Res Cogn Brain Res 24:73–80 130. Case LK, Laubacher CM, Richards EA et al (2017) Inhibitory rTMS of secondary somatosensory cortex reduces intensity but not pleasantness of gentle touch. Neurosci Lett 653:84–91. https://doi.org/10.1016/ j.neulet.2017.05.006 131. Silvanto J, Cattaneo Z (2017) Common framework for “virtual lesion” and state-

dependent TMS: the facilitatory/suppressive range model of online TMS effects on behavior. Brain Cogn 119:32–38. https://doi.org/ 10.1016/j.bandc.2017.09.007 132. Harris JA, Miniussi C, Harris IM, Diamond ME (2002) Transient storage of a tactile memory trace in primary somatosensory cortex. J Neurosci 22:8720–8725 133. Sliva DD, Black CJ, Bowary P et al (2018) A prospective study of the impact of transcranial alternating current stimulation on EEG correlates of somatosensory perception. Front Psychol 9. https://doi.org/10.3389/fpsyg. 2018.02117 134. Bowling NC, Banissy MJ (2017) Modulating vicarious tactile perception with transcranial electrical current stimulation. Eur J Neurosci 46:2355–2364. https://doi.org/10.1111/ ejn.13699 135. Valchev N, Tidoni E, de Hamilton AFC et al (2017) Primary somatosensory cortex necessary for the perception of weight from other people’s action: a continuous theta-burst TMS experiment. NeuroImage 152:195– 2 0 6 . h t t p s : // d o i . o r g / 1 0 . 1 0 1 6 / j . neuroimage.2017.02.075 136. Antal A, Brepohl N, Poreisz C et al (2008) Transcranial direct current stimulation over somatosensory cortex decreases experimentallyinduced acute pain perception. Clin J Pain 24:56–63. https://doi.org/10.1097/ AJP.0b013e318157233b 137. Hilgenstock R, Weiss T, Huonker R, Witte OW (2016) Behavioural and neurofunctional impact of transcranial direct current stimulation on somatosensory learning. Hum Brain Mapp 37:1277–1295. https://doi.org/10. 1002/hbm.23101 138. Hurme M, Railo H (2022) Promise and challenges for discovering transcranial magnetic stimulation induced “numbsense”—commentary on Ro Koenig (2021). Conscious Cogn 98:103265. https://doi.org/10. 1016/j.concog.2021.103265 139. Siebner HR, Funke K, Aberra AS et al (2022) Transcranial magnetic stimulation of the brain: what is stimulated? – a consensus and critical position paper. Clin Neurophysiol. https://doi.org/10.1016/j.clinph.2022. 04.022

INDEX A Abnormalities .................................................96, 227, 228 Activation................................................65, 67, 110, 112, 120, 125, 131–133, 135, 139, 144, 169, 174, 202, 203, 205–208, 210–212, 215–218, 292, 294–296, 312, 345, 400, 402, 409–411, 414–417, 421, 431, 432, 434–438, 441–443 Acuity ......................................... 5–7, 56, 63, 95, 96, 102, 251, 252, 254, 258, 288–292, 295, 441 Acute itch .................................................... 162, 168, 169 Affect..................................... 26, 65, 112, 113, 117, 136, 149, 209, 213, 215, 234, 239, 243–245, 252, 253, 256, 259, 261, 271, 392, 407, 419, 436, 452, 454, 456, 459, 462–465, 468, 469, 471, 474 Affective touch ..................................... 87, 110–113, 117, 121, 122, 200–202, 204, 207–215, 217–219, 228, 230, 231, 233, 238, 244, 417, 439, 443, 444 Afferent fibres.................................................................. 96 Age ......................... 16, 41, 65, 122, 136, 140, 188, 194, 209, 214, 231, 237, 241, 251–254, 258, 382, 441 Anisotropy ................................................. 95, 96, 98–102 Apparent motion ....................71–73, 75, 76, 87, 91, 280 Assessment........................................56, 63, 64, 130–133, 136, 138, 139, 141, 143, 144, 146, 152, 163, 167, 183, 184, 191, 208, 217, 219, 230, 234, 236, 237, 241, 242, 253, 256, 261, 277, 297, 299 Attention ...................................................... 3, 18, 24, 30, 49, 79, 82, 88, 122, 126, 134, 136, 144, 148, 150, 152, 167, 173, 178, 202, 210, 229, 231, 238–240, 242–244, 252, 253, 261, 267, 274, 278, 280, 306, 316, 319, 323, 374, 399, 414, 417, 420, 433–435, 441, 443, 460, 461, 465 Attenuation...................................................36–42, 45, 47 Atypical touch processing .................................... 227–245 Auditory evoked itch (AEI).........................167, 175–176 Autism................................................ 117, 149, 209, 217, 228–231, 238–240, 244

B Barrel cortex ...............................335, 374, 375, 383, 390 Behaviour.....................................29, 112, 135, 136, 138, 167, 168, 173, 174, 178, 200, 203, 236, 242, 253

Biophysics ............................................................. 183, 184 Bodily illusions .............................................................. 280 Body awareness ........................................... 217, 219, 235 Brain..................................12, 14, 16, 18, 26, 38, 41, 57, 58, 65, 89, 122, 131, 132, 135, 147, 148, 153, 169, 202, 203, 212, 216, 231, 238, 240, 252, 253, 270, 279, 297, 309, 311, 322, 324, 334, 343, 345–348, 351, 352, 354, 360–362, 373, 374, 376, 378, 382, 385, 387, 388, 392, 397, 399, 402, 403, 406, 407, 414, 416, 435, 440, 442, 451, 452, 454–459, 463–471, 473–476

C Calcium imaging .................................374, 375, 378, 387 Chemically evoked itch ................................................. 162 Chronic imaging ........................ 379, 381–384, 387, 388 Cortex ..................................................... 57, 65, 110–112, 117, 169, 174, 203, 211, 212, 216, 234, 288, 296, 335, 339, 351, 353, 355, 359, 361, 362, 373–375, 381, 384, 390, 397, 398, 401, 407, 410, 411, 413, 414, 421, 431–437, 442–444, 452–456, 463–465, 467, 469, 474 Cowhage............................ 162–164, 168–171, 176, 177 C-tactile afferent (CTs).............................. 109–111, 116, 118, 120, 124, 201, 417, 435, 443, 444 Cues ................................................... 44, 72, 73, 77, 148, 177, 184, 187, 190, 192, 197, 207, 253, 256, 272, 289, 325, 403, 405, 419 Cutaneous.................................... 5, 6, 8, 78–82, 87, 110, 133, 142, 143, 201, 219, 307, 308, 324, 333, 336, 348, 349, 361, 414, 417, 464, 471, 475 Cutaneous pain .......................... 200, 205–206, 212–219 Cutaneous rabbit........................................ 75, 78–82, 87, 273, 280, 291

D Detection .............................................. 4, 8, 9, 11, 14–16, 18, 20, 22–24, 26, 73, 130, 132, 134, 143, 147, 153, 183, 184, 187–189, 191, 193, 206, 212–215, 228, 230, 233, 237, 238, 241, 243, 258, 260, 261, 289, 294, 299, 356, 357, 359, 374, 376, 453, 459–463, 468, 473

Nicholas Paul Holmes (ed.), Somatosensory Research Methods, Neuromethods, vol. 196, https://doi.org/10.1007/978-1-0716-3068-6, © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2023

483

SOMATOSENSORY RESEARCH METHODS

484 Index

Devices............................................... 6, 8, 10, 12, 13, 24, 36, 42, 49, 58, 59, 65, 72, 75–78, 80, 81, 89–91, 114, 132, 133, 138, 150, 152, 163, 171, 178, 254, 257, 259, 271, 275, 287–294, 296–299, 315, 336–338, 342–347, 376, 387, 404, 406, 417, 439, 441, 444, 455–457, 470–472, 475 Digits .......................................................... 233, 398, 399, 401, 414, 415, 418, 464 Discrimination......................... 3, 5–7, 11, 12, 14–16, 18, 19, 22–24, 26, 29, 56, 59, 63, 64, 71, 102, 153, 228, 230, 236–238, 241, 289, 295, 343, 356, 374, 375, 388, 414, 459, 461, 465, 468, 473 Distance .............................. 6, 13, 36, 38, 72, 73, 78–82, 95–103, 173, 210, 233, 270, 288, 289, 293–297, 314, 325, 356, 376, 381, 393, 444, 474 Distance perception ................................ 95–98, 100, 102

Fluorescence microscopy ............................ 374, 376, 388 Force-discrimination .......................36–39, 42, 43, 46–49 Force-matching .........................................................36–48 Force perception ............................................................. 46 Functional magnetic resonance imaging (fMRI)................................ 10, 26, 37, 65, 67, 72, 100, 174, 211, 216, 277, 309, 398–403, 405–411, 413, 414, 416, 417, 419, 421, 422, 434, 436, 443, 452, 456, 457, 459, 463, 466, 470

E

Haemodynamic response..................................... 400, 401 Hand ...............................7, 9, 18, 24, 35, 36, 38, 40, 44, 47, 57, 59, 60, 62, 63, 66, 74, 85, 87, 95, 96, 98–102, 113, 115–117, 125, 136, 142, 166, 167, 174, 190, 191, 197, 199, 204–206, 210, 213, 215, 219, 244, 254, 256, 258, 260, 261, 268, 270–278, 280–282, 289, 291, 298, 308, 320, 323, 347, 356, 376, 390, 398, 399, 401, 402, 413, 438, 440–443, 457, 458, 462, 463, 469, 470 Haptic illusions..................................................... 280, 291 High-resolution.......................................... 191, 309, 375, 398, 403, 407, 412, 419, 421 Histamine ..................162–164, 168–171, 174, 176, 177 Human............ 4, 41, 42, 110, 112, 115, 116, 119, 161, 162, 166, 168, 189, 202, 218, 257, 258, 291, 305, 306, 308, 324, 333, 334, 347, 397–399, 402, 405, 406, 413, 417, 435, 438, 452, 475 Hydration ................................... 196, 252–254, 258, 259 Hygroreception .................................................... 182, 254

Electrical/mechanically evoked itch ................... 162, 166 Electrode ..............................................58, 59, 64, 65, 90, 117, 119, 122, 123, 126, 137–139, 144, 145, 152, 163, 164, 168, 171, 305, 306, 308, 310–312, 314–316, 318–326, 336, 339, 341–349, 351, 352, 354–356, 358–362, 405, 431, 437, 444, 452, 455, 457, 458, 465, 469, 470 Electroencephalography (EEG) ................ 8, 37, 90, 112, 120, 123, 131, 133, 134, 137, 139, 140, 144–146, 149, 153, 216, 231, 234, 239, 245, 309, 311, 339, 341, 352, 406, 431, 432, 437, 439–441, 444, 452, 457, 458, 463, 466, 467 Electromyography (EMG) .................. 56, 58–61, 64–66, 112, 117, 120, 122, 124, 131, 133, 139, 144, 145, 311, 323, 444, 468 Electrophysiological methods ............311, 325, 375, 436 Event-related potentials (ERPs)......................... 152, 341, 352, 432, 434 Experimental induction .............162, 169, 170, 176, 276 Extracellular recordings ....................................... 168, 375

F Feel.........................................4, 5, 12, 14–16, 23, 24, 26, 28–30, 35–38, 46, 47, 50, 62, 67, 76, 79, 82, 86, 87, 90, 95, 101, 102, 117, 130, 151, 161, 166, 167, 172, 173, 178, 184, 209, 241, 257, 261, 267–273, 276–278, 280, 281, 318, 470, 471 Filling-in ................................................ 75, 78, 83–88, 91 Fingers .........................................6, 8, 11, 12, 16, 18, 19, 24, 26, 28, 30, 37–39, 42–44, 46, 47, 49, 50, 72, 80, 83–85, 87, 88, 90, 146, 152, 172, 174, 182–184, 189–191, 237, 254, 268, 271, 272, 274–278, 280, 281, 288, 320, 326, 398, 399, 404, 414, 415, 433, 434, 437, 438, 441–443, 452, 453, 457, 459–461, 463, 468, 470, 473, 474

G Glabrous skin.............................................. 120, 207, 210, 215, 324, 336, 349, 353, 355, 361, 435 Grating orientation task.......................... 6, 7, 13, 24, 255

H

I Illusion ................... 4, 36, 38, 57–63, 65, 67, 73, 75, 76, 78, 95, 96, 98, 100, 182, 267–280, 282, 297, 438 Illusions of touch .......................................................... 280 Imaging........... 10, 58, 72, 78, 216, 231, 234, 238, 245, 259, 279, 374–379, 381–388, 390–393, 398, 400, 401, 403, 405–407, 413, 414, 417, 419, 470 Impairment...........................................65, 136, 169, 188, 251, 252, 261, 287, 293, 294, 452 Induction .......................... 121, 133, 139, 142, 162–164, 169–172, 176, 276, 348, 351, 378, 382 Insula ..................................................110–112, 169, 174, 202–204, 211, 216, 230, 231 Intensity 5, 8, 14–16, 18, 21, 22, 29, 30, 35–38, 40, 43, 45–50, 62–64, 73, 75, 76, 78, 82, 87, 89, 91, 113, 116, 123, 125, 130, 132–135, 139–146, 152, 153, 162–164, 168–172, 174, 176–178, 205,

SOMATOSENSORY RESEARCH METHODS Index 485 208, 213, 214, 289, 294, 295, 314, 319, 321, 356, 376, 381, 387, 414, 438, 452, 453, 456, 459, 461, 464, 465, 468, 471, 473, 474 Interference ................74, 126, 188, 277, 292, 309–312, 315, 325, 338, 343, 353, 444, 451, 459, 468, 473 Internal forward models ................................................. 40 Interoception............ 199, 200, 209, 215, 217, 219, 244 Itch....................109, 121, 132, 161–178, 201, 212, 417

K Kinaesthetic illusion ........................................................ 67 Kinesthetic illusions ...................................................... 280

L Limb.........................................55–57, 66, 67, 79, 80, 89, 91, 96, 123, 132, 152, 200, 205, 218, 219, 268, 269, 272, 308, 335, 361, 399, 402 Local field potential (LFP) .................................. 339, 351 Localization .................... 4, 15, 79–82, 86, 96, 202, 237, 311, 312, 316, 343, 374, 397, 432, 461, 469, 471

M Matching ....... 16, 56, 79, 167, 272, 276, 355, 390, 471 Mechanoreceptive afferent ................................. 212, 308, 309, 320, 349, 350 Mechanoreceptor ...................................5, 6, 35, 56, 109, 119, 182, 201, 205, 252, 258, 288, 292, 308, 333, 336, 349, 355, 361, 401, 417, 435, 441 Mechanosensation.....................................................5, 334 Mice/mouse ...........................................41, 81, 172, 334, 356, 374, 375, 378, 381–388, 390, 393, 406 Microinjection .......... 335, 339, 342, 346, 347, 360–362 Microneurography ............................ 109, 110, 117, 119, 120, 168, 202, 305–312, 314–318, 321–323, 325, 326, 402, 443, 467 Microscopy .................................................. 374, 376, 388 Microstimulation........................................ 308, 345–347, 356, 358, 359, 404, 417 Moisture ....................182–184, 186, 189, 190, 192–194 Motion..............................................4, 56, 58, 61, 71–91, 170, 235, 256, 257, 260, 268, 291, 343, 390, 407, 409, 440, 443, 456, 457 Motion cues........................................................ 72–74, 76 Motion perception ............................................. vii, 77, 89 Multisensory ................................................. 61, 253, 277, 292, 375, 434, 435, 438, 440–442, 475 Multisensory perception ...................................... 259, 260 Muscle............................................ 3, 4, 56–67, 112, 117, 119, 123, 126, 138, 139, 142, 144, 145, 152, 206, 257, 269, 308, 316, 318, 321, 348, 355, 383, 452, 453, 464, 469, 471, 475, 476 Muscle proprioception.................................................... 58

N Nerve ............................................... 8, 12, 109, 110, 119, 132, 139, 140, 143, 145, 162, 163, 168, 201, 202, 206, 216, 233, 305–326, 334, 336, 340, 345, 348, 349, 352, 358, 404, 417, 432, 435, 436, 443, 464, 467, 468, 475, 476 Neurodevelopmental disorders .................. 227, 229, 230 Neuroimaging ...................... 11, 57, 112, 131, 134, 203, 211, 212, 216, 308, 309, 403, 452, 457, 463, 466 Neuron .........................42, 96, 182, 202, 252, 312, 333, 334, 339, 340, 344–346, 351, 353, 355, 356, 359, 360, 362, 373–378, 388–391, 401, 464, 465 Neuronavigation ........................ 455–458, 470, 473, 475 Neuroplasticity .............................................................. 464 Nociception ....................... 131, 132, 200–202, 216, 333 Nociceptive withdrawal reflex (NWR) ............................................. 139, 140, 145 Numb-spots..................................................78, 83–87, 91

O Objects........................................................ 36, 73, 75–78, 83, 90, 98, 132, 153, 166, 181, 184, 215, 256, 257, 268, 270–272, 275, 276, 290, 293–296, 299, 336, 374, 375, 404, 438

P Pain ....................5, 16, 89, 96, 109, 110, 112, 129–153, 161–164, 173, 176, 199–202, 204–206, 212–219, 323, 382, 402, 417, 418, 420, 436, 471 Pain assessment .................................................... 136, 141 Pain rating .................................................. 130, 138, 143, 146, 149, 213–215 Pain threshold ............................................ 130, 132, 134, 141, 142, 145, 146, 212–214 Pain tolerance ....................................................... 132, 136 Perception................................................. 3–6, 13, 16, 19, 23, 24, 29, 30, 35, 37, 40, 45, 46, 55, 56, 61, 63, 71, 72, 75–78, 87, 89, 90, 96, 98, 102, 110–112, 123, 125, 130–132, 134, 137, 181–184, 186–188, 191–195, 197, 200–202, 204–206, 208, 209, 212–219, 227, 228, 230–232, 235, 237–239, 241, 242, 244, 251–254, 257, 258, 260, 261, 268–270, 274, 280, 288, 291, 334, 335, 355, 359, 361, 373, 374, 434, 438, 441, 444, 453, 459, 468, 469, 473, 476 Peripheral nerve ..............................................6, 139, 188, 201, 202, 206, 305, 307, 312, 324, 336, 345, 352, 358, 432, 453, 467, 468, 475 Physiology .................................................. 117, 183, 232, 323, 324, 335, 435, 436

SOMATOSENSORY RESEARCH METHODS

486 Index

Potentials ....................................... 26, 58, 61, 65, 72, 73, 83, 91, 133, 150, 151, 184, 196, 203, 205, 209, 216, 217, 232, 233, 239, 240, 244, 269, 278, 288, 289, 296, 299, 307, 308, 311, 312, 314, 315, 321, 325, 333, 334, 336, 339, 341, 343, 348, 349, 352, 353, 356, 375, 376, 378, 387, 399, 402, 431, 434, 436, 437, 440, 441, 444, 455, 460, 463, 465, 468, 473, 475 Prediction ..........................................................38, 40, 45, 174, 212, 399, 417 Proprioceptive illusions.............................................62, 63 Pruritus .......................................................................... 161 Psychologically evoked itch ................................. 162, 172 Psychophysics .....................................3–5, 18, 36, 38, 47, 71, 78, 97, 110, 113, 129, 130, 136–138, 143, 153, 187, 238, 242, 244, 245, 270, 402

Q Questionnaires ........................................... 122, 136, 138, 152, 168, 173, 189, 213, 214, 217–219, 227, 228, 230, 232, 234–236, 240–242, 276, 297, 299

R Receptive fields............................................. 5, 6, 96, 312, 314, 315, 321, 326, 340, 348, 406 Receptors ........................4, 21, 35, 56, 58, 75, 125, 146, 164, 182, 201, 202, 257, 290, 297, 314, 321, 322, 333–335, 346, 373, 374, 417, 438, 441, 475 Remapping .................................276, 289, 290, 295, 438 Rodent .............................. 334–337, 339, 341, 345–347, 355, 356, 359, 361, 374, 376, 382, 387 Rubber hand illusion ................. 269, 270, 272, 274–277

S Scanner ................................. 10, 26, 122, 174, 259, 309, 310, 376, 402–405, 407, 417, 419, 420, 459 Sciatic nerve................................................................... 453 Self-movement ................................................................ 89 Sensitivity................... 3, 5, 9, 15, 20, 21, 24, 26, 27, 29, 41, 46, 47, 64, 78, 87, 95, 102, 121, 130, 134, 135, 182, 183, 188, 190, 194, 211, 214, 227, 234, 235, 241, 252, 257–259, 291, 321, 334, 357, 359, 376, 391, 400, 408, 460, 463, 473 Sensory saltation ................. 75, 78–82, 87, 90, 280, 291 Signals .....................................................8, 10, 11, 13, 14, 21, 36, 40, 42, 43, 59–62, 65, 81, 90, 130, 132, 143, 144, 167, 168, 174, 181, 182, 199–203, 205, 211, 212, 215, 216, 219, 252, 257, 273, 279, 291, 292, 306, 308–313, 316, 321, 322, 325, 326, 336, 337, 339, 341, 343, 344, 346, 352–354, 357, 359, 375, 376, 383, 386, 390, 399, 402, 405, 407, 411, 413, 431, 432, 439–441, 444, 457, 463, 466, 470, 473

Single unit ......................................... 110, 117, 168, 307, 315, 316, 326, 339, 344, 348, 349, 353, 355 Size perception ................................................................ 96 Skin .............................3–6, 8, 10–12, 14, 21, 23, 26, 35, 44, 56, 58, 61, 64, 71–79, 83–86, 89, 91, 95, 96, 98, 100–102, 109, 110, 113–117, 119, 121, 125, 131–133, 137–139, 142, 146, 162–174, 176–178, 182–184, 186, 187, 190, 192, 195–197, 200–212, 214–216, 218, 219, 238, 252–259, 268, 270, 271, 275, 277, 278, 289–291, 295, 297, 306, 308, 310, 311, 315, 316, 318–326, 333–337, 348–350, 353, 359, 361, 380, 383, 384, 390, 404, 415, 417, 435, 438, 439, 441, 443, 457, 459, 461, 476 Social touch ................................................ 110, 115, 218, 230, 233, 238, 244, 245 Softness ................................................................. 208, 270 Somatosensory ................................ 5, 21, 30, 36, 38, 40, 44, 45, 72, 96, 110, 112, 132, 134, 152, 174, 188, 212, 237, 238, 240, 252, 253, 257, 288, 289, 297, 333, 348, 351–356, 360, 373, 374, 376, 390, 397, 399, 402–405, 407, 413–415, 417, 419, 421, 422, 431–435, 437–439, 442, 443, 452–455, 459, 461, 463–468, 470, 474, 476 Somatosensory attenuation ............35–38, 40, 41, 44–46 Somatosensory cortex.......................................14, 37, 38, 41, 96, 111, 112, 169, 202, 211, 216, 253, 334, 335, 360, 374, 376, 378, 383, 384, 393, 397, 401–405, 410, 411, 413, 414, 417, 421, 422, 432, 437, 438, 442, 452–455, 462, 463, 465, 466, 469, 472, 473, 475 Somatosensory evoked potentials (SEPs) ...................348, 432–435, 437–444, 457, 465, 468 Somatosensory experiments ........................ 72, 110, 257, 288, 289, 374, 443, 452, 454 Somatosensory illusions......................125, 268, 269, 273 Somatotopic mapping................ 383, 399, 402, 413–417 Spike.............................................73, 119, 312, 321, 322, 326, 339–341, 343, 344, 347–356, 358, 378 Stickiness...........................................................5, 183, 208 Symptoms ..................................................... 41, 138, 141, 148, 152, 218, 227–230, 232, 239, 434, 452

T Tactile ...........................................3–8, 10, 12–16, 18–20, 23, 24, 26, 28, 30, 39, 72–76, 83, 87, 89, 95–102, 109, 110, 112–114, 121, 123, 125, 130, 134, 183, 184, 186, 192, 200–210, 215, 217–219, 227–240, 242–245, 252–259, 261, 268–271, 275, 277, 278, 282, 288–296, 309, 321, 323, 326, 333–335, 349, 353, 355, 359, 361, 373–375, 398, 403, 404, 414, 417, 418, 431, 433–435, 437–444, 453, 454, 457, 459–463, 465, 467–469, 471, 473, 476

SOMATOSENSORY RESEARCH METHODS Index 487 Tactile illusions ......................................95, 268, 270, 280 Tactile motion ............................................ 71–73, 75, 76, 78, 85, 87, 89, 91, 113, 269 Temperature ..............................4, 5, 11, 14, 16, 26, 109, 110, 114, 116, 125, 131, 132, 141, 142, 145, 146, 149, 150, 169, 178, 182–184, 188–194, 196, 197, 201, 203, 205–207, 212, 213, 215, 217–219, 230, 233, 244, 253, 254, 277, 310, 335, 348, 378, 382, 383, 417, 435, 441, 459 Tendon .............................. 56–61, 63–66, 316, 318, 334 Texture....................................................72, 73, 113, 114, 116, 125, 230, 231, 233, 237, 256, 257, 260, 261, 288, 299, 336, 391, 404, 414 Thermosensation........................................................... 183 Ticklishness......................................................... 36–38, 40 Tissue ....................... 135, 138, 142, 143, 150, 173, 187, 234, 258, 307, 310, 311, 314, 315, 318–320, 322, 324, 326, 334, 336, 339, 341–343, 345, 346, 348, 351, 352, 356, 362, 376, 382–385, 387, 390, 392, 407, 408, 411, 412, 432 Touch...........................................3, 4, 12, 13, 16, 29, 30, 35–41, 45, 65, 71–76, 78, 83, 87, 89, 90, 95, 97, 98, 100, 102, 109, 110, 112–123, 125, 132, 144, 161, 166, 170, 175, 182, 184, 187–190, 193, 194, 196, 197, 199, 201–212, 217, 219, 227, 228, 230–233, 235, 239, 241, 243, 244, 252, 253, 256–261, 270, 271, 274–278, 280–282, 288, 290, 297, 308, 319, 320, 323, 325, 334, 336, 355, 375, 385, 386, 403, 404, 417, 421, 432, 434–436, 438–444, 476 Training ..................................... 6, 16, 22–24, 45, 46, 50, 67, 74, 114, 123, 135, 150, 204, 232, 234, 242, 275, 288–290, 292, 296–299, 306, 382, 405 Transcranial alternating current stimulation (TACS)............................................. 453, 465, 466 Transcranial direct current stimulation (TDCS) ..................................................... 453, 465

Transcranial magnetic stimulation (TMS) ....................................13, 29, 57, 451–476 Two photon.......................................................... 375, 387 Two-point discrimination.................................. 6, 98, 102

U Ultrasound.................. 76, 271, 308, 310, 314, 318, 319

V Velocity ............12, 49, 56–58, 62, 63, 75, 87, 110, 111, 113–118, 120–123, 125, 126, 202–204, 207, 210, 211, 216, 276, 334, 346, 387, 417, 436, 444 Vibration..............................4–6, 8, 9, 13, 14, 20–22, 24, 28–30, 56–67, 72–74, 81, 82, 91, 134, 166, 230, 238, 239, 242, 244, 268, 269, 289, 292, 295, 309, 336, 351, 359, 414, 415, 417, 443, 461, 473 Vibrotactile ...................8, 9, 13, 18, 21, 79, 87, 96, 230, 233, 257, 258, 291, 292, 295, 337, 350, 353, 355–357, 399, 403, 404, 414, 417, 419, 461, 470 Vicarious ................... 112, 113, 115, 121, 122, 125, 443 Visceral signals............................................................... 201 Vision ...............30, 61, 71–73, 75, 77, 78, 83, 102, 121, 172, 205, 235, 243, 252, 259, 268, 270, 273, 274, 287–290, 292–294, 299, 406, 434, 435, 441 Vision to touch................................. 72, 73, 83, 274, 288 Visual impairment ....................................... 251, 287, 288 Visually evoked itch (VEI)......... 166, 167, 172–174, 177

W Wetness ..................................... 5, 16, 181–184, 186–196 Whiskers..............................................334–336, 348, 359, 361, 374, 375, 383, 386, 388, 390, 391 Whose hand illusion (WHI) ...............271, 274, 277–282

Y Yes/no detection task ................................................... 356