Advanced Technologies in Rehabilitation: Empowering Cognitive, Physical, Social and Communicative Skills Through Virtual Reality, Robots, Wearable Systems and Brain-computer Interfaces 9781607500186

Citation preview

ADVANCED TECHNOLOGIES IN REHABILITATION

Studies in Health Technology and Informatics This book series was started in 1990 to promote research conducted under the auspices of the EC programmes’ Advanced Informatics in Medicine (AIM) and Biomedical and Health Research (BHR) bioengineering branch. A driving aspect of international health informatics is that telecommunication technology, rehabilitative technology, intelligent home technology and many other components are moving together and form one integrated world of information and communication media. The complete series has been accepted in Medline. Volumes from 2005 onwards are available online. Series Editors: Dr. O. Bodenreider, Dr. J.P. Christensen, Prof. G. de Moor, Prof. A. Famili, Dr. U. Fors, Prof. A. Hasman, Prof. E.J.S. Hovenga, Prof. L. Hunter, Dr. I. Iakovidis, Dr. Z. Kolitsi, Mr. O. Le Dour, Dr. A. Lymberis, Prof. J. Mantas, Prof. M.A. Musen, Prof. P.F. Niederer, Prof. A. Pedotti, Prof. O. Rienhoff, Prof. F.H. Roger France, Dr. N. Rossing, Prof. N. Saranummi, Dr. E.R. Siegel and Dr. P. Wilson

Volume 145 Recently published in this series Vol. 144. B.K. Wiederhold and G. Riva (Eds.), Annual Review of Cybertherapy and Telemedicine 2009 – Advanced Technologies in the Behavioral Social and Neurosciences Vol. 143. J.G. McDaniel (Ed.), Advances in Information Technology and Communication in Health Vol. 142. J.D. Westwood, S.W. Westwood, R.S. Haluck, H.M. Hoffman, G.T. Mogel, R. Phillips, R.A. Robb and K.G. Vosburgh (Eds.), Medicine Meets Virtual Reality 17 – NextMed: Design for/the Well Being Vol. 141. E. De Clercq et al. (Eds.), Collaborative Patient Centred eHealth – Proceedings of the HIT@HealthCare 2008 joint event: 25th MIC Congress, 3rd International Congress Sixi, Special ISV-NVKVV Event, 8th Belgian eHealth Symposium Vol. 140. P.H. Dangerfield (Ed.), Research into Spinal Deformities 6 Vol. 139. A. ten Teije, S. Miksch and P. Lucas (Eds.), Computer-based Medical Guidelines and Protocols: A Primer and Current Trends Vol. 138. T. Solomonides et al. (Eds.), Global Healthgrid: e-Science Meets Biomedical Informatics – Proceedings of HealthGrid 2008 Vol. 137. L. Bos, B. Blobel, A. Marsh and D. Carroll (Eds.), Medical and Care Compunetics 5 Vol. 136. S.K. Andersen, G.O. Klein, S. Schulz, J. Aarts and M.C. Mazzoleni (Eds.), eHealth Beyond the Horizon – Get IT There – Proceedings of MIE2008 – The XXIst International Congress of the European Federation for Medical Informatics

ISSN 0926-9630

Advanced Technologies in Rehabilitation Empowering Cognitive, Physical, Social and Communicative Skills through Virtual Reality, Robots, Wearable Systems and Brain-Computer Interfaces

Edited by

Andrea Gaggioli Catholic University of Milan, Milan, Italy Istituto Auxologico Italiano, Milan, Italy

Emily A. Keshner Temple University, Philadelphia, USA

Patrice L. (Tamar) Weiss University of Haifa, Haifa, Israel

and

Giuseppe Riva Catholic University of Milan, Milan, Italy Istituto Auxologico Italiano, Milan, Italy

Amsterdam • Berlin • Tokyo • Washington, DC

© 2009 The authors and IOS Press. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior written permission from the publisher. ISBN 978-1-60750-018-6 Library of Congress Control Number: 2009927468 Publisher IOS Press BV Nieuwe Hemweg 6B 1013 BG Amsterdam Netherlands fax: +31 20 687 0019 e-mail: [email protected] Distributor in the UK and Ireland Gazelle Books Services Ltd. White Cross Mills Hightown Lancaster LA1 4XS United Kingdom fax: +44 1524 63232 e-mail: [email protected]

Distributor in the USA and Canada IOS Press, Inc. 4502 Rachael Manor Drive Fairfax, VA 22032 USA fax: +1 703 323 3668 e-mail: [email protected]

LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS

Advanced Technologies in Rehabilitation A. Gaggioli et al. (Eds.) IOS Press, 2009 © 2009 The authors and IOS Press. All rights reserved.

v

INTRODUCTION The proportion of the world population over 65 years of age is climbing. Life expectancy in this age group is increasing, and disabling illnesses now occur later in life, so the burden on the working–age population to support health care costs of aging populations continues to increase. These demographic shifts portend progressively greater demands for cost effective health care, including long-term care and rehabilitation. The most influential change in physical rehabilitation practice over the past few decades has been the rapid development of new technologies that enable clinicians to provide more effective therapeutic interventions. New rehabilitation technologies can provide more responsive treatment tools or augment the therapeutic process. However, the absence of education about technological advancements and apprehensions by clinicians related to the role of technology in the treatment delivery process puts us at risk of losing the benefit of an essential partner in achieving successful outcomes with the physically disabled and aging population. There are two reasons that may explain why rehabilitation practitioners do not play an integral role in the development and evaluation of these new technologies. First, the engineers who develop these technologies do not recognize the value they could derive by consulting with rehabilitation professionals in order to make their machine-user interfaces more efficient, user friendly, and effective for specific disabilities. Second, many rehabilitation professionals are uncomfortable with technology and fear that it may take the place of individualized interactions with patients. Funding challenges, a lack of public awareness about technology’s potential, a shortage of trained experts, and poor collaboration among researchers, clinicians, and users are often the cause for an absence of clinical trials that demonstrate the value of near-term and future rehabilitation applications. If technology transfer is to become successful, we need to establish collaborative interactions in which the goals of each discipline become overlapping with the skills and goals of the other fields of endeavor and of the consumer. The rapid rise of technological development is pushing the market place and it is essential that rehabilitation specialists oversee the quality and validity of these new applications before they reach the consumer. It is clear from the chapters in this book that improvements in technology depend on interdisciplinary cooperation among neuroscientists, engineers, computer programmers, psychologists, and rehabilitation specialists, and on adoption and widespread application of objective criteria for evaluating alternative methods. The goal of this book is to bring ideas from several different disciplines in order to examine the focus and aims that drive rehabilitation intervention and technology development. Specifically, the chapters in this book address the questions of what research is currently taking place to further develop rehabilitation applied technology and how we

vi

have been able to modify and measure responses in both healthy and clinical populations using these technologies. In the following sections we highlight some of the issues raised about emergent technologies and briefly describe the chapters from this book that are dedicated toward addressing these issues.

1. Does Training with Technology Add to Functional Gains? Before we can develop a successful intervention, we need to determine what the end goal is. A number of different therapeutic technologies are already available for use in clinics, but their value to the treatment program is not well defined. Developers and clinicians must consider whether a technological device better targets diagnostic or therapeutic interventions. Does it serve as an extension or repetition of conventional therapeutic interventions? Do we want it to perfectly replicate the actions of a therapist or to assist or augment the actions of the therapist? For example, as stated by Reinkensmeyer in his chapter, there has been a rapid increase in the number of robotic devices that are being developed to assist in movement rehabilitation, yet it is still not well understood how these devices enhance movement recovery, and whether they have inherent therapeutic value that can be attributed to their robotic properties. Chapters by Frisoli et al. and Piron et al. present results of clinical trials that demonstrate improvements in functional outcomes on standard clinical scales when compared with more traditional clinical interventions which would suggest value in adding technology to therapeutic interventions.

2. Are there Rules that Govern Recovery of Function? Are learning rules for recovery similar to those for skill acquisition? In particular, should we be concerned mostly with error reduction or feedback enhancement? If we are concerned with recognition of movement error, do we try to increase or decrease that error for learning? How do we instruct patients to attend not only to the error, but also to their own kinematics? If functional recovery depends on plasticity of the central nervous system, can the use of technology enhance this plasticity? If we are attempting to promote plastic changes in the nervous system, then motor learning principles most likely should be adhered to and rules for learning need to be defined including the optimal length and frequency of the intervention and how much interference plays a role in learning. Cameirao et al. use virtual reality to engage patients in task specific training scenarios that adapt to their performance thereby allowing for an individualized training of graded difficulty and complexity. Deutsch provides an overview of virtual reality gaming based technologies to improve motor and cognitive elements required for ambulation and mobility in different patient populations. Levin et al. and Merians et al. demonstrate how movement retraining can be optimized by combining virtual reality with haptic devices if important motor learning elements such as repetition, varied task practice, performance feedback and motivation are incorporated. Riva et al. discuss development of a new open source system that uses the principles of motor learning within real life context in order to increase generalization of recovered motor and cognitive behaviours. Using a combination of robotics and virtual reality, Sanguineti et al. demonstrated functional gains by tailoring their intervention to the different degrees

vii

of impairment and adapting the intervention as performance changed thereby exploiting the nervous system’s capacity for sensorimotor adaptation.

3. Using the Body’s Own Signals to Augment Therapeutic Gains Another rapidly advancing area of technology for rehabilitation is the application of the individual’s own residual sensory and motor signals to augment function. Although wheelchairs are still the most popular assistive device for patients with spinal cord injuries and disabling neurological conditions, many users encounter difficulties in controlling their powered wheelchairs. The wheelchair represents an assistive device that, in large part, requires the person to adapt to the technology rather than having the technology fit the abilities of the individual. Bonato discusses the emerging use of miniature sensors that can be worn by the patient to measure and transmit information about physiologic and motor functions. Carabalona et al. explore research on brain-computer interfaces and discuss how technologies that are driven by or access the signals initiated by each patient can support activity in their environments.

4. Technology Incorporates Cognition and Action Clinicians often voice concerns about using technological interventions because they appear to replace the human interaction which is believed to be a prime factor in the success of rehabilitation programs. Rehabilitation clinicians work with patients using a combination of verbal, visual, and physical interaction as well as a variety of treatment tools and techniques. Delivering equivalent interventions to patients through technological devices presents significant obstacles, but also presents numerous opportunities to enhance the quality, consistency, and documentation of care received. Several chapters in this book explore how rehabilitation technology offers the capacity to individualize treatment approaches by monitoring the specificity and frequency of feedback, providing standardization of assessment and training, and presenting treatment within a functional, purposeful and motivating context. Antonietti presents the field of music therapy as a tool of the mind, using cognition and emotion as the avenue towards accomplishing goals for rehabilitation. Gaggioli et al. demonstrate how virtual reality can be successfully used to support motor imagery techniques for mental practice in stroke rehabilitation. Keshner and Kenyon discuss how cognitive processes such as perception and spatial orientation can be accessed through virtual reality for the assessment and rehabilitation of perceptual-motor disorders.

5. Technology Enhances the Impact of Rehabilitation Programs One of the greatest challenges for healthcare in the coming decade will be accessibility to the increasing numbers of individuals who are unable to travel to rehabilitation facilities or who do not have local rehabilitation facilities that provide the health maintenance and extended care they require. Additionally, most of the responsibility for caring for individuals with physical or psychological disabilities will fall on their family or on health care aides who do not have the training to provide wellness and rehabilitation interventions. The chapters in this book that address improved access to care and

viii

extending the reach of medical rehabilitation service delivery all emphasize the importance of human factors and user-centered design in the planning, developing, and implementation of their systems. Brennan et al. present a brief history of telerehabilitation and tele-care and offer an overview of the technology used to provide these remote rehabilitation services. Mataric et al. demonstrate how combining the technology of non-contact socially assistive robotics and the clinical science of neurorehabilitation and motor learning can promote home-based rehabilitation programs for stroke and traumatic brain injury. Weiss and Klinger discuss the practical and ethical considerations of using virtual reality for multiple users in co-located settings, single users in remote locations, and multiple users in remote locations.

6. Summary Although new technologies and applications are rapidly emerging in the area of rehabilitation, there are still issues that must be addressed before these can be used both effectively and economically. First, we need to demonstrate that these devices are effective through clinical trials. Second, we must determine how to build devices cheaply enough for mass use. Lastly, we need sufficiently educated physicians and therapists to drive the technology development and applications. Although considerable engineering knowledge is required to understand the potential capabilities of the various technologies, engineering alone will not determine the usefulness of these systems. The chapters we have included in this book clearly demonstrate that in order to design appropriate system features and successful interventions, developers and the users need to be familiar with the scientific rationale for motor learning and motor control, as well as the motor impairments presented by different clinical populations. Ultimately, the impact of these new technologies will depend very much on mutual communication and collaboration between clinicians, engineers, scientists, and the people with disabilities that the technology will most directly impact. Emily A. Keshner Temple University Philadelphia, PA, USA

W. Zev Rymer Northwestern University Chicago, Illinois, USA

ix

CONTRIBUTORS Sergei V. ADAMOVICH Department of Biomedical Engineering, New Jersey Institute of Technology, NJ, USA Sergei Adamovich received his Ph.D. degree in physics and mathematics from Moscow Institute of Physics and Technology. He is currently with the department of Biomedical Engineering at New Jersey Institute of Technology, USA. His research is funded by National Institutes of Health and by the National Institute on Disability and Rehabilitation Research.

Michela AGOSTINI Laboratory of Robotics and Kinematics, I.R.C.C.S. San Camillo Venezia, Padova, Italy Michela Agostini obtained the Degrees in Motor Science and in Physical Therapy at the University of Padova. Her studies are focused on the clinical application of virtual reality and telerehabilitation systems for motor recovery after neurological injury, with specific interest in the motor learning principles involved in the human – machine interaction.

Alessandro ANTONIETTI Department of Psychology, Catholic University of the Sacred Heart, Milan, Italy Alessandro Antonietti is Full professor of Cognitive Psychology and head of the Department of Psychology at the Catholic University of the Sacred Heart in Milano. He investigated the role played by media in thinking processes and he is interested in the application of cognitive issues in the field of education and rehabilitation.

Massimo BERGAMASCO PERCRO Laboratory, Scuola Superiore Sant’Anna, Pisa, Italy Massimo Bergamasco is Full Professor of Applied Mechanics at Scuola Superiore Sant’Anna and the current coordinator of the IP EU project SKILLS. His research activity deals with the study and development of haptic interfaces for the control of the interaction between humans and Virtual Environments.

Sergi BERMÚDEZ I BADIA Institute of Audiovisual Studies, Universitat Pompeu Fabra, Barcelona, Spain Dr. Sergi Bermudez i Badia is a PostDoc and head of the Robotic Systems laboratory of SPECS at the Institute of Audiovisual Studies of the Universitat Pompeu Fabra. He received his Ma. in telecommunications engineering from the Universitat Politècnica

x

de Catalunya (UPC) and his PhD from the Swiss Federal Institute of Technology Zürich (ETHZ). David BRENNAN Center for Applied Biomechanics and Rehabilitation Research, National Rehabilitation Hospital, Washington DC, USA David Brennan, MBE, is a Senior Research Engineer at the National Rehabilitation Hospital in Washington, DC. He has worked for over 10 years on telerehabilitation research and development projects with funding from the National Institutes of Health, and the United States Departments of Education and Defense Simon BROWNSELL School of Health and Related Research, University of Sheffield Regent Court, Sheffield, UK Dr. Brownsell is a Research Fellow at the University of Sheffield, UK. He has 12 years experience working in telecare and telehealth and a particular interest in developing evidence based services for older people. He has written over 50 articles, two books, and three book chapters.Mónica S. CAMEIRÃO SPECS-Institut Universitari de l’Audiovisual (IUA), Universitat Pompeu Fabra, Barcelona, Spain Mónica Cameirão is a PhD student in the SPECS group in the University Pompeu Fabra in Barcelona. Mónica’s main interest is the application of new technologies for rehabilitation and she is currently working on the development and clinical assessment of interactive systems for the neurorehabilitation of motor impairments such as the ones originated by stroke. Roberta CARABALONA Biomedical Technology Department (Polo Tecnologico), Fondazione Don C. Gnocchi, Milan, Italy Roberta Carabalona received the B.Sc. in Biomedical Engineering (1996) from “Politecnico di Milano” and the M.Sc. in Biostatistics and Experimental Statistics (2005) from “Università degli Studi di Milano-Bicocca”. She is researcher in the Biosignal Analysis Area at the Biomedical Technology Department of Fondazione Don Carlo Gnocchi (Milan, Italy). Her research interests include bio-signal analysis and brain-computer interfaces. Maria Chiara CARBONCINI Department of Neurosciences, University of Pisa, Pisa, Italy Maria Chiara Carboncini (MD) is the responsible for upper limb rehabilitation and kinesiology at the Neurorehabilitation Unit of the University Hospital of Pisa.

xi

Maura CASADIO Department of Informatics, Systems and Telematics, University of Genoa, Genoa, Italy Maura Casadio received the Master degree in Electronic Engineering (2002), from the University of Pisa, Italy, the Master degree in Biomedical Engineering and the Ph.D. degree in Bioengineering, Material Science and Robotics (2006) from the University of Genoa, Italy. She is now postdoctoral fellow at Rehabilitation Institute of Chicago, USA.

Paolo CASTIGLIONI Biomedical Technology Department (Polo Tecnologico), Fondazione Don C.Gnocchi, Milan, Italy Paolo Castiglioni received the Ph.D. in biomedical engineering (1993) from the “Politecnico di Milano” University, Italy. He is coordinator of the Biosignal Analysis Area at the Biomedical Technology Department of Fondazione Don Carlo Gnocchi (Milan, Italy). His research interests include bio-signal analysis, physiological mechanisms for the cardiovascular control, gravitational physiology, brain-computer interfaces.

Mauro DAM Departement of neurosciences, University of Padova, Padova, Italy Mauro Dam received a specialization in Neurology in 1979. From 1980 to 1982 he was Visiting Fellow, National Institute on Aging, N.I.H., Bethesda, USA. He is currently Associate professor of Neurology and Scientific Vice President of the Italian Scientific Institutes for Research Hospitalization and Health Care, S Camillo Hospital, Venice. His research interests include: brain metabolism, neuropharmacology, dementia, stroke, neurorehabilitation.

Judith E. DEUTSCH Department of Rehabilitation and Movement Sciences, University of Medicine and Dentistry of New Jersey, USA Judith E. Deutsch is Professor and Director of Rivers Lab. Her research focuses on the development and testing of gaming and virtual reality to improve mobility for individuals post-stroke.

Esther DUARTE OLLER Servei de Medicina Física i Rehabilitació, Hospital de L’Esperança, Barcelona, Spain Esther Duarte Oller, MD, is a specialist in Physical Medicine and Rehabilitation since 1987. She is currently the head of the neurological rehabilitation unit in the Physical Medicine and Rehabilitation Department of the Institut Municipal d’Assistència Sanitària (IMAS), Hospitals del Mar i de l’Esperança in Barcelona, Spain.

xii

Jon ERIKSSON Computer Science Department, University of Southern California, Los Angeles, USA Jon Eriksson is a Master student at the Computer Science Department, University of Southern California.

Antonio FRISOLI PERCRO Laboratory, Scuola Superiore Sant’Anna, Pontedera (Pisa), Italy Antonio Frisoli (Eng., PhD) is Assistant Professor of Applied Mechanics at Scuola Superiore Sant’Anna. He is Associate Editor of IEEE Transaction of Haptics and Presence Teleoperators and Virtual Environments journals. His research interests are in the field of robotic assisted rehabilitation, robotics, virtual reality and haptic interfaces.

Andrea GAGGIOLI Faculty of Psychology, Catholic University of Milan, Milan, Italy Andrea Gaggioli received a MSc in Psychology from University of Bologna and a Ph.D. from the Faculty of Medicine of the University of Milan. He is a researcher at the Faculty of Psychology of the Catholic University of Milan and senior researcher at the Applied Technology for Neuro-Psychology Lab of Istituto Auxologico Italiano (Milan, Italy). He is the founder of Positive Technology, a field that studies how technology can be used to promote mental and physical wellbeing.

Psiche GIANNONI School of Medicine, Master program in physiotherapy, University of Genoa, Genoa, Italy Psiche Giannoni is a trained physiotherapist, IBITA Advanced Course Bobath Instructor and EBTA Senior Bobath Instructor (country representative). She teaches and organizes basic and advanced courses for the treatment of adults with hemiplegia, children with cerebral palsy. She is a Professor at the University of Genoa, Physiotherapy School and anauthor of one book and about 30 scientific publications.

Furio GRAMATICA Polo Tecnologico – Biomedical technology Department, Fondazione Don Carlo Gnocchi ONLUS, Milano, Italy Furio Gramatica, physicist, is the coordinator of the Biomedical Technology Department at Fondazione Don Gnocchi, where he also leads a biophysics and nanomedicine team. His main scientific interest is the application of nanotechnology to diagnosis and targeted drug delivery. Formerly, he served as researcher and project manager at CERN (European Laboratory for Particle Physics, Geneva).

Giovanni GREGGIO School of Physical Medicine and Rehabilitation, University of Padua, Rovigo, Italia

xiii

Giovanni Greggio graduated in Medicine at the University of Padua in 2004, and specialized in Physical Medicine and Rehabilitation in 2009. He took part to the European project “I-Learning” about upper limb rehabilitation after stroke.

Robert KENYON Department of Computer Science, University of Illinois, Chicago, USA Robert Kenyon received his Ph.D. in Physiological Optics from the University of California, Berkeley and is a Professor of Computer Science at the University of Illinois at Chicago. His research spans the areas of sensory-motor adaptation, effects of microgravity on vestibular development, visuo-motor and posture control, flight simulation, Tele-immersion, sensory/motor integration for navigation and wayfinding, virtual environments, and the melding of robots and virtual reality for rehabilitation.

Emily A. KESHNER Department of Physical Therapy and Department of Electrical and Computer Engineering, Temple University, Philadelphia, PA, USA Emily Keshner is Professor and Chair of the Department of Physical Therapy, a Professor in the Department of Electrical Engineering and Computer Science, and Director of the Virtual Environment and Postural Orientation Laboratory at Temple University. She is currently President of the International Society for Virtual Rehabilitation. Her research focuses on how the CNS integrates multiple sensory demands with the biomechanical constraints of postural and spatial orientation tasks.

Evelyne KLINGER LAMPA, Arts et Metiers ParisTech Angers, Laval, France Evelyne Klinger, PhD, Eng, is Researcher of Arts et Métiers ParisTech in Laval, France. Her work is dedicated to the design of virtual reality based methods, concepts and systems for cognitive rehabilitation assessment and intervention. She created the VAP-S, a virtual supermarket for executive functions exploration.

Luiz Alberto Manfré KNAUT School of Rehabilitation, University of Montreal, PR, Brazil Mr. Knaut is a physiotherapist (B.Sc.) who graduated from the Universidade Tuiuti do Parana (Brazil) in 2003. He obtained his M.Sc. in Biomedical Sciences from University of Montreal in 2008. He is currently affiliated with the Hospital Center of Rehabilitation Ana Carolina Xavier and the Coritiba Foot Ball Club (Curitiba-Brazil).

Mindy F. LEVIN School of Physical and Occupational Therapy, McGill University, Montreal, Quebec, Canada

xiv

Mindy Levin is a researcher and neurological physiotherapist (McGill-1996). She obtained an MSc (Clinical Sciences, University of Montreal-1985) and a PhD (Physiology, McGill-1990). She was a Professor in the School of Rehabilitation (UdeM-19922004) and Director of the Physical Therapy Program (McGill-2004-08). She holds a Canada Research Chair in Motor Recovery and Rehabilitation.

Eliane C. MAGDALON Department of Biomedical Engineering, University of Campinas, Campinas, SP, Brazil Eliane Magdalon obtained a B.Sc. in Physical Therapy from the Methodist University of Piracicaba in 2000 and her Master’s degree from the University of Campinas in 2004. She is currently completing her PhD in the Department of Biomedical Engineering (Rehabilitation Engineering) of University of Campinas, Campinas, SP, Brazil.

Maja MATARIĆ Computer Science Department, University of Southern California, Los Angeles, USA Maja J. Mataric is Professor of Computer Science and Neuroscience, Director of the Center for Robotics and Embedded Systems (CRES), and the Viterbi School of Engineering Senior Associate Dean for Research at the University of Southern California. She received her Ph.D. in Computer Science and Artificial Intelligence at MIT in 1994. With the goal of getting robots to help people, her research interests include human– robot interaction and robot control and learning in complex environments.

Sue MAWSON Center for Health and Social Care Research, Sheffield Hallam University, Sheffield, UK Sue Mawson is a Professor of Rehabilitation at Sheffield Hallam University, UK. Her research focuses on improving quality of life of people with neurological problems, particularly through exploration of the effectiveness of rehabilitative interventions. She is a partner in the SMART trial, investigating benefits of technology for stroke rehabilitation.

Andrea MENEGHINI Advanced Technology in Rehabilitation Lab Padua Teaching Hospital, Rehabilitation Unit, University of Padua, Padua, Italy Andrea Meneghini, MD, is a physiatrist specialized in Orthopedics and Traumatology. He is head and founder of the Advanced Technology in Rehabilitation Lab at Padua Teaching Hospital. He has more than 25 years of clinical and research experience. He has been studying the use of virtual reality in the rehabilitation of hemiplegia since the early ’90s.

xv

Alma S. MERIANS Department of Rehabilitation and Movement Science, University of Medicine and Dentistry of New Jersey, NJ, USA Dr. Alma Merians is Professor and Chairperson of the Department of Rehabilitation and Movement Sciences. The major focus of her lab is to study basic mechanisms underlying neuromuscular control of human movement and sensorimotor learning, both in healthy populations and in people with neurological diseases like stroke or cerebral palsy.

Pietro MORASSO Dept. of Informatics, Systems, Telematics, University of Genoa, Genova, Italy Pietro Morasso is full professor of Bioengineering at the Genoa University. Since 1970 he has been associated with the Neurophsysiological laboratory of Emilio Bizzi (MIT). His scientific interests include neural control of movement, motor learning, anthropomorphic robotics, and rehabilitation engineering. He is author and co-author of 7 books and over 300 papers (44 indexed in Medline).

Francesca MORGANTI Department of Human Science, University of Bergamo, Bergamo, Italy Francesca Morganti received a MSc in Psychology from Padua University, where she took a specialization in Neuropsychology and Clinical Psychophysiology. She also obtained a PhD in Cognitive Science from the University of Turin. Her research focuses on the application of interactive technologies to experimental psychology and neuroscience, as well as the study of intersubjectivity from the perspectives of neuroscience, cognitive science and social cognition.

Francesco PICCIONE Department of Neurorehabilitation, IRCCS Hospital “San Camillo” Alberoni, Venice, Italy Francesco Piccione has a degree in Medicine and Surgery and Residency in Neurology and Neurophysiopathology. He is currently Director of Unit of Neurodegenerative Disorders and Neurophysiopathology in San Camillo Hospital, Venice. Expert in EMG, EEG and Evoked Potentials, and scientific researcher in the Neurophysiology field applied to disability improvement.

Maurizia PIGATTO Dipartimento di Specialità Medico Chirurgiche, University of Padua, Padua, Italy Maurizia Pigatto is a chartered Physiotherapist. She has over 25 years of clinical experience. She has collaborated with the School of Physioterapy and Master of Musicotherapy at Padua University. She serves as senior research collaborator at Advanced Technology in Rehabilitation Lab at Padua Teaching Hospital.

xvi

Lamberto PIRON Neurorehabilitation Department, I.R.C.C.S. San Camillo Hospital, Venice, Italy Lamberto Piron is a neurologist. He is the director of the “Cerebro-vascular diseases” Operative Unit and of the “Kinematics and Robotics” laboratory at I.R.C.C.S. San Camillo Hospital. His research focuses on the use of virtual environments, robotics and telerehabilitation for training patients with arm motor impairment after neurological lesions.

Ilaria POZZATO Rehabilitation Unit, University of Padua, Padua, Italy Dr. Ilaria Pozzato is currently postgraduate training at the Medical School of Specialization in Physical Medicine and Rehabilitation at Padua University. She has graduated in Medicine at the University of Padua with a thesis on the application of virtual reality and motor imagery training for upper limb rehabilitation of hemiplegic patients.

David J. REINKENSMEYER Department of Mechanical and Aerospace Engineering, University of California at Irvine, CA, USA David J. Reinkensmeyer received his B.S. degree from the Massachusetts Institute of Technology and his M.S. and Ph.D. degrees from the University of California at Berkeley. He was a research associate at the Rehabilitation Institute of Chicago before joining the University of California at Irvine.

Giuseppe RIVA Department of Psychology, Catholic University of Milan, Milan, Italy Giuseppe Riva, Ph.D. is Associate Professor of General Psychology and Communication Psychology at the Catholic University of Milan, Italy; Director of the Interactive Communication and Ergonomics of NEw Technologies – ICE-NET – Lab. at the Catholic University of Milan, Italy, and Head Researcher of the Applied Technology for Neuro-Psychology Laboratory – Istituto Auxologico Italiano (Milan, Italy). His research activities focus on methods and assessment tools in psychology and the use virtual reality in assessment and therapy.

Bruno ROSSI Neurorehabilitation Unit, Department of Neurosciences, University of Pisa, Pisa, Italy Bruno Rossi (MD) is Head of the Neurorehabilitation Unit, Department of Neuroscience University Hospital Pisa, and Full Professor of Physical Medicine and Rehabilitation. His research interests include clinical neurophysiology, EMG in neuromuscular disorders, brain-stem and spinal reflexology, muscle fatigue analysis, clinical neurology, psychophysiology of consciousness disorders and neurorehabilitation.

xvii

William Zev RYMER Rehabilitation Institute of Chicago/Department of Physical Medicine and Rehabilitation, Northwestern University, Chicago, Illinois, USA Dr W, Zev Rymer was trained in Medicine at Melbourne University, Australia, and received a PhD from Monash University. After postdoctoral training at NIH and Johns Hopkins, he was appointed as a Physiology Professor at Northwestern University in 1977, and moved to the RIC in 1989 as the Searle Director of Research.

Vittorio SANGUINETI Dept Informatics Systems Telematics, University of Genoa and Italian Institute of Technology, Genoa, Italy Vittorio Sanguineti was born in Genova, Italy in 1964. He got a Master’s Degree in Electronic Engineering in 1989 and a PhD in Robotics in 1994, both at the University of Genova. Until 1998 he was working, as a post-doctoral fellow, at the Institut de la Communication Parlée, INPG (Grenoble, France); at the Department of Psychology, McGill University, (Montreal, Canada); and at the Department of Physiology, Northwestern University, (Chicago, USA). Since 1999 he has been an assistant professor at the Dipartimento di Informatica, Sistemistica e Telematica (DIST) of the University of Genova.

Valentina SQUERI Dept Informatics Systems Telematics, University of Genoa and Italian Institute of Technology, Genoa, Italy Valentina Squeri received a Master’s Degree in Bioengineering at the University of Genova in 2006. She is currently a PhD student at the University of Genoa and the Italian Institute of Technology. Her areas of interest include motor control, motor learning and their application to robot therapy.

Sandeep SUBRAMANIAN School of Physical and Occupational Therapy, McGill University, Quebec, Canada Sandeep Subramanian, MSc, PT is currently enrolled in the PhD program in Rehabilitation Sciences at the School of Physical and Occupational Therapy, McGill University. His research focuses on the use of feedback for motor learning in patients with chronic stroke and the use of different environments to maximize motor recovery post-stroke.

Adriana TAPUS Computer Science Department, University of Southern California, CA, Los Angeles, USA Dr. Adriana Tapus is a research associate at University of Southern California (USC, USA) in the Interaction Lab/ Robotics Research Lab, Computer Science Department. She received her Ph.D. in Computer Science from Swiss Federal Institute of Technology, Lausanne (EPFL) in 2005, her M.S. in Computer Science from University Joseph

xviii

Fourier, Grenoble, France in 2002 and her degree of Engineer in Computer Science and Engineering from “Politehnica” University of Bucharest, Romania. Her current research interests are socially assistive robotics for post-stroke patients and people suffering from cognitive impairment and/or Alzheimer’s disease, humanoid robotics, machine learning, and computer vision...

Paolo TONIN Department of Neurorehabilitation, IRCCS San Camillo S. Polo, Venice, Italy Paolo Tonin is a neurologist and a physiatrist. He has carried out research in the rehabilitation of Stroke, Parkinson disease, Multiple Sclerosis, Traumatic Brain Injury, with particular reference to the role of emerging technologies in neurorehabilitation. Dr. Tonin is member of the Board of the Italian Society of Neurorehabilitation and of the Management Committee of the World Federation of Neurorehabilitation.

Eugene TUNIK Department of Rehabilitation and Movement Science, University of Medicine and Dentistry of New Jersey, Newark, NJ, USA Dr. Tunik completed his degrees in Physical Therapy at Northeastern University and doctorate at the Center for Molecular and Behavioral Neuroscience at Rutgers University. He studies brain mechanisms involved in motor control and learning in health and disease and how this information can guide therapeutic interventions.

Andrea TUROLLA Laboratory of Robotics and Kinematics, I.R.C.C.S. San Camillo Venezia, Noventa Padovana, Italy Andrea Turolla is Physical Therapist. He obtained the Master’s Degree in Science of Rehabilitation Health Profession at the University of Padua. His research focuses on the application of virtual reality and robotic systems in motor rehabilitation, with specific interest in the motor learning principles involved in human-machine interaction.

Elena VERGARO Department of Informatics, Systems and Telematics, University of Genoa, Genoa, Italy Elena Vergaro received a Master’s degree in Biomedical Engineering (2006), from the University of Genoa, Italy. She is now a Ph.D. student in Bioengineering at the same university. Her area of interest is motor control and motor skill learning.

Paul VERSCHURE Institute of Audiovisual Studies, Universitat Pompeu Fabra, Barcelona, Spain Paul Verschure is a research professor with the Catalan Institute of Advanced Studies (ICREA) and the Universitat Pompeu Fabra. Paul uses synthetic and experimental

xix

methods to find a unified theory of mind and brain and applies the outcomes to novel real-world technologies and quality of life enhancing applications.

Patrice L. (Tamar) WEISS Department of Occupational Therapy, University of Haifa, Haifa, Israel Prof. Weiss, an occupational therapist with graduate training in kinesiology, physiology and biomedical engineering, founded the Laboratory for Innovations in Rehabilitation Technology with the objective of providing a conceptual and experimental environment for the formulation and implementation of research related to the development and evaluation of innovative technologies for rehabilitation.

Carolee J. WINSTEIN Division of Biokinesiology and Physical Therapy at the School of Dentistry, University of Southern California, Los Angeles, USA Carolee J. Winstein, PhD, PT, FAPTA is Professor and Director of Research in Biokinesiology and Physical Therapy at the University of Southern California. She runs an interdisciplinary research program focused on understanding control, rehabilitation and recovery of goal-directed movements that emerge from a dynamic brain-behavior system in brain-damaged conditions.

Carla Silvana ZUCCONI Laboratory of Robotics and Kinematics, I.R.C.C.S. San Camillo, Venice, Italy Carla S. Zucconi is a Physical Therapist. She received a Master’s Degree in Science of Rehabilitation Health Profession from the University of Padova. Her research focuses on the application of virtual reality and robotic systems in motor rehabilitation, with specific interest in the motor learning principles involved in human-machine interaction.

This page intentionally left blank

xxi

CONTENTS Introduction, Emily A. Keshner and W. Zev Rymer

v

Contributors

ix

Section I.

Advanced Technologies in Rehabilitation: An Introduction

Chapter 1.

Rehabilitation as Empowerment: The Role of Advanced Technologies G. Riva and A. Gaggioli

3

Section II. Training and Technology as an Aid in Functional Gains Chapter 2.

Robotic Assistance for Upper Extremity Training after Stroke D.J. Reinkensmeyer

25

Chapter 3.

Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS A. Frisoli, M. Bergamasco, M.C. Carboncini and B. Rossi

40

Chapter 4.

Assessment and Treatment of the Upper Limb by Means of Virtual Reality in Post-Stroke Patients L. Piron, A. Turolla, M. Agostini, C. Zucconi, P. Tonin, F. Piccione and M. Dam

55

Section III. Rules that Govern Recovery of Function Chapter 5.

The Rehabilitation Gaming System: A Review M.S. Cameirão, S. Bermúdez i Badia, E. Duarte Oller and P.F.M.J. Verschure

Chapter 6.

Virtual Reality and Gaming Systems to Improve Walking and Mobility for People with Musculoskeletal and Neuromuscular Conditions J.E. Deutsch

Chapter 7.

Chapter 8.

Chapter 9.

Virtual Reality Environments to Enhance Upper Limb Functional Recovery in Patients with Hemiparesis M.F. Levin, L.A.M. Knaut, E.C. Magdalon and S. Subramanian Virtual Reality to Maximize Function for Hand and Arm Rehabilitation: Exploration of Neural Mechanisms A.S. Merians, E. Tunik and S.V. Adamovich Robot Therapy for Stroke Survivors: Proprioceptive Training and Regulation of Assistance V. Sanguineti, M. Casadio, E. Vergaro, V. Squeri, P. Giannoni and P.G. Morasso

65

84

94

109

126

xxii

Section IV. Using the Body’s Own Signals to Augment Therapeutic Gains Chapter 10. Advances in Wearable Technology for Rehabilitation P. Bonato

145

Chapter 11. Brain-Computer Interfaces and Neurorehabilitation R. Carabalona, P. Castiglioni and F. Gramatica

160

Section V.

Technology Incorporates Cognition and Action

Chapter 12. Why Is Music Effective in Rehabilitation? A. Antonietti

179

Chapter 13. Computer-Guided Mental Practice in Neurorehabilitation A. Gaggioli, F. Morganti, A. Meneghini, I. Pozzato, G. Greggio, M. Pigatto and G. Riva

195

Chapter 14. Postural and Spatial Orientation Driven by Virtual Reality E.A. Keshner and R.V. Kenyon

209

Section VI. Technology Enhances the Impact of Rehabilitation Programs Chapter 15. Telerehabilitation: Enabling the Remote Delivery of Healthcare, Rehabilitation, and Self Management D.M. Brennan, S. Mawson and S. Brownsell Chapter 16. Socially Assistive Robotics for Stroke and Mild TBI Rehabilitation M. Matarić, A. Tapus, C. Winstein and J. Eriksson

231 249

Chapter 17. Moving Beyond Single User, Local Virtual Environments for Rehabilitation P.L. Weiss and E. Klinger

263

Subject Index

279

Author Index

281

SECTION I ADVANCED TECHNOLOGIES IN REHABILITATION: AN INTRODUCTION

The field of physical medicine and rehabilitation has placed increasing emphasis on the construct of empowerment as a potential conceptual cornerstone of identity, and hence, a critical variable in outcome research. In referring to people with disabilities who have rehabilitation service needs, terminology has shifted from patient, to consumer, and most recently to constituent. This reflects a paradigmatic shift from a focus on deficits and dependence toward an emphasis on assets and independence. Empowerment theory provides a useful framework for guiding our work as the field becomes more constituent based. Zimmerman & Warschausky, 1998

This page intentionally left blank

Advanced Technologies in Rehabilitation A. Gaggioli et al. (Eds.) IOS Press, 2009 © 2009 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-60750-018-6-3

3

Rehabilitation as Empowerment: The Role of Advanced Technologies a

Giuseppe RIVAa,b, Andrea GAGGIOLIa,b Applied Technology for Neuro-Psychology Lab., Istituto Auxologico Italiano, Milan, Italy b ICE-NET Lab., Catholic University of Sacred Heart, Milan, Italy Abstract. Rehabilitation is placing increasing emphasis on the construct of empowerment as the final goal of any treatment approach. This reflects a shift in focus from deficits and dependence to assets and independence. According to this approach, rehabilitation should aim to improve the quality of the life of the individual by means of effective support to his/her activity and interaction. Here we suggest that advanced technologies can play a significant role in this process. By enhancing the experienced level of “Presence” - the non-mediated perception of successfully transforming intentions into action - these emerging technologies can foster optimal experiences (Flow) and support the empowerment process. Finally, we describe the “NeuroVR” system (http://www.neurovr.org) as an example of how advanced technologies can be used to support Presence and Flow in the rehabilitation process. Keywords. Empowerment, Rehabilitation, Presence, Virtual Reality, NeuroVR

Introduction The field of rehabilitation is placing increasing emphasis on the construct of empowerment as a critical element of any treatment strategy. This construct integrates perceptions of personal control, participation with others to achieve goals and an awareness of the factors that hinder or enhance one’s efforts to exert control in one's life [1, 2]. The emphasis on empowerment reflects a critical shift in rehabilitation: from a focus on deficits and dependence toward an emphasis on assets and independence. The International Classification of Functioning, Disability and Health (ICF) of the World Health Organization [3] defines disability as a “condition in which people are temporarily or definitively unable to perform an activity in the correct manner and/or at a level generally considered ‘normal’ for the human being.” In this definition the focus is not on deficits but on assets: a person is disabled when he/she is not able to fully exploit his/her relationship with everyday contexts [4]. In this chapter we suggest that the new emerging technologies discussed in the book – with particular reference to robotics and virtual reality - have the right features for improving the rehabilitation process. These technologies can improve the quality of life of the disabled individual through an effective support of his/her activity and interaction [5].

4

1.

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

Empowerment in Rehabilitation

“Empowerment” is a term that is becoming very popular in rehabilitation services. More and more rehabilitation programs claim to “empower” their clients. However, in practice, few researchers and clinicians have specifically targeted aspects of empowerment in rehabilitation programs. The main issue up until now has been the lack of guidelines to assess and enhance empowerment during the rehabilitation process. In general, empowerment refers to processes and outcomes relating to issues of control, critical awareness, and participation [2]. How does this apply to rehabilitation? According to Zimmerman and Warschausky [6] empowerment in rehabilitation should provide a sense of and motivation to control and the knowledge and skills to help the patient to adapt to and influence his/her own environment. This approach underlines the role of participation and control, supporting wellness versus illness, and competence versus deficiency. In this view, the final goal of rehabilitation is to help patients to become as independent as possible, by developing skills for changing conditions that pose barriers in their lives. To put this approach into practice, the next step is the definition of clear empowerment outcomes. Table 1 provides a brief comparison of empowering processes, goals and outcomes across the different levels of analysis (intrapersonal, interactional and social) involved in a typical rehabilitation program. Our analysis will focus on the first two levels – intrapersonal and interactional. We believe that it is at these levels that emerging technologies can play a critical role. The intrapersonal component refers to how the patients think about themselves[6]. At the intrapersonal level, the main goals of the rehabilitation process are to help the individual in gaining control over his/her life. Specifically, the patient needs to recover his/her decision-making power through full access to information and resources. How is it possible to evaluate the success of an intrapersonal rehabilitation strategy? According to the psychological literature, the key outcome variables are [6]: • •

self-efficacy: this refers to perceptions about one's ability to achieve the desired outcomes; sense of control: this refers to perceptions about one's ability to regulate and manage the different domains of his/her personal experience. Table 1. Empowerment outcomes in rehabilitation

Levels Process Patient (Intrapersonal) Receiving help from therapist to gain control over his/her life

Therapist/Caregiver (Interactional)

Health Care Institution/System (Social)

Helping patients and their family to evaluate/understand their actual skills/situation Helping patients gain control over their lives Providing opportunities for patients to develop and practice skills

Goals To have decision-making power

Outcomes Self-efficacy

To have access to information and resources To change perceptions of patient's competency and capacity to act

Sense of control Critical awareness

Not to feel alone; to feel part of a group To effect change in one's life and one's community

Participatory behaviors Effective resource management

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

5

The interactional component refers to how people think about and relate to their social environment. This component of any empowering rehabilitation strategy involves the transactions between people and the environments (family, clinical setting, work, etc.) that they are involved in. On the one hand, it includes the decision-making and problem-solving skills necessary to actively engage in one's environment. On the other, it includes the ability to mobilize and obtain resources. Again, how is it possible to evaluate the success of an interactional rehabilitation strategy? According to the psychological literature, the key outcome variables are [6]: • •

critical awareness: this refers to one's understanding of the resources needed to achieve a desired goal, knowledge of how to acquire those resources, and skills for managing resources once they are obtained; participatory behaviors: this refers to one’s social activities affording the opportunity for individual participation.

An increasing number of empirical studies are addressing empowerment in rehabilitation. These studies focus on a variety of participatory programs targeting a broad range of population groups and goals. Few authors, however, have investigated the role of technology in this process. In this chapter, we argue that the advanced technologies presented in this book can enhance this process by supporting the experience of “Presence”, defined as the “feeling of being there” [7]. The creation of a feeling of Presence can help patients to cope with their context in an effective and transparent way. In this view, technologies are used for triggering a broad empowerment process within the optimal experience induced by a high sense of Presence [8]. 2.

Advanced Technologies in Rehabilitation: The Role of Presence

In recent years it has been possible to identify a clear trend in the design and development of rehabilitation technologies: the shift from a general user-centered approach to a specific activity-centered approach. In this last perspective, the goal of technology should be the improvement of the quality of life of the individual, through an effective support of his/her activity and interaction [4]. In this vision, “…if a person is able to write a paper with a pen and another person is limited in the pen use but is able to write the same paper using a computer keyboard, none of them is defined as disabled. On the contrary if both of them will be in a condition in which the tool, that allows them to write the paper, is not available in a specific moment they will be both disabled in performing the activity.” (p. 286). This “compensatory” approach in rehabilitation is usually divided [9] into personoriented and environmentally oriented interventions (see Figure 1).

6

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

Figure 1. The role of advanced technologies in rehabilitation

Person-oriented interventions include the recruitment of alternate cognitive or physical resources to achieve a desired outcome. Environmentally oriented interventions offer external cues to the subject in order to improve his/her handling of the activity [10]. As noted by Crosson and colleagues [10], environmentally oriented interventions may be: “the only practical means for dealing with neurologically based deficits. Although not ideal, external modification can be effective in many circumstances” (p. 53). This viewpoint stresses the need of developing technological tools for providing alternative affordances in planning specific activities. Moreover, as noted by Kirsk and colleagues [9], any rehabilitation device has to support activity in a transparent way: “In regard to device features, an ideal intervention will be one that is minimally intrusive, provides assistance without assuming unnecessary control, and does not demand of the user an uncharacteristic level of comfort with technological aids.” (p. 201). In summary, rehabilitation technologies become empowerment tools when they help people in coping with their context in an effective and transparent way. But how can we assess whether rehabilitation technologies meet these requirements? A possible answer to this question is “through Presence”. We will detail this point in the next paragraph. 2.1 Presence: a first definition The term “Presence” entered the general scientific debate in 1992 when Sheridan and Furness used it in the title of a new journal dedicated to the study of virtual reality systems and teleoperations: Presence, Teleoperators and Virtual Environments. In the first issue, Sheridan clearly refers to Presence as an experience elicited by technology use [11]: the effect felt when controlling real world objects remotely as well as the effect people feel when they interact with and immerse themselves in virtual environments.

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

7

However, as remarked by Biocca [12], and agreed upon by most researchers in the area, “while the design of virtual reality technology has brought the theoretical issue of Presence to the fore, few theorists argue that the experience of Presence suddenly emerged with the arrival of virtual reality.” Rather, as suggested by Loomis [13], Presence may be described as a basic state of consciousness: the attribution of sensation to some distal stimulus, or more broadly to some environment. Due to the complexity of the topic, and the interest in this concept, different conceptualizations of Presence have been proposed in the literature. A first definition of “Presence” is introduced by the International Society of Presence Research (ISPR). ISPR researchers define “Presence” (a shortened version of the term “telePresence”) as: “a psychological state in which even though part or all of an individual’s current experience is generated by and/or filtered through human-made technology, part or all of the individual’s perception fails to accurately acknowledge the role of the technology in the experience” [14]. This definition suggests that rehabilitation technology should provide a strong feeling of Presence: the more the user experiences Presence in using a rehabilitation technology, the more it is transparent to the user, the more it helps the user in coping with his/her context in an effective way . Nevertheless, the above definition has two limitations. First, what is Presence for? Why do we experience Presence? As underlined by Lee [15]: “Presence scholars, may find it surprising and even disturbing that there have been limited attempts to explain the fundamental reason why human beings can feel Presence when they use media and/or simulation technologies.” (p. 496). Second, is Presence related to media only? As commented by Biocca [12], and agreed by most researchers in the area: “while the design of virtual reality technology has brought the theoretical issue of Presence to the fore, few theorists argue that the experience of Presence suddenly emerged with the arrival of virtual reality.” (online: http://jcmc.indiana.edu/vol3/issue2/biocca2.html) Recent insights from cognitive sciences suggest that Presence is a neuropsychological process that results in a sense of agency and control [16-18]. For instance, Slater suggested that presence is a selection mechanism that organizes the stream of sensory data into an environmental gestalt or perceptual hypothesis about current environment [19, 20]. Within this framework, supported by ecological/ethnographic studies [21-28], any rehabilitation technology, virtual or real, does not provide undifferentiated information or ready-made objects in the same way for everyone. It offers different opportunities and creates different levels of Presence according to its ability in supporting the users' intentions. 2.2 Presence: A Second Definition Recent findings in cognitive science suggest that Presence is a neuropsychological phenomenon, evolved from the interplay of our biological and cultural inheritance,

8

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

whose goal is the enaction of volition: Presence is the perception of successfully transforming intentions into action (enaction). Recent research by Haggard and Clark [29, 30] on voluntary and involuntary movements, provides direct support for the existence of a specific cognitive process binding intentions with actions. In their words [30]: “Taken as a whole, these results suggest that the brain contains a specific cognitive module that binds intentional actions to their effects to construct a coherent conscious experience of our own agency.” (p. 385). Varela and colleagues [31] define “enaction” in terms of two intertwined and reciprocal factors: first, the historical transformations which generate emergent regularities in the actor's embodiment; second, the influence of an actor's embodiment in determining the trajectory of behaviors. As suggested by Whitaker [32] these two aspects reflect two different usages of the English verb “enact”. On the one hand is “to enact” in the sense of “to specify, to legislate, to bring forth something new and determining of the future”, as in a government enacting a new law. On the other is “to enact” in the sense of “to portray, to bring forth something already given and determinant of the present”, as in a stage actor enacting a role. In line with these two meanings, Presence has a dual role: -

First, Presence "locates" the self in an external physical and/or cultural space: the Self is “present” in a space if he/she can act in it Second, Presence provides feedback to the Self about the status of its activity: the Self perceives the variations in Presence and tunes its activity accordingly.

First, we suggest that the ability to feel “present” in the interaction with a rehabilitation technology - an artifact - basically does not differ from the ability to feel “present” in our body. Within this view, “being present” during agency means that 1) the individual is able to successfully enact his/her intentions 2) the individual is able to locate him/herself in the physical and cultural space in which the action occurs. When the subject is present during a mediated action (that is, an action supported by a tool), he/she incorporates the tool in his/her peri-personal space, extending the action potential of the body into virtual space [33]. In other words, through the successful enaction of the actor’s intentions using the tool, the subject becomes “present” in the tool. The process of Presence can be described as a sophisticated but covert form of monitoring action and experience, transparent to the self but critical for its existence. The result of this process is a sense of agency: the feeling of being both the author and the owner of one’s own actions. The more intense the feeling of Presence, the higher the quality of experience perceived during the action[34]. However, the agent directly perceives only the variations in the level of Presence: breakdowns and optimal experiences [16]. Why do we monitor the level of Presence? Our hypothesis is that this high-level process has evolved to control the quality of action and behaviors. According to Csikszentmihalyi [35, 36], individuals preferentially engage in opportunities for action associated with a positive, complex and rewarding state of consciousness, defined by him as “optimal experience” or “Flow”. The key feature of this experience is the perceived balance between great environmental opportunities for action (challenges) and adequate personal resources in facing them (skills). Additional characteristics are deep concentration, clear rules for and unambiguous feedback from

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

9

the task at hand, loss of self-consciousness, control of one’s actions and environment, positive affect and intrinsic motivation. Displays of optimal experience can be associated with various daily activities, provided that individuals perceive them as complex opportunities for action and involvement. An example of Flow is the case where a professional athlete is playing exceptionally well (positive emotion) and achieves a state of mind where nothing else is attended to but the game (high level of Presence). From the phenomenological viewpoint, both Presence and Flow are described as absorbing states, characterized by a merging of action and awareness, loss of self-consciousness, a feeling of being transported into another reality, and an altered perception of time. Further, both Presence and optimal experience are associated with high involvement, focused attention and high concentration on the ongoing activity. Starting from these theoretical premises, can we design rehabilitation technologies that elicit a state of Flow by activating a high level of Presence (maximal Presence) [4, 37, 38]? This question will be addressed in the following section. 2.3 The Presence Levels How can we achieve a high level of Presence during interaction with a rehabilitation technology? The answer to this question requires a better understanding of what intentions are. According to folk psychology, the intention of an agent performing an action is his/her specific purpose in doing so. However, the latest cognitive studies clearly show that any action is the result of a complex intentional chain that cannot be analyzed at a single level [39-41]. Pacherie identifies three different “levels” or “forms” of intentions, characterized by different roles and contents: distal intentions (D-intentions), proximal intentions (Pintentions) and motor intentions (M-intentions): •

•

•

D-intentions (Future-directed intentions). These high-level intentions act both as intra- and interpersonal coordinators, and as prompters of practical reasoning about means and plans: “helping my elderly father” is a D-intention, the object that drives the activity “finding a nurse” (see Figure 2) of the subject. P-intentions (Present-directed intentions). These intentions are responsible for high-level (conscious) forms of guidance and monitoring. They have to ensure that the imagined actions become current through situational control of their unfolding: “posting a request for a nurse” is a P-intention driving the action “going to the hospital’s bulletin board (see Figure 2), M-intentions (Motor intentions). These intentions are responsible for low-level (covert) forms of guidance and monitoring: we may not be aware of them and have only partial access to their content. Further, their contents are not propositional: in the operation “putting the post on the board” (see Figure 2), the motor representations required to move the arm are M-intentions.

Any intentional level has its own role: the rational (D-intentions), situational (PIntention) and motor (M-Intention) guidance and control of action. They form an intentional cascade [40, 41] in which higher intentions generate lower intentions.

10

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

Figure 2. The intentional cascade

We previously defined Presence as the perception of successfully transforming intentions into action (enaction). However, even if we experience a single feeling of Presence during the enaction of our intentions, the three-level structure of the intentional cascade suggests that Presence - on the process side - can be divided into three different layers or sub-processes (for a broader and more in-depth description see [21, 42]), described in Figure 3: -

-

-

Extended Presence (D-Intentions/Activities): The role of “Extended Presence” is to verify the relevance to the Self of possible/future events in the external world (Self vs. possible/future external world). The more the Self is able to identify mediated affordances (that cannot be enacted directly) in the external world, the higher the level of extended Presence will be. Core Presence (P-Intentions/Actions): This can be described as the activity of selective attention made by the Self on perceptions (Self vs. present external world). The more the Self is able to identify direct affordances (that can be enacted directly with a movement of the body) in the external world, the higher the level of core Presence will be. Proto Presence (M-Intentions/Operations): This is the process of internal/external separation related to the level of perception-action coupling (Self vs. non-Self). The more the Self is able to use the body for enacting direct affordances in the external world, the higher the level of proto Presence will be.

As underlined by Dillon and colleagues [43], converging lines of evidence from diverse perspectives and methodologies support this three-layered view of Presence. In their analysis they identify three dimensions common to all the different perspectives, relating to a "spatial" dimension (M-intentions), a dimension relating to how consistent the media experience is with the real world, "naturalness" (P-intentions), and an "engagement" dimension (D-intentions).

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

11

Figure 3. Activity and Presence

The role of the different layers will be related to the complexity of the activity done: the more complex the activity, the more layers will be needed to produce a high level of Presence (Figure 3). At the lower level – operations – proto Presence is enough to induce a satisfying feeling of Presence. At the higher level – activity – the media experience has to support all three layers. As suggested by Juarrero [44] high level intentions (Future Intentions/Objects) channel future deliberation by narrowing the scope of alternatives to be subsequently considered (cognitive reparsing). In practice, once the subject forms an intention, not every logical or physically possible alternative remains open, and those that do are encountered differently: once I decide to do A, non-A is no longer a viable alternative and should it happen, I will consider non-A as a breakdown [45]. 2.4 How to design rehabilitation technologies that foster Presence and Flow This perspective allows us to predict under which mediated situations the feeling of Presence can be enhanced or reduced. First, minimal Presence results from an almost complete lack of integration of the three layers discussed above, such as is the case when attention is mostly directed towards contents of extended consciousness that are unrelated to the present external environment (e.g., I’m in the office trying to write a letter but I’m thinking about how to find a nurse for my father). By the same reasoning, maximal Presence arises when proto Presence, core Presence and extended Presence are focused on the same external situation or activity [28]. Maximal Presence thus results from the combination of all three layers with a tight focus on the same content. This experience is supported by a rehabilitation technology that offers an optimal combination of form and content, able to support the activity of the user in a meaningful way. The concepts described above are summarized by the following points:

12

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

1) The lower the level of activity, the easier it is to induce maximal Presence. The object of an activity is wider and less targeted than the goal of an action. So, its identification and support is more difficult for the designer of a rehabilitation technology. Furthermore, the easiest level to support is the operation. In fact, its conditions are more “objective” and predictable, being related to the characteristics (constraints and affordances) of the artifact used: it is easier to automatically open a door in a virtual environment than to help the user in finding the right path for the exit. At the lower level – operations – proto Presence is enough to induce a satisfying feeling of Presence. At the higher level – activity – the media experience has to support all the three levels. 2) We have maximal Presence when the environment is able to support the full intentional chain of the user: this can explain i) the success of the Nintendo Wii over competing consoles (it is the only one to fully support M-intentions); ii) the need for a long-term goal to induce a high level of Presence after many experiences of the same rehabilitation technology. 3) Subjects with different intentions will not experience the same level of Presence, even when using the same rehabilitation technology: this means that understanding and supporting the intentions of the user will improve his/her Presence during the interaction with the technology. 4) Action is more important than perception: I’m more present in a perceptually poor virtual environment (e.g. a textual MUD) where I can act in many different ways than in a real-like virtual environment where I cannot do anything. 3.

Transformation of Flow in Rehabilitation using Advanced Technologies

As we have seen previously, authentic rehabilitation implies the active participation of patients in their contexts, their exposure to opportunities for action and development and their freedom to select the opportunities which they perceive as most challenging and meaningful for the subject [46, 47]. According to this vision, a critical asset potentially offered by advanced technologies to the rehabilitation process is that they can foster optimal (Flow) experiences triggering the empowerment [48]. Optimal experiences promote individual development. As underlined by Massimini and Delle Fave, [49]: “To replicate it, a person will search for increasingly complex challenges in the associated activities and will improve his or her skill, accordingly. This process has been defined as cultivation; it fosters the growth of complexity not only in the performance of Flow activities but in individual behavior as a whole.” (p. 28). This process can be also activated after a major trauma. As noted by Delle Fave [50], to cope with dramatic changes in daily life and to access environmental opportunities for action, individuals may develop a strategy defined as transformation of Flow: the ability of the subject to use an optimal experience for identifying and exploiting new and unexpected resources and sources of involvement.

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

13

Figure 4. Transformation of Flow

We hypothesize that it is possible to use advanced technologies to activate a transformation of Flow to be used for rehabilitative purposes [8]. The proposed approach is the following (Figure 4): first, to identify an enriched environment that contains functional real-world demands; second, using the technology to enhance the level of Presence of the subject in the environment and to induce an optimal experience; third, allowing cultivation, by linking this optimal experience to the actual experience of the subject. It is well known that process of sequential development of the brain and the sequential development of function, is guided by experience. The brain develops and modifies itself in response to experience. Neurons and neuronal connections (synapses) change in an activity-dependent fashion. Thanks to specific experiences, the brain can even relocate functions to new areas if the primary site is destroyed [51, 52]. For example, stroke victims can gain control over movements with therapy designed to disable their abler body (Constraint-Induced Movement therapy) forcing the brain to establish new circuits to control the areas with little or no control [53, 54]. The only continuing limitation seems to be that some areas of the brain are only open to maximum flexibility during short periods of life. Apparently the transformation of Flow approach could be able to open new plasticity phases, thus improving the possibility of recovery of the subject. Below are reported some examples of technology-driven transformation of Flow. 3.1 Multi-Sensory Environment A first example of the proposed approach is the Multi-Sensory Environment (MSE) method used in the rehabilitation of neurological disabilities, learning disabilities and older people with dementia [55-57]. The concept of multi-sensory environments (Snoezelen) was developed in the 1980’s at the Haarendael Institute, Holland: MSEs are purpose-built units or rooms using advanced sensory stimulating equipment that targets the five senses of sight, hearing, touch, taste and smell. Their goal is the stimulation of the primary senses to generate pleasurable sensory experiences in an atmosphere of trust and relaxation without the need for intellectual activity. Exposure to an MSE occurs through the agency of the caregiver, nurse or therapist who facilitates the development of a relaxing and supportive environment [58]. The results from a randomized controlled trial (N = 50) showed the efficacy of this approach in the treatment of older people with dementia [59]. In particular, the use of a Multi-Sensory Environment appeared to have a greater influence on aspects of communication in comparison to one-to-one activity and led to improvements in behavior and mood at a four-week follow-up.

14

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

Moreover, positive results were obtained in the treatment of children recovering from severe brain injury [60] and in the management of Rett disorder [61]. As underlined by Collier [55], the best results in using MSEs are achieved under transformation of Flow (p. 364): “…the MSE should include an appropriate level of stimulation that challenges the individual to reach their maximum potential (sensory stimulation versus sensory deprivation). The activity should be designed to address individual sensory needs, such as offering a stronger stimulus if initial attempts are unnoticed, and be offered alongside familiar activities and routines to enhance sensory awareness. The activity should occur on a regular basis and offer a ‘just right challenge’ as the person with brain injury will find it easier to cope with the demands of the environment if adequate stimulation is provided… Finally, if the complexity of the activity, individual needs, and MSE demands are matched, engagement in this activity may be achieved.” 3.2 Robots The development of robots that interact socially with people and assist them in everyday life has been a long-term goal of modern science [62, 63]. Within this broad area of research, robotic psychology and robotherapy focus on the psychological meaning of person–robotic creature communication and its intertwining with psychophysiological and social elements. As suggested by Libin & Libin [63]: “Robotherapy is defined as a framework of human–robot interactions aimed at the reconstruction of a person’s negative experiences through the development of coping skills, mediated by technological tools in order to provide a platform for building new positive experiences.” (p.370). Recent research suggests that now, after more than 25 years of research, low-level information, such as animacy, contingency, and visual appearance, can trigger longterm bonding and socialization both in children [64] and in the elderly [65]: rather than losing interest, the interaction between users and the robot improved over time. Interestingly, the results highlighted the particularly important role that haptic behaviors (motor intentions) played in the socialization process [64]: the introduction of a simple touch-based contingency had a breakthrough effect in the development of social behaviors toward the robot. Also, as predicted by our model, the ability to address all levels of Presence in the interaction with the rehabilitative robot helps in maintaining patients' interest high during execution of the assigned tasks [66]. 3.3 Virtual Reality The basis of the Virtual Reality (VR) idea is that a computer can synthesize a threedimensional (3D) graphical environment from numerical data [67]. Using visual, aural or haptic devices, the human operator can experience the environment as if it were a part of the world. A VR system is the combination of the hardware and software that enables developers to create VR applications. The hardware components receive input from user-controlled devices and convey multi-sensory output to create the illusion of a virtual world. The software component of a VR system manages the hardware that makes up VR system.

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

15

Many researches using VR underline the link between this technology and optimal experiences. However, given the limited space available, we focus on the ones that are most relevant to the contents of this chapter. A first set of results comes from the work of Gaggioli [46, 47]. Gaggioli compared the experience reported by a user immersed in a virtual environment with the experience reported by the same individual during other daily situations. To assess the quality of experience the author used a procedure called Experience Sampling Method (ESM), which is based on repeated on-line assessments of the external situation and personal states of consciousness [47]. Results showed that the VR experience was the activity associated with the highest level of optimal experience (22% of self-reports). Reading, TV viewing and using other media – both in the context of learning and of leisure activities – obtained lower percentages of optimal experiences (15%, 8% and 19% of self-reports respectively). To verify the link between advanced technologies and optimal experiences, the “V-STORE Project” investigated the quality of experience and the feeling of Presence in a group of 10 patients with Frontal Lobe Syndrome involved in VR-based cognitive rehabilitation [68].They used the ITC-Sense of Presence Inventory [69] to evaluate the feeling of Presence induced by the VR sessions. Findings highlighted the association of VR sessions with both positive affect and a high level of Presence. Miller and Reid [70] investigated the personal experiences of children with cerebral palsy engaging in a virtual reality play intervention program. The results show that participants experienced a sense of control and mastery over the virtual environment. Moreover, they perceived experiencing Flow and both peers and family reported perceived physical changes and increased social acceptance. These results were confirmed in two later studies with the same population group [71, 72]. The other hypothesis we suggested in this chapter is that the transformation of Flow may also exploit the plasticity of the brain producing some form of functional reorganization [73]. Optale and his team [74-76] investigated the experience of subjects with male erectile disorders engaging in a virtual reality rehabilitative experience. The results obtained - 30 out of 36 patients with psychological erectile dysfunction and 28 out of 37 clients with premature ejaculation maintained partial or complete positive response after 6-month follow up - showed that this approach was able to hasten the healing process and reduce dropouts. However, the most interesting part of the work is the PET analysis carried out in the study. Optale used PET scans to analyze regional brain metabolism changes from baseline to follow-up in the experimental sample [77]. The analysis of the scans showed, after the VR protocol, different metabolic changes in specific areas of the brain connected with the erection mechanism. Recent experimental results from the work of Hoffman and his group in the treatment of chronic pain [78-81] might also be considered as fostering this vision. Hoffman and colleagues verified the efficacy of VR as an advanced distraction tool [82] in different controlled studies. The result showed dramatic drops in pain ratings during VR compared to controls [83]. Further, using a functional magnetic resonance imaging (fMRI) scanner they measured pain-related brain activity for each participant when virtual reality was not present and when virtual reality was present (order randomized). The team studied five regions of the brain known to be associated with pain processing - the anterior cingulate cortex, primary and secondary somatosensory cortex, insula, and thalamus - and found that during VR the activity in all regions showed significant reductions [84]. In particular, the results showed direct modulation

16

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

of human brain pain responses by VR distraction: the amount of reduction in painrelated brain activity ranged from 50 percent to 97 percent. Interestingly, as predicted by our model, the level of pain reduction was directly correlated to the level of Presence experienced in VR [79, 85]: the more the Presence, the less the pain. 4.

Transformation of Flow in Virtual Reality: The NeuroVR project

Although VR certainly has potential as a rehabilitation technology [86] [87, 88], most of the actual applications in this area are still in the laboratory or at investigation stage. In a recent review [89], Riva identified four major issues that limit the use of VR in this field: • • • •

the lack of standardization in VR hardware and software, and the limited possibility of tailoring virtual environments (VEs) to the specific requirements of the clinical or experimental setting; the low availability of standardized protocols that can be shared by the community of researchers; the high costs (up to 200,000 US$) required for designing and testing a clinical VR application; most VEs in use today are not user-friendly; expensive technical support or continual maintenance is often required.

To address these challenges, we developed NeuroVR (http://www.neurovr.org) in 2007 – a free virtual reality platform based on open-source elements [90]. The software allows non-expert users to adapt the content of 14 pre-designed virtual environments to the specific needs of the clinical or experimental setting. The key characteristics that make NeuroVR suitable as rehabilitation tool are the high level of control over interaction with the tool, and the enriched experience provided to the patient. These features transform NeuroVR into an “empowering environment”, a special, sheltered setting where patients can start to explore and act without feeling threatened. Nothing the patient fears can “really” happen to them in VR. With such assurance, they can freely explore, experiment, feel, live, and experience feelings and/or thoughts. Following the feedback of over 700 users who downloaded the first version, we developed a new version – NeuroVR 1.5 – that improves the possibility for the therapist to enhance the patient’s feeling of familiarity and intimacy with the virtual scene by using external sounds, photos or videos. The NeuroVR Editor is built using Python scripts that create a custom graphical user interface for Blender. The Pythonbased GUI allows all the richness and complexity of the Blender suite to be hidden, thus revealing only the controls needed to customize existing scenes and to create the proper files to be viewed in the player. NeuroVR Player leverages two major opensource projects in the VR field: Delta3D (http://www.delta3d.org) and OpenSceneGraph (http:// www.openscenegraph.org). Both are building components that the NeuroVR player integrates with an ad-hoc code to handle the simulations. NeuroVR software was designed with the goal of enabling therapists to create virtual environments that can enhance the feeling of Presence and support the transformation of Flow. To accomplish this goal, the design process followed the requirements derived from the three-layered theory of Presence summarized in par. 3.4 and developed in [8] and [19]:

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

17

1) The lower the level of activity, the easier it is to induce maximal Presence. The object of an activity is wider and less targeted than the goal of an action. The virtual exercises developed with NeuroVR can simulate a number of finegrained activities, such as opening the fridge, grabbing the water and closing the fridge. These activities may in turn broken down to an even finer level – depending on the goals and the complexity of the exercise. 2) We have maximal Presence when the environment is able to support the full intentional chain of the user. The virtual environments developed using NeuroVR support the three hierarchical levels indicated by the Presence theory [19]: - Extended Presence: NeuroVR allows the presentation of mediated affordances that supports the Self in generating complex action plans; - Core Presence: the VE can be programmed to present the patient with direct affordances. For instance, it is possible to program the appearance/disappearance of virtual objects/images that that trigger the attention of the user. These objects/images can be activated by user’s actions and behavior or by therapist’s commands. - Proto Presence: the combined use of sensors and actuators supports perception-action coupling and permits the patient to use his/her body for enacting direct affordances in the virtual environment. Movements can in turn be captured and recorded by means of different input devices and wearable sensors (i.e. head tracking). 3) Subjects with different intentions will not experience the same level of Presence, even when using the same rehabilitation technology. Since the reduction of psychomotor performance can vary significantly among patients suffering from neurological damages, complexity of virtual exercises can be tailored to match the level of impairment of each patient. In this way, even patients with a low level of cognitive functioning can successfully accomplish virtual exercises, thereby increasing their feeling of presence, empowerment and motivation for therapy. 4) Action is more important than perception: NeuroVR was explicitly designed to find an optimal trade-off between perceptual realism and naturalness of interaction. Whilst finding this trade-off was not an easy task, the level of realism supported by the player is at least adequate to provide patients with the feeling of “being there”. As several Presence scholars have pointed out [24], [25], [32] the experience of Presence depends to a greater extent on the ability of a medium to support users’ action in a transparent and natural way, and is affected to a lesser extent by the quantity and quality of realism cues depicted in the simulated environment. 5.

Conclusions

The field of rehabilitation is placing increasing emphasis on the construct of empowerment as a critical element in any treatment approach. This construct integrates perceptions of personal control, participation with others to achieve goals, and a critical

18

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

awareness of the factors that hinder or enhance one's efforts to exert control in one's life [1, 2]. In this chapter we suggested that the new emerging technologies discussed in the book – from Virtual Reality to Robotics – have the right features to improve the course of rehabilitation. Specifically, we claim that they are able to improve the quality of life of the individual, by improving his/her level of “Presence”. To be precise, by enhancing the experienced level of Presence, emerging technologies can foster optimal (Flow) experiences triggering the empowerment process (transformation of Flow). The vision underlying this concept arises from “Positive Psychology” [91]. According to this vision, rehabilitation technologies should include positive peak experiences because they serve as triggers for a broader process of motivation and empowerment. Within this context, the transformation of Flow can be defined as a person's ability to draw upon an optimal experience and use it to marshal new and unexpected psychological resources and sources of involvement. Although different technologies can be used to achieve this goal, one of the most promising is Virtual Reality. On the one hand, it can be described as an advanced form of human–computer interface that allows the user to interact with and become immersed in a computer-generated environment in a naturalistic fashion. On the other, VR can also be considered as an advanced imaginal system: an experiential form of imagery that is as effective as reality in inducing emotional responses. To this end, we developed NeuroVR, an “empowering rehabilitation tool” that allows the creation of virtual environments where patients can start to explore and act without feeling threatened [92, 93]. Nothing the patient fears can “really” happen to them in VR. With such assurance, they can freely explore, experiment, feel, live, and experience feelings and/or thoughts. VR thus becomes a very useful intermediate step between the therapist’s office and the real world [94]. Clearly, further improving NeuroVR and building new virtual environments is important so that therapists will continue to investigate the application of these tools in their day-to-day clinical practice. In fact, in most circumstances, the clinical skills of the rehabilitator remain the key factor in the successful use of VR systems. Future research should also deepen analysis of the link between cognitive processes, motor activities, Presence and Flow. This will allow the creation of a new generation of rehabilitation technologies which are truly able to support the empowerment process. References [1] [2] [3] [4]

[5]

M.A. Zimmerman, Taking aim on empowerment research: On the distinction between individual and psychological conceptions. American Journal of Community Psychology, (1984), 18(1): p. 169-177. D.D. Perkins and M.A. Zimmerman, Empowerment theory: Research and applications. American Journal of Community Psychology, (1995), 23: p. 569–579. WHO, International Classification of Functioning, Disability and Health. 2004, World Health Organization. F. Morganti and G. Riva, Ambient Intelligence in Rehabilitation, in Ambient Intelligence: The evolution of technology, communication and cognition towards the future of the human-computer interaction, G. Riva, F. Davide, F. Vatalaro, and M. Alcañiz, Editors. 2004, IOS Press. On-line: http://www.emergingcommunication.com/volume6.html: Amsterdam. p. 283-295. R.L. Glueckauf, J.D. Whitton, and D.W. Nickelson, Telehealth: The new frontier in rehabilitation and health care, in Assistive technology: Matching device and consumer for successful rehabilitation, M.J. Scherer, Editor. 2002, American Psychological Association: Washington, DC. p. 197-213.

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies [6] [7]

[8] [9]

[10]

[11] [12]

[13] [14] [15] [16]

[17] [18] [19] [20] [21]

[22]

[23] [24] [25]

[26]

[27]

[28]

19

M.A. Zimmerman and S. Warschausky, Empowerment Theory for Rehabilitation Research: Conceptual and Methodological Issues. Rehabilitation Psychology, (1998), 43(1): p. 3-16. G. Riva, F. Davide, and W.A. IJsselsteijn, eds. Being There: Concepts, effects and measurements of user presence in synthetic environments. Emerging Communication: Studies on New Technologies and Practices in Communication, ed. G. Riva and F. Davide. 2003, Ios Press. Online: http://www.emergingcommunication.com/volume5.html: Amsterdam. G. Riva, G. Castelnuovo, and F. Mantovani, Transformation of flow in rehabilitation: the role of advanced communication technologies. Behavior Research Methods, (2006), 38(2): p. 237-44. L.N. Kirsch, M. Shenton, E. Spirl, J. Rowan, R. Simpson, D. Schreckenghost, and E.F. LoPresti, WebBased Assistive Technology Interventions for Cognitive Impairments After Traumatic Brain Injury: Selective Review and Two Case Studies. Rehabilitation Psychology, (2004), 49(3): p. 200-212. B. Crosson, P. Barco, C. Velozo, M.M. Bolesta, P.V. Cooper, D. Wefts, and T.C. Brobeck, Awareness and compensation in post-acute head injury rehabilitation. Journal of Head Trauma Rehabilitation, (1989), 4(46-54). T.B. Sheridan, Musing on telepresence and virtual presence. Presence, Teleoperators, and Virtual Environments, (1992), 1: p. 120-125. F. Biocca, The Cyborg's Dilemma: Progressive embodiment in virtual environments. Journal of Computer Mediated-Communication [On-line], (1997), 3(2): Online: http://jcmc.indiana.edu/vol3/issue2/biocca2.html. J.M. Loomis, Distal attribution and presence. Presence, Teleoperators, and Virtual Environments, (1992), 1(1): p. 113-118. International Society for Presence Research, The concept of presence: explication statement. 2000. K.M. Lee, Why Presence Occurs: Evolutionary Psychology, Media Equation, and Presence. Presence, (2004), 13(4): p. 494-505. G. Riva, Being-in-the-world-with: Presence meets Social and Cognitive Neuroscience, in From Communication to Presence: Cognition, Emotions and Culture towards the Ultimate Communicative Experience. Festschrift in honor of Luigi Anolli, G. Riva, M.T. Anguera, B.K. Wiederhold, and F. Mantovani, Editors. 2006, IOS Press. Online: http://www.emergingcommunication.com/volume8.html: Amsterdam. p. 47-80. G. Riva, F. Mantovani, and A. Gaggioli, Are robots present? From motor simulation to “being there”. Cyberpsychology & Behavior, (2008), 11(631-636). G. Riva, Virtual Reality and Telepresence. Science, (2007), 318(5854): p. 1240-1242. M. Slater, Presence and the sixth sense. Presence: Teleoperators, and Virtual Environments, (2002), 11(4): p. 435–439. M.V. Sanchez-Vives and M. Slater, From presence to consciousness through virtual reality. Nature Review Neuroscience, (2005), 6(4): p. 332-9. G. Riva, J.A. Waterworth, and E.L. Waterworth, The Layers of Presence: a bio-cultural approach to understanding presence in natural and mediated environments. Cyberpsychology & Behavior, (2004), 7(4): p. 405-419. G. Mantovani and G. Riva, "Real" presence: How different ontologies generate different criteria for presence, telepresence, and virtual presence. Presence, Teleoperators, and Virtual Environments, (1999), 8(5): p. 538-548. G. Mantovani and G. Riva, Building a bridge between different scientific communities: on Sheridan's eclectic ontology of presence. Presence: Teleoperators and Virtual Environments, (2001), 8: p. 538-548. J.J. Gibson, The ecological approach to visual perception. 1979, Hillsdale, NJ: Erlbaum. A. Spagnolli and L. Gamberini, A Place for Presence. Understanding the Human Involvement in Mediated Interactive Environments. PsychNology Journal, (2005), 3(1): p. 6-15. On-line: www.psychnology.org/article801.htm. A. Spagnolli, D. Varotto, and G. Mantovani, An ethnographic action-based approach to human experience in virtual environments. International Journal of Human-Computer Studies, (2003), 59(6): p. 797-822. L. Gamberini and A. Spagnolli, On the relationship between presence and usability: a situated, actionbased approach to virtual environments, in Being There: Concepts, Effects and Measurement of User Presence in Synthetic Environments, G. Riva, W.A. IJsselsteijn, and F. Davide, Editors. 2003, IOS Press: Amsterdam. p. 97-107. Online: http://www.emergingcommunication.com/volume5.html. J.A. Waterworth and E.L. Waterworth, Presence as a Dimension of Communication: Context of Use and the Person, in From Communication to Presence: Cognition, Emotions and Culture towards the Ultimate Communicative Experience, G. Riva, M.T. Anguera, B.K. Wiederhold, and F. Mantovani, Editors. 2006, IOS Press: Amsterdam. p. 80-95. Online: http://www.emergingcommunication.com/volume8.html.

20

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

[29] P. Haggard and S. Clark, Intentional action: conscious experience and neural prediction. Conscious Cogn, (2003), 12(4): p. 695-707. [30] P. Haggard, S. Clark, and J. Kalogeras, Voluntary action and conscious awareness. Nat Neurosci, (2002), 5(4): p. 382-5. [31] F.J. Varela, E. Thompson, and E. Rosch, The embodied mind: Cognitive science and human experience. 1991, Cambridge, MA: MIT Press. [32] R. Whitaker, Self-Organization, Autopoiesis, and Enterprises,. ACM SIGOIS Illuminations series, (1995: p. online: http://www.acm.org/sigs/siggroup/ois/auto/Main.html. [33] A. Clark, Natural Born Cyborgs: Minds, technologies, and the future of human intelligence. 2003, Oxford: Oxford University Press. [34] P. Zahoric and R.L. Jenison, Presence as being-in-the-world. Presence, Teleoperators, and Virtual Environments, (1998), 7(1): p. 78-89. [35] M. Csikszentmihalyi, Beyond Boredom and Anxiety. 1975, San Francisco: Jossey-Bass. [36] M. Csikszentmihalyi, Flow: The psychology of optimal experience. 1990, New York: HarperCollins. [37] G. Riva, The psychology of Ambient Intelligence: Activity, situation and presence, in Ambient Intelligence: The evolution of technology, communication and cognition towards the future of the human-computer interaction, G. Riva, F. Davide, F. Vatalaro, and M. Alcañiz, Editors. 2004, IOS Press. On-line: http://www.emergingcommunication.com/volume6.html: Amsterdam. p. 19-34. [38] E.L. Waterworth, M. Häggkvist, K. Jalkanen, S. Olsson, J.A. Waterworth, and W. H., The Exploratorium: An environment to explore your feelings. PsychNology Journal, (2003), 1(3): p. 189201. On-line: http://www.psychnology.org/File/PSYCHNOLOGY_JOURNAL_1_3_WATERWORTH.pdf. [39] J. Searle, Intentionality: An essay in the philosophy of mind. 1983, New York: Cambridge University Press. [40] E. Pacherie, Toward a dynamic theory of intentions, in Does consciousness cause behavior?, S. Pockett, W.P. Banks, and S. Gallagher, Editors. 2006, MIT Press: Cambridge, MA. p. 145-167. [41] E. Pacherie, The phenomenology of action: A conceptual framework. Cognition, (2008), 107(1): p. 179-217. [42] G. Riva, Enacting Interactivity: The Role of Presence, in Enacting Intersubjectivity: A cognitive and social perspective on the study of interactions, F. Morganti, A. Carassa, and G. Riva, Editors. 2008, IOS Press: Online: http://www.emergingcommunication.com/volume10.html: Amsterdam. p. 97-114. [43] C. Dillon, J. Freeman, and E. Keogh. Dimension of Presence and components of emotion. in Presence 2003. 2003. Aalborg, Denmark: ISPR. [44] A. Juarrero, Dynamics in action: Intentional behavior as a complex system. (1999. [45] M.E. Bratman, Shared cooperative activity. Philosophical Review, (1992), 101: p. 327-341. [46] A. Gaggioli, M. Bassi, and A. Delle Fave, Quality of Experience in Virtual Environments, in Being There: Concepts, effects and measurement of user presence in synthetic environment, G. Riva, W.A. IJsselsteijn, and F. Davide, Editors. 2003, Ios Press. Online: http://www.emergingcommunication.com/volume5.html: Amsterdam. p. 121-135. [47] A. Gaggioli, Optimal Experience in Ambient Intelligence, in Ambient Intelligence: The evolution of technology, communication and cognition towards the future of human-computer interaction, G. Riva, F. Vatalaro, F. Davide, and M. Alcañiz, Editors. 2004, IOS Press. On-line: http://www.emergingcommunication.com/volume6.html: Amsterdam. p. 35-43. [48] J.A. Waterworth, Virtual Realisation: Supporting creative outcomes in medicine and music. PsychNology Journal, (2003), 1(4): p. 410-427. http://www.psychnology.org/pnj1(4)_waterworth_abstract.htm. [49] F. Massimini and A. Delle Fave, Individual development in a bio-cultural perspective. American Psychologist, (2000), 55(1): p. 24-33. [50] A. Delle Fave, Il processo di trasformazione di Flow in un campione di soggetti medullolesi [The process of flow transformation in a sample of subjects with spinal cord injuries], in La selezione psicologica umana, F. Massimini, A. Delle Fave, and P. Inghilleri, Editors. 1996, Cooperativa Libraria IULM: Milan. p. 615-634. [51] N. Doidge, The Brain that Changes Itself: Stories of Personal Triumph from the frontiers of Brain Science. 2007, New York: Penguin Books. [52] S. Begley, The Plastic Mind. 2008, London: Constable & Robinson. [53] S.L. Wolf, C.J. Winstein, J.P. Miller, E. Taub, G. Uswatte, D. Morris, C. Giuliani, K.E. Light, and D. Nichols-Larsen, Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: the EXCITE randomized clinical trial. Jama, (2006), 296(17): p. 2095-104. [54] L.V. Gauthier, E. Taub, C. Perkins, M. Ortmann, V.W. Mark, and G. Uswatte, Remodeling the brain: plastic structural brain changes produced by different motor therapies after stroke. Stroke, (2008), 39(5): p. 1520-5.

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

21

[55] L. Collier and J. Truman, Exploring the multi-sensory environment as a leisure resource for people with complex neurological disabilities. NeuroRehabilitation, (2008), 23(4): p. 361-7. [56] S.B.N. Thompson and S. Martin, Making sense of multi-sensory rooms for people with learning disabilities. British Journal of Occupational Therapy, (1994), 57: p. 341-344. [57] K.W. Hope, The effects of multi-sensory environments on older people with dementia. Journal of Psychiatric and Mental Health Nursing, (1998), 5: p. 377-385. [58] K.W. Hope and H.A. Waterman, Using Multi-Sensory Environments (MSEs) with people with dementia. Dementia, (2004), 3(1): p. 45-68. [59] R. Baker, S. Bell, E. Baker, S. Gibson, J. Holloway, R. Pearce, Z. Dowling, P. Thomas, J. Assey, and L.A. Waering, A randomized controlled trial of the effects of multi-sensory stimulation (MSS) for people with dementia. British Journal of Clinical Psychology, (2001), 40(1): p. 81-96. [60] G.A. Hotz, A. Castelblanco, I.M. Lara, A.D. Weiss, R. Duncan, and J.W. Kuluz, Snoezelen: a controlled multi-sensory stimulation therapy for children recovering from severe brain injury. Brain Inj, (2006), 20(8): p. 879-88. [61] M. Lotan and J. Merrick, Rett syndrome management with Snoezelen or controlled multi-sensory stimulation. A review. Int J Adolesc Med Health, (2004), 16(1): p. 5-12. [62] M.M. Behrmann and L. Lahm, Babies and robots: technology to assist learning of young multiple disabled children. Rehabil Lit, (1984), 45(7-8): p. 194-201. [63] E. Libin and A. Libin, New diagnostic tool for robotic psychology and robotherapy studies. Cyberpsychol Behav, (2003), 6(4): p. 369-74. [64] F. Tanaka, A. Cicourel, and J.R. Movellan, Socialization between toddlers and robots at an early childhood education center. Proc Natl Acad Sci U S A, (2007), 104(46): p. 17954-8. [65] M.R. Banks, L.M. Willoughby, and W.A. Banks, Animal-assisted therapy and loneliness in nursing homes: use of robotic versus living dogs. J Am Med Dir Assoc, (2008), 9(3): p. 173-7. [66] R. Colombo, F. Pisano, A. Mazzone, C. Delconte, S. Micera, M.C. Carrozza, P. Dario, and G. Minuco, Design strategies to improve patient motivation during robot-aided rehabilitation. J Neuroeng Rehabil, (2007), 4: p. 3. [67] G. Riva and A. Gaggioli, Virtual clinical therapy. Lecture Notes in Computer Sciences, (2008), 4650: p. 90-107. [68] G. Castelnuovo, C. Lo Priore, D. Liccione, and G. Cioffi, Virtual Reality based tools for the rehabilitation of cognitive and executive functions: the V-STORE. PsychNology Journal, (2003), 1(3): p. 311-326. Online: http://www.psychnology.org/pnj1(3)_castelnuovo_lopriore_liccione_cioffi_abstract.htm. [69] J. Lessiter, J. Freeman, E. Keogh, and J. Davidoff, A Cross-Media Presence Questionnaire: The ITCSense of Presence Inventory. Presence: Teleoperators, and Virtual Environments, (2001), 10(3): p. 282297. [70] S. Miller and D. Reid, Doing play: competency, control, and expression. Cyberpsychol Behav, (2003), 6(6): p. 623-32. [71] D. Reid, The influence of virtual reality on playfulness in children with cerebral palsy: a pilot study. Occup Ther Int, (2004), 11(3): p. 131-44. [72] K. Harris and D. Reid, The influence of virtual reality play on children's motivation. Can J Occup Ther, (2005), 72(1): p. 21-9. [73] B.B. Johansson, Brain plasticity and stroke rehabilitation. The Willis lecture. Stroke, (2000), 31(1): p. 223-30. [74] G. Optale, A. Munari, A. Nasta, C. Pianon, J. Baldaro Verde, and G. Viggiano, Multimedia and virtual reality techniques in the treatment of male erectile disorders. International Journal of Impotence Research, (1997), 9(4): p. 197-203. [75] G. Optale, F. Chierichetti, A. Munari, A. Nasta, C. Pianon, G. Viggiano, and G. Ferlin, PET supports the hypothesized existence of a male sexual brain algorithm which may respond to treatment combining psychotherapy with virtual reality. Studies in Health Technology and Informatics, (1999), 62: p. 249251. [76] G. Optale, Male Sexual Dysfunctions and multimedia Immersion Therapy. CyberPsychology & Behavior, (2003), 6(3): p. 289-294. [77] G. Optale, F. Chierichetti, A. Munari, A. Nasta, C. Pianon, G. Viggiano, and G. Ferlin, Brain PET confirms the effectiveness of VR treatment of impotence. International Journal of Impotence Research, (1998), 10(Suppl 1): p. 45. [78] H.G. Hoffman, T.L. Richards, B. Coda, A.R. Bills, D. Blough, A.L. Richards, and S.R. Sharar, Modulation of thermal pain-related brain activity with virtual reality: evidence from fMRI. Neuroreport, (2004), 15(8): p. 1245-1248.

22

G. Riva and A. Gaggioli / Rehabilitation as Empowerment: The Role of Advanced Technologies

[79] H.G. Hoffman, T. Richards, B. Coda, A. Richards, and S.R. Sharar, The illusion of presence in immersive virtual reality during an fMRI brain scan. CyberPsychology & Behavior, (2003), 6(2): p. 127-131. [80] H.G. Hoffman, D.R. Patterson, J. Magula, G.J. Carrougher, K. Zeltzer, S. Dagadakis, and S.R. Sharar, Water-friendly virtual reality pain control during wound care. Journal of Clinical Psychology, (2004), 60(2): p. 189-195. [81] H.G. Hoffman, T.L. Richards, T. Van Oostrom, B.A. Coda, M.P. Jensen, D.K. Blough, and S.R. Sharar, The analgesic effects of opioids and immersive virtual reality distraction: evidence from subjective and functional brain imaging assessments. Anesth Analg, (2007), 105(6): p. 1776-83, table of contents. [82] H.G. Hoffman, D.R. Patterson, E. Seibel, M. Soltani, L. Jewett-Leahy, and S.R. Sharar, Virtual reality pain control during burn wound debridement in the hydrotank. Clin J Pain, (2008), 24(4): p. 299-304. [83] H.G. Hoffman, J.N. Doctor, D.R. Patterson, G.J. Carrougher, and T.A. Furness, 3rd, Virtual reality as an adjunctive pain control during burn wound care in adolescent patients. Pain, (2000), 85(1-2): p. 3059. [84] H.G. Hoffman, T.L. Richards, A.R. Bills, T. Van Oostrom, J. Magula, E.J. Seibel, and S.R. Sharar, Using FMRI to study the neural correlates of virtual reality analgesia. CNS Spectr, (2006), 11(1): p. 4551. [85] H.G. Hoffman, S.R. Sharar, B. Coda, J.J. Everett, M. Ciol, T. Richards, and D.R. Patterson, Manipulating presence influences the magnitude of virtual reality analgesia. Pain, (2004), 111(1-2): p. 162-8. [86] P.L. Weiss and N. Katz, The potential of virtual reality for rehabilitation. J Rehabil Res Dev, (2004), 41(5): p. vii-x. [87] D. Rand, R. Kizony, and P.T. Weiss, The Sony PlayStation II EyeToy: low-cost virtual reality for use in rehabilitation. J Neurol Phys Ther, (2008), 32(4): p. 155-63. [88] A. Rizzo, M.T. Schultheis, K. Kerns, and C. Mateer, Analysis of assets for virtual reality applications in neuropsychology. Neuropsychological Rehabilitation, (2004), 14(1-2): p. 207-239. [89] G. Riva, Virtual reality in psychotherapy: review. CyberPsychology & Behavior, (2005), 8(3): p. 22030; discussion 231-40. [90] G. Riva, A. Gaggioli, D. Villani, A. Preziosa, F. Morganti, R. Corsi, G. Faletti, and L. Vezzadini, NeuroVR: an open source virtual reality platform for clinical psychology and behavioral neurosciences. Studies in Health Technology and Informatics, (2007), 125: p. 394-9. [91] M.E.P. Seligman and M. Csikszentmihalyi, Positive psychology. American Psychologist, (2000), 55: p. 5-14. [92] C. Botella, C. Perpiña, R.M. Baños, and A. Garcia-Palacios, Virtual reality: a new clinical setting lab. Studies in Health Technology and Informatics, (1998), 58: p. 73-81. [93] F. Vincelli, From imagination to virtual reality: the future of clinical psychology. CyberPsychology & Behavior, (1999), 2(3): p. 241-248. [94] C. Botella, S. Quero, R.M. Banos, C. Perpina, A. Garcia Palacios, and G. Riva, Virtual reality and psychotherapy. Stud Health Technol Inform, (2004), 99: p. 37-54.

SECTION II TRAINING AND TECHNOLOGY AS AN AID IN FUNCTIONAL GAINS

Many people question why we don't just have subjects perform motor tasks in the real world. The answer to this question is that virtual reality offers us the opportunity to bring the complexity of the physical world into the controlled environment of the laboratory. Virtual reality gives us the potential to move away from reductionism in science and towards the measurement of natural movement within natural complex environments. Keshner, 2004

This page intentionally left blank

Advanced Technologies in Rehabilitation A. Gaggioli et al. (Eds.) IOS Press, 2009 © 2009 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-60750-018-6-25

25

Robotic Assistance for Upper Extremity Training after Stroke a

David J. REINKENSMEYERa Department of Mechanical and Aerospace Engineering University of California at Irvine, CA, USA

Abstract. There has been a rapid increase in the past decade in the number of robotic devices that are being developed to assist in movement rehabilitation of the upper extremity following stroke. Many of these devices have produced positive clinical results. Yet, it is still not well understood how these devices enhance movement recovery, and whether they have inherent therapeutic value that can be attributed to their robotic properties per se. This chapter reviews the history of robotic assistance for upper extremity training after stroke and the current state of the field. Future advances in the field will likely be driven by scientific studies focused on defining the behavioral factors that influence motor plasticity. Keywords. upper extremity, rehabilitation, robotics, motor control, plasticity

Introduction In the early 1990’s there were a handful of robotic devices being developed for upper extremity training after stroke. Today there are tens of prototypes and several companies selling commercial devices [1]. However, use of robotic devices in rehabilitation clinics is still rare. This chapter reviews the history of the field, and identifies factors that limit clinical acceptance and important directions for future scientific research. Section 1 reviews why engineers started investigating robots for use in rehabilitation therapy, and initial reactions by clinicians to these efforts. Section 2 reviews key design decisions that had to be made for the first robotic therapy devices, which in some ways defined the flow of the field. Section 3 reviews clinical results from the field and two important scientific questions that these results have raised. Section 4 discusses recent developments in robotic assistance for the upper extremity. The chapter concludes by suggesting directions for future research.

1. Robotic Assistance: Beginnings and Therapist Response 1.1. Precursors from Therapists The development of robotic devices for rehabilitation therapy can be seen as the logical progression of a stream of technological development activity begun by therapists

26

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

Figure 1. Pre-cursors of robotic therapy devices. The three devices on the left (Swedish sling, arm skateboard, and JAECO mobile arm support) are designed to provides assistance for arm movement without using actuators. The device on the right is the Biodex Active Dynamometer, which is a single degree-offreedom robot that can be adjusted to assist or resist movement around different joints.

themselves. Rehabilitation professionals have long taken an active interest in developing and using technology to assist in rehabilitation (Figure 1). Therapy catalogs such as the Sammons-Preston catalog (http://www.sammonspreston.com/) contain dozens of devices designed to assist in upper extremity therapy after stroke. Much of this technology tries to meet one or more of three goals: increasing activity, providing assistance, and assessing outcomes (Table 1). Implicit in the development of this technology was the idea of partial automation; that is, the technology might allow patients to practice some of the repetitive aspects of rehabilitation therapy on their own, without the continuous presence of the rehabilitation therapist. 1.2. Enter the Engineers In the late 1980’s and early 1990’s engineers began to realize that robotic devices could potentially be adapted to better fulfill these same goals [2, 3]. This work was a logical continuation of work on what were probably the first robotic devices for rehabilitation therapy: the active dynamometers, such as the Lido and Biodex machines, Table 1. Typical goals of older, simpler therapy technology, and how robotic devices further these goals. Goals of therapy technology Increase Activity: provide activities that allow stroke patients to independently exercise and practice functional tasks. Provide Assistance: assist patients in positioning or moving the hand or arm with a therapeutic goal.

Assess Outcomes: measure the movement performance of patients.

Example of simple, existing technology therabands, pegboards, blocks

Splints, arm supports

Grip force measurement devices, electrogoniometers, timers

How robotic devices further these goals Robots can simulate a variety of computerized activities and quickly and automatically switch between them. Robots can generate arbitrary patterns of assistance or resistance force against the patient’s limb, and automatically adjust this force based on performance. Robots can assess performance in an integrated and objective way using their sensors.

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

27

Figure 2. Some of the first robotic therapy devices for the arm to undergo clinical testing (left to right: MITMANUS [2], MIME [4], the ARM Guide [5]). These devices were designed to provide active assistance to patients during reaching movements with the arm.

developed in the late 1970’s and early 1980’s (Figure 1). Here we define a robot to be a device that can move in response to commands (cf. American Heritage Dictionary). Active dynamometers incorporate a computer-controlled motor, and thus fit this general definition of a robot. They include a kit of levers and bars that can be attached to the motor. The levers are designed to work with different limbs and joints (e.g. elbow flexion/extension, or should abduction/adduction), allowing patients to exercise a joint while the motor resists or assists movement. The dynamometer senses the torque and limb rotation that the patient generates, and displays this information to the patient and therapist for visual feedback and outcomes documentation. Robotics engineers realized that not only one-joint robotic devices with simple controllers like active dynamometers could be used in therapy, but also more sophisticated robotic mechanisms with more than one joint and more sophisticated controllers (Figure 2). Engineers began to delineate possible benefits of robots, in a way that aligned with many of the therapists’ technological goals defined above (Table 1). Engineers also explicitly promoted the goal of partial automation: robots had the potential to allow the patient to practice some of the repetitive aspects of rehabilitation therapy on their own, without the continuous presence of the rehabilitation therapist. 1.3. A Skeptical Reception by Some Clinicians, and a Collaborative Approach by Others Some clinicians expressed skepticism toward the idea that robots could help them meet rehabilitation goals. Skeptical clinicians had good reasons to be skeptical that included the following points: 1) Robots cannot match therapists’ expertise and skill. Therapy involves manual skills that are learned over the course of years by experience under the guidance of expert mentors. Some of these skills require sophisticated manual manipulations of complex joints (e.g. mobilizing the patient’s scapula). An alert and perceptive therapist alters her therapy goals and assistance based on a complex, ongoing consideration of the patient’s state and progress. In brief: hands-on therapy requires expertise and is complex; it seems doubtful that a robot could replicate hands-on therapy effectively. 2) Robots are unsafe: robots are dangerous because they can move patient’s limbs but are not intelligent and sensitive to contra-indications to imposed movement like human therapists. They could move a patient in a harmful way. 3) Robots might replace therapists. Also implicit in the dubious reception by some therapists was a concern that robots might replace therapists, just as

28

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

robots had replaced assembly workers in factories. Indeed, another definition of a robot is “a machine designed to replace human beings in performing a variety of tasks, either on command or by being programmed in advance.” (American Heritage Science Dictionary). Most engineers interested in robotic therapy probably never assumed that a robot could replace a therapist, because the job of a therapist is multifaceted and interpersonal, involving much more than just rotely moving limbs. Rather, the goal in the mind of most engineers was consistent with that of therapists’ own previous technological developments (Figure 1): to provide a means for patients to practice therapy on their own so that they could get more therapy at less cost (i.e. partial automation). Other clinicians were of course more receptive to the idea or robot-assisted therapy, perhaps because they saw robotic therapy devices as the logical evolution of technology already being used in therapy. Robotic devices were an opportunity to try to improve on the forms of technology already used in clinics to partially automate repetitive aspects of therapy. 1.4. Incentives for Forging Ahead Several research groups went ahead and developed robotic therapy devices for the arm, notably, MIT-MANUS [2], MIME [4], and the ARM Guide [5] (Figure 2), collaborating with the rehabilitation professionals who saw potential for these devices. These engineering teams were perhaps bolstered by the insights that robotics, control theory, and computational approaches were giving to the understanding of human motor control in the 1980’s (e.g. [6]). If engineering concepts and technology could help improve understanding of normal human motor control, could they also improve understanding of motor control after neurologic injury? The prospect of developing computational models of motor plasticity using robotic tools was intriguing. Another motivation in most research team’s minds was the possible business opportunity presented by robotic therapy: more people than ever before were in need of rehabilitation after stroke because of the demographics of aging in industrialized nations and the increased stroke survival rates, and this trend was expected to continue. At the same time, rehabilitation units were being forced to deliver less repetitive therapy because of cost-saving attempts in the health care industry. For example, the average length of stay for stroke survivors in inpatient rehabilitation facilities in the U.S. decreased from 31 days to 14 days after prospective payment system reimbursement was instituted in 1983 [7]. And yet rehabilitation science was finding with increasing certainty that recovery could be influenced by activity: training enhanced use-dependent plasticity (e.g. [8, 9]). Developers of robotic therapy devices thought that robots might help people with a stroke by allowing them access to a greater quantitative of repetitive therapy at less cost than would be possible with oneon-one interactions with a clinician. This access might allow the creation of new businesses, providing an additional incentive to pursue device development.

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

29

2. Initial Design Decisions 2.1. But what should the robot do? To this point, I have spoken of “robot assistance” in general terms – the robot assists the therapist and patient in some way that promotes rehabilitation. When it came time to actually build robotic therapy devices, however, engineers had to determine exactly what the robots were to do – for example, they had to write the computer program that controlled the motors on the robot. Here, engineers encountered a problem: we discovered that the specific movement and assistance patterns that were effective for therapy were relatively unknown. Despite a history of over one hundred years, and the presence of somewhat dogmatic schools of therapy (e.g. Neurodevelopmental Treatment, Brunstrom Technique, Proprioceptive Neural Facilitation [10]), the field of rehabilitation science had at that time few randomized controlled trials that defined the elements of therapy that specifically aided recovery [11]. Clinical practice varied widely, with details of therapeutic techniques sometimes in opposition to each other in different clinics (e.g. should the therapist promote movement within synergy or avoid it? Is movement against resistance therapeutic, or does it increase spasticity?), depending on which school of therapy the clinic’s therapists had been educated in. The general lack of evidence for specific motions to be practiced or assistance patterns to be applied had the practical result that there was not a well-defined scientific basis on which the design of robots and computer algorithms for movement training could be based. 2.2. A Logical Target: Active Assist Exercise Despite this uncertainty, or perhaps because of it, the therapeutic target that the robotic therapy research teams chose for MIT-MANUS, MIME, and the ARM Guide was the same: active assist exercise, and, indeed, this technique has continued to be the primary target for robotic therapy devices. In this technique, the therapist manually assists the patient in achieving desired movements. The “active” refers to the patient being active and engaged; i.e. the patient tries to move during the exercise. The “assist” refers to the therapist manually assisting the patient, but only as much as needed. Researchers chose this technique as a target because most of the schools of therapy seemed to incorporate active assist exercise as an element [10]. As a result, application of this technique could be witnessed on almost any day on a visit to almost any rehabilitation clinic. The technique was also amenable to robotic implementation – assisting movement was something robots could do. It was also straightforward to conceive of a scientific rationale for active assist therapy, although the rationale was speculative rather than verified: 1) Suppleness Enhancement: at the lowest level of motor control of biomechanics and reflexes, active assist exercise stretches soft tissue and muscles, which might be helpful for preventing contracture and reducing spasticity. 2) Plasticity Enhancement: at a middle level of motor control, active assist exercise provides the patient’s motor system with somatosensory stimulation that would normally not be available because the patient is paretic. Somatosensory input had recently been shown to drive cortical plasticity [12].

30

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

3) Motivation Enhancement: at a high level of the motor system, active assist exercise may motivate patients to exercise. If a patient cannot move well on his own, he are she may be disinclined to try to move. Active assist exercise allows the patient to be successful in achieving a desired movement, presumably motivating practice and effort [13]. However, it should be noted, assisting too much with a robot may decrease effort [14]. As stated earlier, the field of rehabilitation science was not well established and none of these rationales was scientifically proven at the time. They still remain largely unproven today, even though most robotic therapy devices still focus on implementing active assist exercise. 2.3. But what joints? A decision also had to be made about which joints of the upper extremity to focus on, as development of a robotic exoskeleton that can assist in all joint movements of the upper extremity was and remains an unsolved problem, especially for the hand and shoulder complex. The first robotic therapy devices for the upper extremity that were clinically tested (i.e. MIT-MANUS, MIME, and the ARM Guide, Figure 2) focused on providing active assist exercise for elbow flexion/extension and for limited shoulder movements (e.g. shoulder flexion below 90 degrees and limited external rotation). Three reasons for this choice were: 1) Simplicity: these joints were viewed as simpler than the hand, wrist, and complex shoulder movements. 2) Availability of tools: robots had already been developed to study motor control at these joints, and thus there were technological precedents and scientific concepts from which to build. For example, MIT-MANUS was essentially the same robot that was concurrently being used in early, influential studies of motor adaptation [15]. MIME used an industrial robot that had the scale of human arm movements. 3) Pragmatism: the hand often appears to be hopelessly impaired following stroke, and shoulder problems such as subluxation are governed by complex biomechanical and neurological mechanisms which would be very difficult for a robot to address. Reaching movements with the arm are needed for a lot of functional activities. Robotic therapy research teams therefore aimed to achieve functional improvements by making robots that focused on reaching movements with the arm. It is worth noting that it is still unclear which joints to focus on for an optimal therapeutic result because of a lack of clinical trials addressing this question. Intriguingly, a device focused on simple wrist and forearm movements, the BiManuTrac, has produced the largest changes in impairment observed with robotic therapy to date [16]. 2.4. And what types of movements? Finally a related decision had to made about what types of movements the patient would perform with robot assistance. Should the movements be single-joint or multiple joint? Should they be fast-as-possible or slow? Should they avoid abnormal synergy

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

31

patterns or work to build strength in those patterns? Bimanual, with two robots, or unimanual? Should they have a functional goal? The motions used by MIT-MANUS in the first clinical trials were unimanual pointing movements in the horizontal plane [17]. The patient was instructed to move a cursor to a target. After attaining the target, the target moved to a new location. The robot helped the patient to make the movement to the target, following a normative trajectory (minimum jerk trajectory) [17]. This type of paradigm had been used often previously in motor control research. It required multiple-joint coordination, and was functional in a sense, since pointing (or reaching) is a component of many activities of daily living. MIME and the ARM Guide also focused on unimanual reaching movements. MIME incorporated some bimanual reaching exercises also.

3. Initial Clinical Tests and the Questions they Raised 3.1. First Clinical Results The basic findings of the initial clinical tests with the first three robotic therapy devices for the arm (MIT-Manus, MIME, and the ARM Guide) were as follows (for detailed reviews, see: [1, 9, 18]): 1.

2.

3.

Statistically Significant Motor Gains: An additional dose of active assist exercise, delivered with a robotic device with an intensity of several hours per week for several weeks, significantly (in a statistical sense) improved motor recovery in the acute or chronic stage following a stroke, as measured with quantitative measures of range of motion or strength, or clinical impairment scales (Figure 3). Patients typically maintained this improvement at long-term follow-up (i.e. months later). Modest Motor Gains: While statistically significant, the gains due to robotic therapy were small – typically 2-6 points on the upper extremity Fugl-Meyer scale [19], which ranges from 0-66 (Figure 3). Functional gains, as measured with clinical ADL scales, typically were even smaller and sometimes not significant [19]. Comparable Motor Gains: The gains due to robotic therapy were roughly the same size as those due to a matched amount of conventional rehabilitation therapy, or to unassisted rehabilitation practice, as well as comparable between the different robots used (Figure 3). In other words, comparisons between different types of therapy often led to statistically inconclusive results.

Clinical testing of second generation robotic therapy devices has essentially been confirmatory of these findings, as reviewed in a recent systematic review [19].

32

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

Figure 3. Change in Fugl-Meyer Upper-Extremity Score with one to two months of training several hours per week after chronic stroke, for three robotic devices (MIT-MANUS [17], MIME [4], and Gentle-S [20]), and with conventional table-top exercise [21] and with the TWREX non-robotic exoskeleton [21] (see Figure 4). The Fugl-Meyer score varies from 0 (complete paralysis) to 66 (normal movement ability).

3.2. Questions Raised by Initial Clinical Testing This initial clinical testing raised two important questions: 1.

2.

The Question of Necessity: Was the robot necessary for the observed therapeutic benefit? I think the clearest way to express this question is as follows [22]: Consider a control group for which the motors of the robot are removed but the joints are allowed to move freely such that the robot allows movement but does not assist movement. The unactuated robot provides the same audiovisual stimulation, and the control group undergoes a matched duration of unactuated therapy. Would this control group recover less than a group that exercised with the actuated robot? If not, this would suggest that the robotic properties themselves (i.e. the programmable actuators) were superfluous. This result is scientifically plausible because, with regards to motor plasticity after stroke, we know that practice is a key (or perhaps the key) stimulant for motor plasticity. The Question of Optimization. If one accepts that the robotic properties of robotic therapy are helpful for enhancing recovery, a logical question is how sensitive are therapeutic benefits to the optimization of the robotic parameters? The first robotic therapy devices elicited therapeutic benefits comparable to each other, even though they were fairly different in their design and approach (e.g. number of degrees of freedom, details of the form of assistance provided, stiffness levels). Can tuning the robot geometry and control algorithm increase the therapeutic benefits? Or will any reasonable robot (or non-robotic therapy) give approximately the same result?

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

33

4. State of the Field Today 4.1. Progress in answering questions about the necessity and optimization of robotic actuation Few randomized controlled trials have yet addressed whether robotic actuation is necessary for therapeutic benefit and how much it can be optimized. A recent exception was a study that found that chronic stroke patients who received a fixed dose of active assist therapy for the hand from a robotic device (HWARD) recovered significantly better than a group that received half as much active assist therapy [23]. The number of patients included in this study was small (n = 13) and the baseline characteristics of the subjects were slightly mismatched, however, so the result needs to be examined with a larger study. The additional advantage due to more active assist therapy was moderate (about 3 extra Fugl-Meyer points). Notably, the process of answering the necessity and optimization questions is theoretically endless because of the problem of “unlimited alternatives”. That is, even if a randomized controlled trial demonstrates that the robotic properties being tested were unnecessary to generate the observed benefits (i.e. a group trained with an unactuated technique at similar dosage receives similar therapeutic benefits), or even if an interesting tweak of a robot’s parameters does not substantially alter the clinical outcomes, such a negative finding would of course only be for one particular instantiation of robot therapy. Other robots or different control algorithms, some maybe as yet unconceived, may produce better results. Since there are an infinite number of possible robots and robot control algorithms, it may be impossible to provide definitive answers to these questions. In addition, establishing negative results (i.e. no difference between therapy groups) with a high level of precision requires large subject populations because of the high inter-subject variability in stroke patients and the nature of statistical power, again adding effort, cost, and time to the process. 4.2. Trends in the Field If the field has not focused on answering the necessity and optimization questions with clinical trials, what has it focused on? Three trends mark the field of robotic therapy for the upper extremity today: 1.

2.

Rapid Proliferation of Innovative Hardware. Many cleverly designed robotic devices have been or are being developed to assist at different joints, at more joints, or at the same joints as before with improved weight, mass, or control properties (Figure 4, see review: [1]). Non-robotic approaches are also being developed, such as devices that passively relieve the weight of the arm [21, 24]. Initial testing suggests that passive devices may have similar clinical benefits with lower cost and theoretically-better safety [21] (Figure 4). Several companies are now selling upper extremity devices, and sales of these devices number in the hundreds. Development of New Control Strategies. Most current research on control strategies still focuses on active assist exercise. To improve active assistance algorithms, researchers are exploring several strategies, including:

34

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

Figure 4. Recently developed robotic and non-robotic therapy devices. Upper left: NeReBot: a 5 DOF cable robot that can be used next to a patient’s bed [27]. Bottom left: ArmIn: a highly responsive robot that allows naturalistic arm movement, including shoulder translation [28]. Middle top: Rupert: a lightweight exoskeleton actuated with pneumatic muscles, which can be worn by the subject [29]. Middle bottom: TWREX – a non-robotic arm support device [21]. Upper right: HWARD – a 3 DOF hand and wrist robot [23]. Lower right: A cable driven glove that can be worn, and driven by a motor or the patients shoulder shrugs [30].

•

•

•

Improved Compliance and Feedforward Control: These efforts include methods to make robots more compliant but still able to assist in spatial movement, by incorporating feedforward control [25, 26]. Compliance may have the advantage of making the patient feel more in control of therapy, and thus more engaged. It also preserves the relationship between motor commands that the patient generates and actual movement direction, which may allow patients to better optimize motor commands, since they receive accurate information about the results of a change in their motor command, whereas a stiff robot will always enforce the same trajectory. Adaptive Control: Several groups are making the controller adaptive, so that the robot changes its assistance based on ongoing sensing of patient performance [25, 31, 32]. The key concept here is that patient ability changes during therapy, and it would theoretically be best to keep the patient appropriately challenged, for provoking motor learning. Optimization: Optimization theory allows the goals of the therapy to be expressed as a high-level control objective. For example, for active assistance, my research group has proposed to minimize a weighted sum of patient movement error and robot assistance force [33]. Minimization of this cost function thus helps the patient achieve a desired trajectory, but with as little robot force as possible (Assistance-as-needed). Optimization theory provides a means to derive the robot therapy controller that mathematically optimizes the cost function. Within an optimization framework, robotic therapy controllers can be rigorously proven to satisfy a “high level” goal, rather than being based on ad hoc strategies devised by the research team.

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

•

35

Neuro-Computational Modeling: My own research group has also begun to develop computational models that model what the patient’s brain is computing during therapy to gain insight into how better to design robotic therapy controllers [34, 35]. The concept here is that if we can mathematically model how behavioral signals drive adaptation, then we should be able to design control strategies that mathematically optimize adaptation.

Other therapeutic paradigms besides active assistance are also being explored including: •

•

Error amplification strategies [36, 37]: The concept behind this approach is that movement errors drive motor adaptation, and thus assistance may be the wrong approach to take if the goal is to enhance motor adaptation, since assistance reduces movement errors. Amplifying errors may improve the rate or extent of motor adaptation by better provoking motor plasticity. Clinically, this technique has only been shown to be effective in reducing curvature errors during supported-arm reaching in the shortterm [38]. Virtual environments (see review: [39]) Another alternate therapeutic paradigm that differs from the active assistance paradigm that dominates the field is to use the robot to create a virtual environment that simulates different therapeutic activities. In this paradigm, the robot may not physically assist or resist movement, but instead just provide a training environment that simulates reality. Potential advantages of training in a haptic environment over training in physical reality include: a haptic simulator can create many different interactive environments simulating a wide range of real-life situations; quickly switch between these environments without a “set-up” time, automatically grade the difficulty of the training environment by adding or removing virtual features; make the environments more interesting than a typical rehabilitation environment; automatically “reset” itself if virtual objects are dropped or misplaced; and provide novel forms of visual and haptic feedback regarding performance. In this haptic simulation framework, robotics may benefit rehabilitation therapy not by provoking motor plasticity with special assisting or resisting control schemes, but rather by providing a diverse, salient, and convenient environment for semi-autonomous training.

3. Rehabilitation Therapists are Accepting Robots as Scientific but not Clinical Tools. A third trend is that while rehabilitation therapists are not widely incorporating commercial robotic therapy devices for clinical use, they are using robots in their research. The research therapists in the conference that led to this book are setting the pace: they are doing groundbreaking scientific work using robotics and related technology, as can be read in this book’s other chapters (see chapter by Mataric, for example).

36

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

5. Conclusion As mentioned in the Introduction, in the early 1990’s there were only a handful of robotic devices being developed for upper extremity training after stroke. In 2008, there are dozens of devices being developed. However, robotic therapy has not become a standard therapeutic treatment in most clinics. What impedes clinical acceptance? One important factor is that the therapeutic benefits of robotic therapy are modest, and have not been shown to be decisively better than other, less expensive approaches that can partially automate therapy (Figure 1). In other words, the necessity question remains unanswered. There is little motivation for most clinics to buy expensive robots until it is proven that the robots yield therapeutic or cost benefits that are substantially better than current approaches. The field seems to be investing the majority of its resources in developing new devices, rather than in understanding and optimizing the content of robotic therapy. One explanation for this phenomenon is that there is a lack of devices for certain movements and applications, such as hand movement and naturalistic arm movement, and the new technology addresses this lack, as well as improving features such as portability and force control response (Figure 4). But another possible factor is that engineers like to build devices and are good at it. Engineers’ motivation and expertise for scientifically exploring the clinical effects of their devices is more limited, and this may signal the need for an even greater role by clinician scientists. The field will likely have to evolve to place more focus on scientific studies of the mechanisms of motor plasticity to optimize technology, improve the benefits of robotic therapy, and determine if routine clinical use makes sense. The question of “What are the maximum benefits that we can obtain with robotic therapy?” can be illustrated by a boy playing with a stomp rocket (Figure 5). A dose of robotic therapy is like stomping on the air bladder. The altitude that the rocket reaches is like the resulting improvement in motor control. The boy can increase the rocket altitude by stomping harder, just like a robotic device can increase recovery if it uses an optimal training paradigm, but there is a limit to how the rocket, and likely recovery also, can go. For upper extremity recovery, the limit is probably dictated by the number of spared corticospinal neurons following stroke. The limit for the rocket is well short of the Eiffel tower, despite the perspective shown in Figure 5. Does a trick of perspective make us think that the limits for recovery enhancement that are possible with robotic therapy are higher than they really are, if indeed the amount of cell loss defines them? Addressing the following two key questions would help answer this question, and advance robotic therapy development: 1.

What behavioral signals provoke plasticity during rehabilitation? Knowing these signals would allow us to design robots that optimally influence those signals. This would provide answers to questions like “What type of forces (error attenuating or error amplifying)?, “What joints?”, “What movements?”, and “What type of feedback?”.

2.

What are the fundamental limits to the plasticity that can be provoked with behavioral signals? Answering this question would define the limits we should expect of robotic therapy optimization. It would thus allow us to determine how much time to invest in optimizing robotic therapy itself. If the cost function is relatively flat and we are already close to an optimum, it may

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

37

Figure 5. What are the maximum benefits that we can obtain with robotic therapy?

make sense to focus more attention on approaches that combine cell- or molecule based regeneration techniques with robotic therapy, in search for a synergy that improves clinical results beyond that achievable with either robots or regeneration alone.

Acknowledgements The contents of this chapter were developed in part with support from NIDRR H133E070013 and NIH N01-HD-3-3352.

References [1] [2] [3]

[4]

[5]

[6] [7]

[8]

B.R. Brewer, S.K. McDowell, and L.C. Worthen-Chaudhari, Poststroke upper extremity rehabilitation: a review of robotic systems and clinical results, Top. Stroke Rehabilitation 14 (2007), 22-44. H.I. Krebs, N. Hogan, M.L. Aisen, and B.T. Volpe, Robot-aided neurorehabilitation, IEEE Transactions on Neural Systems and Rehabilitation Engineering 6 (1998), 75-87. P.S. Lum, D.J. Reinkensmeyer, and S.L. Lehman, Robotic assist devices for bimanual physical therapy: preliminary experiments, IEEE Transactions on Neural Systems and Rehabilitation Engineering 1 (1993),185-191. P.S. Lum, C.G. Burgar, S.P.C.M. Majmundar, and M. Van der Loos, Robot-assisted movement training compared with conventional therapy techniques for the rehabilitation of upper limb motor function following stroke, Archives of Physical Medicine and Rehabilitation 83 (2002), 952-9. D. Reinkensmeyer, L. Kahn, M. Averbach, A. McKenna, B. Schmit, and W. Rymer, Understanding and promoting arm movement recovery after chronic brain injury: Progress with the ARM Guide, Journal of Rehabilitation Research & Development submitted (2000). C.G. Atkeson, Learning arm kinematics and dynamics, Annual Review of Neuroscience 12 (1989), 157183. S.M. Schmidt, L. Guo, and S.J. Scheer, Changes in the status of hospitalized stroke patients since inception of the prospective payment system in 1983, Archives of Physical Medicine and Rehabilitation 83 (2002), 894-898. R.J. Nudo, B.M. Wise, F. SiFuentes, and G.W. Milliken, Neural substrates for the effects of rehabilitative training on motor recovery after ischemic infarct, Science 272 (1996), 1791-1794.

38

[9]

[10] [11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19] [20]

[21]

[22]

[23] [24]

[25]

[26] [27] [28] [29]

[30]

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

G. Kwakkel, R. van Peppen, R.C. Wagenaar, S. Wood-Dauphinee, C. Richards, A. Ashburn, K. Miller, N. Lincoln, C. Partridge, I. Wellwood, and P. Langhorne, Effects of augmented exercise therapy time after stroke: A meta-analysis, Stroke 35 (2004), 2529-2539. C.A. Trombly, Occupational Therapy for Dysfunction, 4th Edition, Baltimore: Williams and Wilkins, 1995. G. Gresham, P. Duncan, W. Stason, H. Adams, A. Adelman, D. Alexander, D. Bishop, L. Diller, N. Donaldson, C. Granger, A. Holland, M. Kelly-Hayes, F. McDowell, L. Myers, M. Phipps, E. Roth, H. Siebens, G. Tarvin, and C. Trombly, Post-Stroke Rehabilitation. Rockville, MD: U.S. Department of Health and Human Services. Public Health Service, Agency for Health Care Policy and Research, 1995. M.M. Merzenich, and W.M. Jenkins, Reorganization of cortical representations of the hand following alterations of skin inputs induced by nerve injury, skin island transfers, and experience, Journal of Hand Therapy 6 (1993), 89-104. D.J. Reinkensmeyer, and S.J. Housman, If I can't do it once, why do it a hundred times?: Connecting volition to movement success in a virtual environment motivates people to exercise the arm after stroke, Proc. Virtual Rehabilitation Conference (2007), 44-48. J.F. Israel, D.D. Campbell, J.H. Kahn, and T.G. Hornby, Metabolic costs and muscle activity patterns during robotic- and therapist-assisted treadmill walking in individuals with incomplete spinal cord injury, Physical Therapy 86 (2006),1466-78. R. Shadmehr, and F.A. Mussa-Ivaldi, Adaptive representation of dynamics during learning of a motor task, Journal of Neuroscience 14 (1994), 3208-3224. S. Hesse, C. Werner, M. Pohl, S. Rueckriem, J. Mehrholz, and M.L. Lingnau, Computerized arm training improves the motor control of the severely affected arm after stroke: a single-blinded randomized trial in two centers, Stroke 36 (2005),.1960-6. S. Fasoli, H. Krebs, J. Stein, W. Frontera, and N. Hogan, Effects of robotic therapy on motor impairment and recovery in chronic stroke, Archives of Physical Medicine and Rehabilitation 84 (2003), 477-82. L.E. Kahn, M.L. Zygman, W.Z. Rymer, and D.J. Reinkensmeyer, Robot-assisted reaching exercise promotes arm movement recovery in chronic hemiparetic stroke: A randomized controlled pilot study, Journal of Neuroengineering and Neurorehabilitation 3 (12) (2006). G. Kwakkel, B.J. Kollen, and H.I. Krebs, Effects of robot-assisted therapy on upper limb recovery after stroke: a systematic review, Neural Repair 22 (2008), 111-121. F. Amirabdollahian, R. Loureiro, E. Gradwell, C. Collin, W. Harwin, and G. Johnson, Multivariate analysis of the Fugl-Meyer outcome measures assessing the effectiveness of GENTLE/S robotmediated stroke therapy, Journal of Neuroengineering Rehabilitation 19 (2007), 4. S.J. Housman, V. Le, T. Rahman, R.J. Sanchez, and D.J. Reinkensmeyer, Arm-Training with T-WREX after Chronic Stroke: Preliminary Results of a Randomized Controlled Trial, To Appear, 2007 IEEE International Conference on Rehabilitation Robotics, 2007. L. Kahn, P. Lum, W. Rymer, and D. Reinkensmeyer, Robot-assisted movement training for the strokeimpaired arm: Does it matter what the robot does?, Journal of Rehabilitation Research and Development 43 (2006), 619-630. C.D. Takahashi, L. Der-Yeghiaian, V. Le, R.R. Motiwala, and S.C. Cramer, Robot-based hand motor therapy after stroke, Brain 131 (2008), 425-437. A.H.A. Stienen, E.E.G. Hekman, F.C.T. Van der Helm, G.B. Prange, M.J.A. Jannink, A.M.M. Aalsma, and H. Van der Kooij, Freebal: dedicated gravity compensation for the upper extremities, IEEE 10th International Conference on Rehabilitation Robotics (2007), 804-808. E. T. Wolbrecht, D.J. Reinkensmeyer, and J.E. Bobrow, Optimizing compliant, model-based robotic assistance to promote neurorehabilitation, IEEE Transactions Neural Systems and Rehabiltation Engineering 16 (2008), 286-297. M. Mihelj, T. Nef, and R. Riener, A novel paradigm for patient-cooperative control of upper-limb rehabilitation robots, Advanced Robotics 21 (2007), 843-867. S. Masiero, A. Celia, G. Rosati, and M. Armani, Robotic-assisted rehabilitation of the upper limb after acute stroke, Archives of Physical Medicine and Rehabilitation 88 (2007), 142-9. T. Nef, and R. Riener, ARMin – Design of a Novel Arm Rehabilitation Robot, Proceedings of the 2005 IEEE International Conference on Rehabilitation Robotics, Chicago, Illinois, 2005, pp. 57-60. T.G. Sugar, J. He, E.J. Koeneman, J.B. Koeneman, R. Herman, H. Huang, R.S. Schultz, D.E. Herring, J. Wanberg, S. Balasubramanian, P. Swenson, and J.A. Ward, Design and control of RUPERT: a device for robotic upper extremity repetitive therapy, IEEE Transactions Neural Systems and Rehabilitation Engineering 15 (2007), 336-346. H.C. Fischer, K. Stubblefield, T. Kline, X. Luo, R.V. Kenyon, and D.G. Kamper, Hand rehabilitation following stroke: a pilot study of assisted finger extension training in a virtual environment. Top Stroke Rehabilitation 14 (1)(2007), 1-12.

D.J. Reinkensmeyer / Robotic Assistance for Upper Extremity Training After Stroke

39

[31] H. Krebs, J. Palazzolo, L. Dipietro, M. Ferraro, J. Krol, K. Rannekleiv, B. Volpe, and N. Hogan, Rehabilitation robotics: performance-based progressive robot-assisted therapy, Auto. Rob. 15 (2003), 720. [32] R. Riener, L. Lunenburger, S. Jezernik, M. Anderschitz, G. Colombo, and V. Dietz, Patient-cooperative strategies for robot-aided treadmill training: first experimental results, IEEE Transactions Neural Systems and Rehabilitation Engineering 13 (2005), 380-394. [33] J.L. Emken, R. Benitez, and D.J. Reinkensmeyer, Human-robot cooperative movement training: learning a novel sensory motor transformation during walking with robotic assistance-as-needed, Journal of Neuroengineering Rehabilitation 4 (2007), 8. [34] J.L. Emken, R. Benitez, A. Sideris, J.E. Bobrow, and D.J. Reinkensmeyer, Motor adaptation as a greedy optimization of error and effort, Journal of Neurophysiology 97 (2007), 3997-4006. [35] D.J. Reinkensmeyer, E. Wolbrecht, and J. Bobrow, A computational model of human-robot load sharing during robot-assisted arm movement training after stroke, IEEE Engineering in Medicine and Biology Society 2007 (2007), 4019-4023. [36] J.L. Patton, M.E. Phillips-Stoykov, M. Stojakovich, and F.A. Mussa-Ivaldi, Evaluation of robotic training forces that either enhance or reduce error in chronic hemiparetic stroke survivors, Experimental Brain Research 168 (2005), 368-383. [37] J.L. Emken, and D.J. Reinkensmeyer, Robot-enhanced motor learning: accelerating internal model formation during locomotion by transient dynamic amplification, IEEE Transactions Neural Systems and Rehabilitation Engineering 13 (2005), 33-9. [38] J. Patton, M. Kovic, and F. Mussa-Ivaldi, Custom-designed haptic training for restoring reaching ability to individuals with poststroke hemiparesis. Journal of Rehabilitation Research and Development 43 (2006), 643-56. [39] J.L. Patton, G. Dawe, C. Scharver, F.A. Muss-Ivaldi, and R. Kenyon, Robotics and virtual reality: A perfect marriage for motor control research and rehabilitation, Assistive Technology 18 (2006), 181195.

40

Advanced Technologies in Rehabilitation A. Gaggioli et al. (Eds.) IOS Press, 2009 © 2009 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-60750-018-6-40

Robotic assisted rehabilitation in Virtual Reality with the L-EXOS Antonio FRISOLIa, Massimo BERGAMASCOa, Maria C. CARBONCINIb and Bruno ROSSIb a PERCRO Laboratory, Scuola Superiore Sant’Anna, Pisa, Italy b Neurorehabilitation Unit, Department of Neurosciences, University of Pisa, Italy Abstract. This study presents the evaluation results of a clinical trial of robotic-assisted rehabilitation in Virtual Reality performed with the PERCRO L-Exos (Light-Exoskeleton) system, which is a 5-DoF force-feedback exoskeleton for the right arm. The device has demonstrated itself suitable for robotic arm rehabilitation therapy when integrated with a Virtual Reality (VR) system. Three different schemes of therapy in VR were tested in the clinical evaluation trial, which was conducted on a group of nine chronic stroke patients at the Santa Chiara Hospital in Pisa-Italy. The results of this clinical trial, both in terms of patients performance improvements in the proposed exercises and in terms of improvements in the standard clinical scales which were used to monitor patients receovery are reported and discussed. The evaluation both pre and post-therapy was carried out with both clinical and quantitative kinesiologic measurements. Statistically significant improvements were found in terms of Fugl-Meyer scores, Ashworth scale, increments of active and passive ranges of motion of the impaired limb, and quantitative indexes, such as task time and error. Keywords. Exoskeleton, robotic-assisted rehabilitation, task-oriented movement, reaching target, clinical protocol, Virtual Reality, Range of Motion, Fugl-Meyer assessment

Introduction Several studies demonstrate the importance of an early, constant and intensive rehabilitation following cerebral accidents. This kind of therapy is an expensive procedure in terms of human resources and time, and the increase of both life expectance of world population and incidence of stroke is making the administration of such therapies more and more important. The impairment of upper limb function is one of the most common and challenging consequences following stroke, that limits the patient’s autonomy in daily living and may lead to permanent disability [1]. Well-established traditional stroke rehabilitation techniques rely on thorough and constant exercise [2, 3], which patients are required to carry out within the hospital with the help of therapists, as well as during daily life at home. Early initiation of active movements by means of repetitive training has proved its efficacy in guaranteeing a good level of motor capability recovery [4]. Such techniques allow stroke patients to partially or fully recover motor functionalities during the acute stroke phase, due to the clinical evidence of a period of rapid sensorimotor recovery in the first three months after

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

41

stroke, after which improvement occurs more gradually for a period of up to two years and perhaps longer [5, 6]. However after usual therapies, permanent disabilities are likely to be present in the chronic phase, and in particular a satisfying upper extremity motor recovery is much more difficult to obtain with respect to lower extremities [7]. Several studies have attempted to investigate the efficacy of stroke rehabilitation approaches [8, 9]. Intensive and task oriented therapy for the upper limb, consisting of active, highly repetitive movements, is one of the most effective approaches to arm function restoration [10, 11]. The driving motivations to apply robotic technology to stroke rehabilitation are that it may overcome some of the major limitations that manual assisted movement training suffers from, i.e. lack of repeatability, lack of objective estimation of rehabilitation progress, and high dependence on specialized personnel availability. Robotic devices for rehabilitation can help to reduce the costs associated with the therapy and lead to new effective therapeutic procedures. In addition, Virtual Reality can provide a unique medium where therapy can be provided within a functional and highly motivating context, that can be readily graded and documented. The cortical reorganization and associated functional motor recovery after Virtual Reality treatments in patient with chronic stroke are documented also by fRMN [12]. Among leg rehabilitation robot devices, Lokomat [13] has become a commercial and widely diffused lower limb robotic rehabilitation device. It is a motorized orthosis able to guide knee and ankle movements while the patient walks on a treadmill. Concerning arm rehabilitation devices, both cartesian and exoskeleton-based devices have been developed in the last 10 years. MIT Manus [14, 15] and its commercial version InMotion2 [16] are pantograph-based planar manipulators, which have extensively been used to train patients on reaching exercises and have been constantly evaluated by means of clinical data analysis [17]. It has been designed to be backdrivable as much as possible and to have a nearly isotropic inertia. ARM-guide [18, 19] is a device which is attached to the patient’s forearm and guides the arm along a linear path having a variable angle with respect to the horizontal position. Constraint forces and range of motion are measured throughout the exercises. The MIME (Mirror Image Movement Enabler) system [20] is a bimanual robotic device which uses an industrial PUMA 560 robot that applies forces to the paretic limb during 3-dimensional movements. The system is able to replicate the movements of the non-paretic limb. Exoskeletons are robotic systems designed to work linked with parts of the human body and, unlike robots, are not designed to perform specific tasks autonomously in their workspace [21]. In such a condition, the issue of the physical interaction between robots and humans is considered in terms of safety. The design of exoskeleton systems stems from opposite motivations that intend the robotic structure to be always maintained in contact with the human operators limb. Such a condition is required for several applications that include the use of master robotic arms for teleoperation, active orthoses and rehabilitation [22]. Experiments on exoskeletons have been performed at the JPL during 1970s [23]. Sarcos [24] developed a master arm used for the remote control of a robotic arm, while at PERCRO arm exoskeletons have been developed for interaction with virtual environments since 1994 [22, 25, 26]. Exoskeletons can be suitably employed in robotic assisted rehabilitation [27].

42

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

Two exoskeleton-based systems have been developed at Saga University, Japan. The older one [28] is a 1-DoF interface for the human elbow motion, where angular position and impedance of the robot are tuned relying on biological signals used to interpret the human subjects intention. The newer neuro-fuzzy controlled device [29] is a 2-DoF interface used to assist human shoulder joint movement. Another device, the ARMin, has been developed at ETH, Switzerland [30, 31]. This device provides three active DoFs for shoulder and one active DoF for elbow actuation. The patient is required to perform task-oriented repetitive movements having continuous visual, auditory and haptic feedback. The Salford Exoskeleton [32], which is based on pneumatic Muscle Actuators (pMA) and provides an excellent power over weight ratio, has also been used in physiotherapy and training. A recent survey [33] on the efficacy of different robot assisted therapies outlines that robotic-aided therapy allows a higher level of improvement of motor control if compared to conventional therapy. Nevertheless, it is to be noted that no consistent influence on functional abilities has yet been found. This chapter presents the results of an extended clinical trial employing the LExos system [34], a 5-DoF force-feedback exoskeleton for the right arm; the system was installed at the Neurorehabilitation Unit of the University of Pisa, where it was used for the robotic assisted VR-based rehabilitation in a group of 9 chronic stroke patients[35, 36]. This work is intended to extend previous works concerning a pilot study with the L-Exos system by providing significant therapy and clinical data from a much larger set of patients. Section 1 presents a general description of the L-Exos system, underlining the main features which make the device useful for rehabilitation purposes, and a description of the developed VR applications may be found in Section 2. Section 3 and Section 4 discuss the main results which have been obtained with the L-Exos both in terms of improvements in the metrics used to assess patient performance in the therapy exercises and in terms of improvements in the standard clinical scales which have been used to monitor patients’ recovery. Conclusions and perspectives opened by this pilot study are briefly reported in Section 5.

1. The L-EXOS system L-Exos (Light Exoskeleton) is a force feedback exoskeleton for the right human arm. The exoskeleton is designed to apply a controllable force of up to 100 N at the center of the user’s hand palm, oriented along any spatial direction and it can provide active and tunable arm weight compensation. The device mechanical structure has been extensively described in [37], whereas a description of the model of its novel tendon transmission may be found in [38]. For sake of clarity, a brief review of the device kinematics will be provided in this section. L-Exos has 5 DoFs, 4 of which are actuated and are used to define the position of the end-effector in space (see Figure 1). The system is therefore redundant, allowing different joint configurations corresponding to the same end-effector position, which is fundamental for chronic stroke patients. Such subjects are likely to implement compensatory strategies in order to overcome force and Range of Motion (ROM) limitations remaining after stroke rehabilitation [39]. The 5th DoF

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

43

Figure 1. L-Exos kinematics.

is passive and allows free wrist pronation and supination movements. Moreover, design optimizations allow total arm mobility to a healthy subject wearing the device. The structure of the L-Exos is open, the wrist being the only closed joint, and can therefore be easily wearable by post-stroke patients with the help of a therapist. In order to use the L-Exos system for rehabilitation purposes, an adjustable height support was made, and a chair was placed in front of the device support, in order to enable patients to be comfortably seated while performing the tasks. The final handle length is also tunable, according to the patient’s arm length. After wearing the robotic device, the subject’s elbow is kept attached to the robotic structure by means of a belt. If necessary, the wrist may also be tightly attached to the device end-effector by means of a second belt, which was used for patients who were not able to fully control hand movements. A third belt can easily be employed in order to block the patient’s trunk when necessary. The L-Exos device was integrated with a projector used to display on a wide screen placed in front of the patient different virtual scenarios in which to perform rehabilitation exercises. The VR display is therefore a mono screen in which a 3D scene is rendered. Three Virtual Rehabilitation scenarios were developed using the XVR Development Studio [40]. The photo shown in Figure 2 was taken during a therapy session, while one of the admitted patients was performing the required exercises, and is useful to visualize the final clinical setup.

Figure 2. One admitted patient performing the robotic-aided therapy exercises.

44

2.

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

Methods

A clinical pilot study involving 9 subjects with the main objective of validating robotic assisted therapy with the L-Exos system was carried out at the Santa Chiara Hospital of Pisa, Italy, between March and August 2007. Potential subjects to be enrolled in the clinical protocol were contacted to take part in a preliminary test session used to evaluate patients acceptance of the device. Most of the patients gave an enthusiastic positive feedback about the opportunity. Patients who were declared fit for the protocol and agreed to sign an informed consent form concerning the novel therapy scheme were admitted to the clinical trials. The protocol consisted of 3 one-hour rehabilitation sessions per week for a total of six weeks (i.e., 18 therapy sessions). Each rehabilitation session consisted in three different VR mediated exercises. A brief description of the goal of each exercise will be provided in the next paragraphs, whereas a more detailed description of the VR scenarios developed may be found in previous works [35, 36]. Some relevant control issues concerning the proposed exercises will be reported as well. The patient was on a seat as shown in Figure 3(D), with his/her right forearm wearing the exoskeleton and a video projector displaying frontally the virtual scenario. A preliminary clinical test was conducted to evaluate the ergonomics of the system and the functionality as a rehabilitation device on a set of three different applications. The test was intended to demonstrate that the L-Exos could be successfully employed by a patient, and to measure the expected performance during therapy. To assess the functionality of the device, three different scenarios and corresponding exercises were devised: - A reaching task; - A motion task constrained to a circular trajectory; - An object manipulation task. The tasks were designed in order to be executed in succession within one therapy session of the duration of about one hour, repeated three times per week.

(A)

(B)

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

(C)

45

(D)

Figure 3. The arm exoskeleton during the execution of the reaching task. A: the starting position of the reaching task; B: a subject in the middle of the path of the reaching task; C: a subject at the end-point of the path of the reaching task; D: The overall system.

2.1. Reaching task In the first task, the represented scenario is composed of a virtual room, where different fixed targets are displayed to the patient as gray spheres disposed on a horizontal row, as shown in Figure 4. The position of the hand of the patient is shown as a green sphere, that is moved according to the end-effector movements. The starting position of the task was chosen as a rest position of the arm, with the elbow flexed at 90°, as shown in Figure 3(A). In this position, the exoskeleton provides the support for the weight of the arm, so that the patient can comfortably lean his arm on the exoskeleton. When one of the fixed targets is activated, a straight trajectory connecting the starting point and the final target is displayed in the simulation. The patient is instructed to actively follow the position of a yellow marker, whose motion is generated along the line connecting the start and end points according to a minimum jerk model [41], approximated by a 5th degree polynomial with a displacement profile as represented in Figure 5. The patient is asked to move the arm to reach the final target with a given velocity, minimizing the position error between the yellow marker that moves automatically toward the target, and his/her own marker, represented by the green sphere. The yellow marker reaches the target with zero velocity, and comes back on the blue line towards the initial position. The patient is alerted of the start of the exercise by a sound, that is generated automatically by the system. The therapist can set the maximum speed of the task, by choosing among three maximum speeds (v1 = 5 cm/s, v2 = 10 cm/s and v3 = 15 cm/s) and change the position of the fixed targets that should be reached by the patient, both in terms of target height and depth within the virtual room. The movement towards multiple targets disposed on the same row and backwards is activated in sequence, so that the patient can perform movements in both medial and lateral planes, reaching targets at the same height. There are 7 fixed targets placed symmetrically respect to the sagittal plane of the subject and the fixed targets can be disposed at two different heights relative to the start position of

46

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

the task (h1 = 0.01 m and h2 = 0.12 m). During each series, the height of the fixed target is not changed, and the following steps are executed in succession for each series: 1) The first movement is executed towards the leftmost fixed target; 2) Once the fixed target is reached the moving marker returns back to its start position, it stops for 2 seconds, and then it starts again towards the next target on the right; 3) The last target of each series is the rightmost one. In order to leave the patient the possibility to actively conduct the task and be passively guided by the robot only when he/she is unable to complete the reaching task, a suitable impedance control was developed. The control of the device is based on two concurrent impedance controls acting respectively along tangential and orthogonal directions to the trajectory. 2.2. Constrained motion task In the second exercise the patient is asked to move freely along a circular trajectory, as shown in Figure 6, where it is constrained by an impedance control. The virtual constraint is activated through a button located on the handle. Position, orientation and scale of the circular trajectory can be changed online, thus allowing the patient to move within different effective workspaces. No guiding force is

Figure 4. The virtual scenario visualized in the reaching task.

Figure 5. The motion profile to be followed by the patient in the reaching task.

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

47

Figure 6. Example of the free motion constrained to a circular trajectory.

applied to the patient’s limb when he/she is moving within the given trajectory, along which the patient is constrained by means of virtual springs. Also in this task the therapist can actively compensate the weight of the patient’s arm through the device, until the patient is able to autonomously perform the task. This is accomplished by applying torques at the level of the joints, based on a model of the human arm, with masses distributed along the different limbs with a proportion derived from anatomical data. The absolute value of the each limb mass is determined according to the weight of the subject. 2.3. Free motion task In this task the patient is asked to move cubes represented in the virtual environment, as shown for instance in figure 7, and to arrange them in a order decided by the therapist, e.g. putting the cubes with the same symbol or with the same color in a row, or putting together the fragments of one image. For this task the device is controlled with a direct force control, with the interaction force computed by a physics module based on the Ageia PhysX physics engine [42]. By pressing a button on the handle, the patient can decide to select which cube wants to move and release the cube through the same button. Collision with and between the objects are simulated through the physics engine, so that it is actually possible to perceive all the contact forces during the simulation. Also in this task the device can apply an active compensation of the weight of the patient arm, leaving to the therapist the possibility to decide the amount of weight reduction.

Figure 7. An example of the manipulation of objects task.

48

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

3. Therapy results The following paragraphs will describe the metrics used in order to quantitatively evaluate patients’ performance in the reaching task and in the path following task exercises. No quantitative data was computed for the last proposed task. A first obvious possible quantitative measure, such as task completion time, was thought as being not significant to evaluate patient performance improvements. This was due to the high variability in the task difficulty among different therapy sessions (initial cube disposition was randomly chosen by the control PC), and to the high variability in patient’s attitude to consider the exercise as completed, i.e. the accepted amount of cube misalignment and hence the amount of time spent in trying to perform fine movements to reduce such misalignment. 3.1. Reaching task Figure 8 shows a typical path followed by a patient during the reaching task. The cumulative error for each task was chosen as being the most significant metric to analyze reaching data. After the definition of a target position and of a nominal task speed, the cumulative error in the reaching task is computed for iterations corresponding to the given target position and speed. The cumulative error curves are then fitted in a least square sense by a sigmoid-like 3-parameter curve, represented with Eq. (1), where s is the cumulative error at time t, whereas a, b and c are fitting parameters. Fitting curves are then grouped and averaged on a therapy session basis, each set containing the fitting curves computed for a single rehabilitation session. Sample data resulting from this kind of analysis are shown in Figure 9, where a greater dash step indicates a later day when a given target was required to be reached with a given peak speed. It is to be said that statistically significant improvements in the average fitting curves from Week 1 to Week 6 are recognizable for more than half targets in only 4 out of 9 patients enrolled in the protocol. A typical improvement pattern for a sample target is shown in Panel A of Figure 9 for Patient 6. This patient is constantly improving his performance in the exercise, leading to a significant (1)

Figure 8. Typical path followed during a reaching task – Blue straight line: ideal trajectory, Red: actual trajectory.

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

49

decrease in the final cumulative error for a given target. A reducing of the mean slope of the central segment of the fitting curve is therefore present, thus indicating a higher ability to maintain a constant average error throughout the task. Panel B of Figure 9 reveals an interesting aspect of the application of the belt used to avoid undesired back movements. During the first therapy sessions, no belt was present, and each therapy session registered a comparable value of the cumulative error. As soon as the trunk belt is introduced, the error increases dramatically, as formerly employed compensatory strategies are not allowed. However, due to the fact that active patient’s movements become much more stimulated, the cumulative error fitting curve improves significantly. It is to be noted that, by the end of the therapy, values which are nearly comparable to the ones obtained in the no-belt condition are reached. 3.2. Path following task Total time required to complete a full circular path was the quantitative parameter used to assess patient improvement for the constrained motion task. 3D position data were projected onto a best fitting plane (in the sense of least squares), and the best fit circle was computed for the projected points. Time to complete a turn was then evaluated with regard to trajectory. Curvature along the trajectory, which is irregular for the three patients, was not evaluated. In particular, due to the deliberately low value of the stiffness which realizes the motion constraint, patients sometimes move in an unstable way, bouncing from the internal side to the external side of the trajectory and vice versa, requiring some time to gain the control of their movements again. This behavior has detrimental effects on curvature computation. Although three of the patients report no significant decrease of the completion time from Week 1 to Week 6, three patients report a decrease of about 50% in the task completion time, whereas three other patients report a decrease of about 70% of the same performance indicator. Such results are significant from a statistical point of view (p < 0.001 for the t-Student test for each patient showing improvements). Sample data from Patient 3 are shown in Figure 10, in order to visualize a typical trend which was found in the patients reporting improvements in the motion constrained exercise. It is interesting to note that, along with the significant reduction in the mean time required to complete a circle, a significant reduction of

A Figure 9. A: sample reaching results for Patient 6;

B B: sample reaching results for Patient 3.

50

A. Frisoli et al. / Robotic Assisted Rehabilitation in Virtual Reality with the L-EXOS

Figure 10. Sample constrained motion task results - Patient 3.

the associated standard deviation is recognizable, hence suggesting an acquired ability of performing the exercise with a much higher regularity level.

4. Clinical results All patients were evaluated by means of standard clinical evaluation scales: • Fugl-Meyer scale: this scale [43] is used for the evaluation of motor function, of balance, and of some sensation qualities and joint function in hemiplegic patients. The Fugl-Meyer assessment method applies a cumulative numerical score. The whole scale consists of 50 items, for a total of 100 points, each item being evaluated in a range from 0 to 2.33 items concern upper limb functions (for a total of 66 points) and are used for the clinical evaluations. • Modified Ashworth scale: it is the most widely used method for assessing muscle spasticity in clinical practice and research. Its items are marked with a score ranging from 0 to 5, the greater the score, the greater being the spasticity level. Only patients with modified Ashworth scale values ≤ 2 were admitted to this study. • Range Of Motion: it is the most classical and evident parameter used to assess motor capabilities of impaired patients. Clinical improvements in each scale have been observed by the end of the therapy protocol for every patient, and they will now be discussed. 4.1. Fugl-Meyer assessment The Fugl-Meyer assessment was carried out before and after robotic therapy. Every patient reported a significant increment ranging from 1 to 8 points, 4 points (out of 66) being the average increment (p