Atlas of Sleep Medicine [3 ed.] 3031346246, 9783031346248

Citation preview

Robert J. Thomas Sushanth Bhat Sudhansu Chokroverty Editors

Atlas of SSleep leep Medicine Third Edition

Atlas of Sleep Medicine

Robert J. Thomas  •  Sushanth Bhat Sudhansu Chokroverty Editors

Atlas of Sleep Medicine Third Edition

Editors Robert J. Thomas Beth Israel Deaconess Medical Center Boston, MA, USA Sudhansu Chokroverty HMH JFK University Medical Center and Neuroscience Institute Hackensack Meridian School of Medicine at Seton Hall University South Orange and Nutley, NJ, USA

Sushanth Bhat JFK Neuroscience Institute Hackensack Meridian Health-JFK University Medical Center Edison, NJ, USA

Rutgers Robert Wood Johnson University Medical Center New Brunswick, NJ, USA

ISBN 978-3-031-34624-8    ISBN 978-3-031-34625-5 (eBook) https://doi.org/10.1007/978-3-031-34625-5 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To my patients, whose misfortunes and the never-ending struggle to reduce their pain motivated this work. To my wife Subha Ramani, who never fails to keep me happy, honest, and grounded, including “you cannot save the world.” Well, I will try. To my friend Sudha Seshadri, for memories created, experiences shared, and some lessons learnt. To Rena Holzer and Kimberly Campbell, nurse practitioners extraordinaire, for holding the metaphorical fort, enabling this indulgence. Robert J. Thomas Dedicated to my parents and family, and with gratitude to all the students, residents, and fellows who, over the years, have kept the spirit of learning alive during busy clinics, hectic rounds, and piles of pending notes and sleep study reports. Sushanth Bhat

Preface

One more Atlas? For centuries, visual depiction of pathology has been a core feature of medical sciences. As new technologies have entered the field, the visuals have become more detailed and complex and been taking forms never previously imagined. The teaching and learning potential of visual imagery is nearly infinite, with so many possible interpretations of any individual “work.” The First Edition of the Atlas of Sleep Medicine (2005), led by the incomparable Dr. Sudhansu Chokroverty and where I was privileged to assist (with Dr. Meeta Bhatt), aimed to “teach sleep medicine through images.” The field of sleep medicine/science is extraordinarily rich with physiological and pathological patterns, blending elements from already visually rich areas of medicine, including neurophysiology, recording technologies, pulmonology, cardiology, and neurology. We hoped to tap into these riches for large-scale benefit and think that we reasonably succeeded. The Second Edition in 2013 continued with that goal. There was some hesitation about embarking on a Third Edition. Clearly, there had been tremendous progress in sleep sciences/practice, but was there enough new visual material? Dr. Chokroverty stepped back slightly to hand me the responsibility to shepherd this edition, and we also welcomed a new co-editor, Dr. Sushanth Bhat. It was quickly evident that there was SO MUCH new visual information! Even better, leading proponents of the art and science of sleep medicine were willing to share their riches—and thus, this Third Edition. We do not pretend that this is a “Textbook of Sleep Medicine” but a journey through some of the most exciting data realms and elements of the field. This Third Edition attests to the vibrancy of sleep medicine in general, and the change across editions reflects the evolution of the field and some future directions, which the editors are grateful to have the opportunity to showcase. Boston, MA, USA Nutley, NJ, USA  Edison, NJ, USA 

Robert J. Thomas Sushanth Bhat Sudhansu Chokroverty

vii

Acknowledgments

To the numerous contributors to this Atlas—no words of appreciation may be adequate. They contributed time, effort, and knowledge during an incredibly difficult period for humanity, with a specific type of toll on healthcare providers and researchers. This book was “conceived” just prior to the Covid-19 pandemic, and of all the challenges we anticipated, the virus was not one of them. The publishers and our managing editor Lee Klein showed infinite patience as deadlines showed distinct shape-shifting, but here we are finally. Dr. Chokroverty would like to acknowledge his wife, Manisha Chokroverty, for her patience and forbearance during the production of this edition despite her considerable disability and comorbidities.

ix

Contents

Introduction�������������������������������������������������������������������������������������������������������������������������   1 Robert J. Thomas, Sushanth Bhat, and Sudhansu Chokroverty Part I Basic Science  Neuronal Circuits, Anatomical Substrates, and Mechanism of Sleep-Wakefulness�����   5 Sudhansu Chokroverty, Sushanth Bhat, and Robert J. Thomas Part II Laboratory Techniques Polysomnographic Recording Technique�������������������������������������������������������������������������  17 Sudhansu Chokroverty and Sushanth Bhat  Hypnogram and Compliance Graph Analysis�����������������������������������������������������������������  43 Robert J. Thomas, Sudhansu Chokroverty, and Sushanth Bhat  Electroencephalography for the Sleep Specialist�������������������������������������������������������������  69 Sudhansu Chokroverty, Eli S. Neiman, and Sushanth Bhat  Odds Ratio Product (ORP): Description and Implications to Understanding and Management of Sleep Disorders ��������������������������������������������������������������������������������������� 109 Magdy Younes Sleep Stage Scoring������������������������������������������������������������������������������������������������������������� 125 Raman K. Malhotra and Alon Y. Avidan Sleep-Disordered Breathing: Scoring������������������������������������������������������������������������������� 165 Robert J. Thomas, Sushanth Bhat, Federica Provini, and Sudhansu Chokroverty Upper Airway Radiology��������������������������������������������������������������������������������������������������� 201 Richard J. Schwab and Andrew Wiemken Part III Other Sleep Techniques Multiple Sleep Latency Test (MSLT) and Maintenance of Wakefulness Test (MWT)������������������������������������������������������������������������������������������������������������������������� 235 Sushanth Bhat, Stacey D. Elkhatib Smidt, and Sudhansu Chokroverty Actigraphy��������������������������������������������������������������������������������������������������������������������������� 243 Marco Zucconi, Samantha Mombelli, and Sudhansu Chokroverty  Cyclic Alternating Pattern (CAP): Scoring Rules and Clinical Applications��������������� 261 Liborio Parrino, Irene Pollara, Francesco Rausa, Marcello Luigi Salvatelli, and Carlota Mutti

xi

xii

Recommendations for Practical Use of Pulse Transit Time as a Tool for Respiratory Effort Measurement During Sleep and Microarousal Recognition������������������������������������������������������������������������������������������� 277 Jean-Louis Pépin, Sébastien Baillieul, and Renaud Tamisier Peripheral Arterial Tonometry ����������������������������������������������������������������������������������������� 293 Robert J. Thomas EEG Spectrograms������������������������������������������������������������������������������������������������������������� 299 Robert J. Thomas and Haoqi Sun Cardiopulmonary Coupling����������������������������������������������������������������������������������������������� 307 Robert J. Thomas Drug-Induced Sleep Endoscopy���������������������������������������������������������������������������������������� 325 Phillip Huyett Upper Airway Endotypes��������������������������������������������������������������������������������������������������� 327 Andrew Wellman and Ran R. Liu Autonomic Function Testing Including Autonomic Nervous System and Sleep Interaction��������������������������������������������������������������������������������������������������������� 333 Sudhansu Chokroverty and Sushanth Bhat  Scoring Techniques for Sleep-Related Movements ��������������������������������������������������������� 341 Ambra Stefani, Birgit Högl, and Raffaele Ferri  Artificial Intelligence in Sleep Medicine��������������������������������������������������������������������������� 355 Haoqi Sun, Wolfgang Ganglberger, M. Brandon Westover, and Robert J. Thomas Neuroimaging Techniques ������������������������������������������������������������������������������������������������� 371 Zara Duquette, Nathan Cross, Aurore A. Perrault, Pierre Maquet, Martin Desseilles, and Thien Thanh Dang-Vu  High-Resolution EEG Characterization of Sleep Neurophysiology������������������������������� 389 Shijing Zhou, Kyle Morgan, Evan Hathaway, Roma Shusterman, Phan Luu, Miranda Lim, Ruth Benca, and Don M. Tucker  Intracranial EEG/MEG Recording and Sleep Medicine ����������������������������������������������� 417 Richard A. Wennberg  Pictorial Diagnosis of Circadian Rhythm Sleep-Wake Disorders ��������������������������������� 439 Sabra M. Abbott, Roneil Malkani, and Phyllis C. Zee  Consumer Sleep Technology: Wearables and Nearables������������������������������������������������� 447 Sushanth Bhat and Liudmila Lysenko Part IV Clinical Aspects  Motor Disorders in Sleep��������������������������������������������������������������������������������������������������� 459 Sushanth Bhat, Liudmila Lysenko, Stacey D. Elkhatib Smidt, Federica Provini, Marco Zucconi, Mauro Manconi, and Sudhansu Chokroverty Cardiac Arrhythmias��������������������������������������������������������������������������������������������������������� 485 Robert J. Thomas, Sudhansu Chokroverty, and Sushanth Bhat Sleep and Epilepsy ������������������������������������������������������������������������������������������������������������� 511 Lino Nobili, Angelica Montini, Marco Zucconi, Sudhansu Chokroverty, and Federica Provini

Contents

Contents

xiii

Sleep Dysfunction and Sleep-­Disordered Breathing in Miscellaneous Neurological Disorders������������������������������������������������������������������������������������������������������� 525 Sudhansu Chokroverty, Sushanth Bhat, Federica Provini, and Greta Mainieri Atypical PSG Patterns ������������������������������������������������������������������������������������������������������� 555 Sudhansu Chokroverty, Sushanth Bhat, and Robert J. Thomas Part V Therapy in Sleep-Disordered Breathing: Non-invasive  CPAP, Bilevel, APAP, High and Low Loop Gain Syndromes����������������������������������������� 595 Robert J. Thomas Adaptive Servo Ventilation������������������������������������������������������������������������������������������������� 657 Robert J. Thomas  Oral Appliances for Obstructive Sleep Apnea Syndrome (OSAS) Therapy����������������� 671 Fernanda R. Almeida and Mona M. Hamoda Part VI Therapy in Sleep-Disordered Breathing: Invasive Neurostimulation and Sleep Apnea����������������������������������������������������������������������������������� 679 Robert J. Thomas  Upper Airway Surgery for Obstructive Sleep Apnea ����������������������������������������������������� 697 John A. Fleetham Part VII Pediatric Sleep Medicine (Clinical and Laboratory Techniques) Pediatric Polysomnography����������������������������������������������������������������������������������������������� 709 Jisu Han, Alexandra Schwarz, Stacey D. Elkhatib Smidt, and Timothy F. Hoban Part VIII Video Vignettes  Sleep Disturbances in Autoimmune Encephalitis������������������������������������������������������������ 741 Margaret S. Blattner  Video-Clinical (SHE, FFI, RBD, and Others) ����������������������������������������������������������������� 757 Federica Provini and Greta Mainieri Narcolepsy: Unequivocal Diagnosis After Split-Screen, Video-­Polysomnographic Analysis of a Prolonged Cataplectic Attack��������������������������������������������������������������������� 759 Mark Eric Dyken, Deborah C. Lin-Dyken, and Jason Maxfield  Obstructive Sleep Apnea Associated with Cerebral Hypoxemia ����������������������������������� 763 Mark Eric Dyken, Christine L. Glenn, and George B. Richerson  The REM Sleep Behavior Disorder Leading to a Subdural Hemorrhage��������������������� 769 Mark Eric Dyken, Deborah C. Lin-Dyken, and Mark Raymond Dyken  Isolated Sleep Paralysis: An REM-“Sleep” Polysomnographic Phenomenon, as Documented with Simultaneous Clinical and Electrophysiologic Assessment������������� 773 Mark Eric Dyken, Deborah C. Lin-Dyken, and Nivedita Jerath Confusional Arousal����������������������������������������������������������������������������������������������������������� 777 Mark Eric Dyken, Deborah C. Lin-Dyken, and Nivedita Jerath Sleepwalking����������������������������������������������������������������������������������������������������������������������� 779 Mark Eric Dyken, Deborah C. Lin-Dyken, and Siddharth Bajpai

xiv

Hypnagogic Hallucination������������������������������������������������������������������������������������������������� 783 Mark Eric Dyken, Deborah C. Lin-Dyken, and Nivedita Jerath Sleep Terrors����������������������������������������������������������������������������������������������������������������������� 787 Mark Eric Dyken, Deborah C. Lin-Dyken, and Jeffrey Boyle  Adult Seizure from REM Sleep After Obstructive Apnea����������������������������������������������� 791 Mark Eric Dyken and Mark Raymond Dyken  Pediatric Sleep-Related Seizure with Obstructive Apnea����������������������������������������������� 795 Mark Eric Dyken and Mark Raymond Dyken  Nocturnal Frontal Lobe Epilepsy ������������������������������������������������������������������������������������� 797 Mark Eric Dyken and Mark Raymond Dyken Sleep-Related Bruxism������������������������������������������������������������������������������������������������������� 799 Mark Eric Dyken and Mark Raymond Dyken Sleep-Related Eating Disorder������������������������������������������������������������������������������������������� 803 Mark Eric Dyken and Mark Raymond Dyken  Sleep-Related Rhythmic Movement Disorder����������������������������������������������������������������� 805 Mark Eric Dyken and Mark Raymond Dyken Sleep-Breathing Dynamics������������������������������������������������������������������������������������������������� 807 Robert J. Thomas Part IX Epilogue Epilogue������������������������������������������������������������������������������������������������������������������������������� 811 Robert J. Thomas, Sushanth Bhat, and Sudhansu Chokroverty Index������������������������������������������������������������������������������������������������������������������������������������� 813

Contents

Contributors

Sabra  M.  Abbott  Department of Neurology, Northwestern University Feinberg School of Medicine, Chicago, IL, USA Fernanda R. Almeida  University of British Columbia Oral Health Sciences, Vancouver, BC, Canada Alon Y. Avidan  Department of Neurology, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA Sébastien Baillieul, MD, PhD  Univ. Grenoble Alpes, Inserm, CHU Grenoble Alpes, Service Universitaire de Pneumologie Physiologie, Grenoble, France Siddharth Bajpai  Northern Virginia Psychiatric Group, Fairfax, VA, USA Ruth  Benca Department of Psychiatry and Behavioral Medicine, Wake Forest School of Medicine, Winston-Salem, NC, USA Sushanth Bhat  Department of Neurology, Hackensack Meridian School of Medicine, Nutley, NJ, USA JFK Neuroscience Sleep Center, Hackensack Meridian JFK University Medical Center, Edison, NJ, USA JFK Neuroscience Institute Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA Margaret  Blattner, MD, PhD Department of Neurology, Beth Israel Deaconess Medical Center, Boston, MA, USA Jeffrey Boyle  Avera Medical Group Neurology, Sioux Falls, SD, USA Sudhansu Chokroverty  HMH JFK University Medical Center and Neuroscience Institute, Hackensack Meridian School of Medicine at Seton Hall University, South Orange and Nutley, NJ, USA Rutgers Robert Wood Johnson University Medical Center, New Brunswick, NJ, USA Nathan  Cross Institut Universitaire de Gériatrie de Montréal and CRIUGM, CIUSSS du Centre-Sud-de-l’Ile-de-Montréal, Montreal, QC, Canada PERFORM Centre, Concordia University, Montreal, QC, Canada Center for Studies in Behavioral Neurobiology, Department of Health, Kinesiology and Applied Physiology, Concordia University, Montreal, QC, Canada Thien  Thanh  Dang-Vu Institut Universitaire de Gériatrie de Montréal and CRIUGM, CIUSSS du Centre-Sud-de-l’Ile-de-Montréal, Montreal, QC, Canada Department of Health, Kinesiology and Applied Physiology, Center for Studies in Behavioral Neurobiology and PERFORM Center, Concordia University, Montreal, QC, Canada PERFORM Centre, Concordia University, Montreal, QC, Canada Martin Desseilles  Department of Psychology, University of Namur, Namur, Belgium xv

xvi

Zara  Duquette Department of Health, Kinesiology and Applied Physiology, Center for Studies in Behavioral Neurobiology and PERFORM Center, Concordia University, Montreal, QC, Canada Mark Eric Dyken  The Department of Neurology, The University of Iowa Roy J. and Lucille A. Carver College of Medicine, Iowa City, IA, USA Mark  Raymond  Dyken  Department of Neurology, University of Iowa Roy J. and Lucille A. Carver College of Medicine, Iowa City, IA, USA The University of Iowa Roy J and Lucille A Carver College of Medicine, Iowa City, IA, USA Stacey  D.  Elkhatib  Smidt JFK Neuroscience Sleep Center, Hackensack Meridian JFK University Medical Center, Edison, NJ, USA JFK Neuroscience Institute, Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA Raffaele  Ferri Sleep Research Centre, Department of Neurology IC, Oasi Research Institute—IRCCS, Troina, Italy John A. Fleetham  University of British Columbia, Vancouver, BC, Canada Wolfgang  Ganglberger  Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA Christine  L.  Glenn  Department of Neurology Sleep Disorders Center, University of Iowa Roy J. and Lucille A. Carver College of Medicine, Iowa City, IA, USA Mona M. Hamoda  Department of Oral Health Sciences, Faculty of Dentistry, University of British Columbia, Vancouver, BC, Canada Jisu Han  Neuroscience Sleep Center, Hackensack Meridian JFK University Medical Center, Edison, NJ, USA Evan Hathaway  Brain Electrophysiology Laboratory Company, Eugene, OR, USA Timothy F. Hoban  Departments of Pediatrics and Neurology, The Michael S. Aldrich Sleep Disorders Laboratory, University of Michigan, Ann Arbor, MI, USA Birgit  Högl Sleep Disorders Clinic, Department of Neurology, Medical University of Innsbruck, Innsbruck, Austria Phillip  Huyett Department of Otolaryngology-Head and Neck Surgery, Harvard Medical School, Boston, MA, USA Nivedita Jerath  AdventHealth, Winter Park, FL, USA Miranda Lim  Department of Psychology, University of Oregon, Eugene, OR, USA Deborah  C.  Lin-Dyken The Division of Developmental and Behavioral Pediatrics, The University of Iowa Roy J. and Lucille A. Carver College of Medicine, Iowa City, IA, USA Ran R. Liu  Harvard Medical School, Massachusetts General Hospital, Boston, MA, USA Phan Luu  Department of Psychology, University of Oregon, Eugene, OR, USA Liudmila Lysenko  Ochsner Health System, New Orleans, LA, USA Greta Mainieri  Department of Biomedical and Neuromotor Sciences, University of Bologna, Bologna, Italy Raman  K.  Malhotra Sleep Medicine Section, Department of Neurology, Washington University School of Medicine, St. Louis, MO, USA

Contributors

Contributors

xvii

Roneil  Malkani Neurology Service, Department of Neurology, Northwestern University Feinberg School of Medicine, Jesse Brown Veterans Affairs Medical Center, Chicago, IL, USA Mauro Manconi  Sleep Medicine Unit, Neurocenter of the Southern Switzerland, Regional Hospital of Lugano, Lugano, Switzerland Faculty of Biomedical Sciences, Università della Svizzera Italiana, Lugano, Switzerland Pierre Maquet  Department of Neurology, University Hospital of Liège, University of Liège, Liège, Belgium Jason Maxfield  Freeman Health System Sleep Laboratory, Joplin, MO, USA Samantha  Mombelli Department of Clinical Neurosciences, Neurology, Sleep Disorders Center, IRCCS San Raffaele Scientific Institute, Milan, Italy Center for Advanced Research in Sleep Medicine, Research center of the Centre intégré universitaire de santé et de services sociaux du Nord de l’Île-de-Montréal, Montreal, Canada IRCCS San Raffaele Scientific Institute, Milan, Italy Angelica  Montini Department of Biomedical and Neuromotor Sciences, University of Bologna, Bologna, Italy Kyle Morgan  Brain Electrophysiology Laboratory Company, Eugene, OR, USA Carlota Mutti  Sleep Disorders Center, University of Parma, Parma, Italy Eli S. Neiman  St. Francis Medical Center, Boca Raton, FL, USA Lino Nobili  IRCCS Istituto delle Scienze neurologiche di Genova, Istituto Giannina Gaslini, Genova, Italy Liborio Parrino  Neurology Unit, Parma University Hospital, Parma, Italy Jean-Louis Pepin, MD, PhD  Univ. Grenoble Alpes, Inserm, CHU Grenoble Alpes, Service Universitaire de Pneumologie Physiologie, Grenoble, France Aurore A. Perrault  Institut Universitaire de Gériatrie de Montréal and CRIUGM, CIUSSS du Centre-Sud-de-l’Ile-de-Montréal, Montreal, QC, Canada PERFORM Centre, Concordia University, Montreal, QC, Canada Center for Studies in Behavioral Neurobiology, Department of Health, Kinesiology and Applied Physiology, Concordia University, Montreal, QC, Canada Irene Pollara  Sleep Disorders Center, Department of Medicine and Surgery, University of Parma, Parma, Italy Federica  Provini IRCCS Istituto delle Scienze Neurologiche, UOC NeuroMet, Bellaria Hospital, Bologna, Italy Department of Biomedical and Neuromotor Sciences, University of Bologna, Bologna, Italy Francesco Rausa  Sleep Disorders Center, University of Parma, Parma, Italy George  B.  Richerson Department of Neurology, University of Iowa Roy J. and Lucille A. Carver College of Medicine, Iowa City, IA, USA Marcello Luigi Salvatelli  Sleep Disorders Center, University of Parma, Parma, Italy Richard  J.  Schwab Division of Sleep Medicine, Department of Medicine, University of Pennsylvania Medical Center, Philadelphia, PA, USA Alexandra  Schwarz Neuroscience Sleep Center, Hackensack Meridian JFK University Medical Center, Edison, NJ, USA

xviii

Roma Shusterman  Brain Electrophysiology Laboratory Company, Eugene, OR, USA Ambra  Stefani Sleep Disorders Clinic, Department of Neurology, Medical University of Innsbruck, Innsbruck, Austria Haoqi  Sun Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA Renaud  Tamisier, MD, PhD, MBA  Univ. Grenoble Alpes, Inserm, CHU Grenoble Alpes, Service Universitaire de Pneumologie Physiologie, Grenoble, France Robert J. Thomas  Department of Medicine, Division of Pulmonary, Critical Care and Sleep, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA Don M. Tucker  Department of Psychology, University of Oregon, Eugene, OR, USA Andrew Wellman  Division of Sleep Medicine, Harvard Medical School, Boston, MA, USA Division of Sleep Medicine and Circadian Disorders, Harvard Medical School, Boston, MA, USA Richard  A.  Wennberg University of Toronto, Krembil Brain Institute, Toronto Western Hospital, Toronto, ON, Canada M.  Brandon  Westover  Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Belmont, MA, USA Andrew  Wiemken Division of Sleep Medicine, Department of Medicine, University of Pennsylvania Medical Center, Philadelphia, PA, USA Magdy Younes  Sleep Disorders Centre, University of Manitoba, Winnipeg, MB, Canada Phyllis  C.  Zee Department of Neurology, Center for Circadian and Sleep Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, USA Shijing Zhou  Brain Electrophysiology Laboratory Company, Eugene, OR, USA Marco Zucconi  Department of Clinical Neurosciences, Neurology, Sleep Disorders Center, IRCCS San Raffaele Scientific Institute, Milan, Italy

Contributors

Introduction Robert J. Thomas, Sushanth Bhat, and Sudhansu Chokroverty

Every squig and shape recorded during sleep has a story. This Atlas tries to tell their story. Humans are highly visual animals. More than half the brain is devoted to visual processing. There are exquisite mechanisms to detect visual contrast, movement and color, while visual creativity from artists to animations to architecture keeps pushing boundaries. We incessantly seek visual stimulation in nature/big screen/small screen, and light/dark cycles have profound effects on our biology. Some fields of medical practice are highly visual—pathology, surgery, radiology, dermatology and neurophysiology in general, but in visual data is “everywhere” (an Atlas of—seems to be on every corner). Throughout the history of medical practice and research, we have tried to simplify the designation of visually complex data, creating better digestible reports containing numbers, ratios, indices and tabulations. There is a real risk that in this process of data reduction, information of interest is lost. Sleep Medicine is the epitome of lost treasures. Polysomnography by its very definition is visually complex. Numerous data streams from biological subsystems of interest are collected and displayed, then scored, and con-

R. J. Thomas (*) Department of Medicine, Division of Pulmonary, Critical Care and Sleep, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA e-mail: [email protected] S. Bhat Department of Neurology, Hackensack Meridian School of Medicine, Nutley, NJ, USA JFK Neuroscience Institute, Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA S. Chokroverty HMH JFK University Medical Center and Neuroscience Institute, Hackensack Meridian School of Medicine at Seton Hall University, South Orange and Nutley, NJ, USA Rutgers Robert Wood Johnson University Medical Center, New Brunswick, NJ, USA

verted into a few discrete numbers. We are then essentially blind to the majority of the data content, living in various degrees of darkness. Computer-assisted scoring and machine learning have largely focused on task efficiency than interpretive sophistication. Even reduced data including single streams like actigraphy and oximetry can be analyzed by a wide range of methods, each with its own simplified graphical/numerical output. Vertical (across subsystems) and horizontal (time of recording) integration of data brings out the “art” in experienced readers of sleep data, but largely remains an individual skill developed over time and at best partially transferred to trainees. While the future may bring into routine practice artificial intelligence approaches which can mimic and even exceed the experienced “sleep study reader,” we seem to be a long way from there and something more immediately accessible is a need to better use the visual data waterfall soaking us daily (nightly). The third edition of the Atlas of Sleep Medicine’s goal is to bring some light into this darkness. This edition can stand on its own, but visual data rarely becomes obsolete, so previous editions remain relevant. Much of the material is new, and besides an effort to bring in new images, there is an abundance of new chapters, including upper airway endotypes, updated sleep-breathing chapters with apnea endotypes and phenotypes, electroencephalographic spectrograms of clinical polysomnograms, artificial intelligence, drug-­ induced sedation endoscopy, hypoglossal and phrenic nerve stimulation, high density and intracranial electroencephalography, autoimmune brain syndromes, the inevitable chapter on wearables, and new videos. We had to stop somewhere to keep the size of the book manageable, so there are some missing or relatively low-key elements which perhaps are best targeted by independent works or are somewhat out of scope for this Atlas, e.g., intensive care electroencephalography and sleep, slow and infra-slow oscillations (SO, ISO) of non-rapid eye movement (NREM) sleep (though there is a chapter on the Cyclic Alternating Pattern, which is an ISO), detailed discussions of ventilator waveforms, an exhaustive review of home sleep-­breathing therapy data, and the uni-

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. J. Thomas et al. (eds.), Atlas of Sleep Medicine, https://doi.org/10.1007/978-3-031-34625-5_1

1

2

verse of circadian visual data elements across human and non-human organisms. The very nature of visual data is that it is subject to individual differences in interpretation. Contributors to this Atlas have provided one view for any given image, which may at times be generally agreed u pon, but other images are likely open to other equally valid explanations. This Atlas is for everyone—technicians (who face this data routinely), sleep medicine trainees at any level (welcome to Sleep Medicine!), practicing sleep physicians (who are unlikely to have individually experienced all the presented data), and sleep researchers (to better understand the enemies we face in the clinic and to better focus their tools).

R. J. Thomas et al.

How to best use this Atlas? A full read from start to finish is best for those less familiar with sleep data, but any individual chapter can be read in any order for benefit. There is a certain joy in grabbing an interesting screenshot or video and stashing it away for future learning and teaching. This joy is evident in the effort put in by the various authors contributing to the Atlas; indeed, new chapters and images were slipping in until just before initiating the final layout; our very capable Publishing Editor Lee Klein finally said “stop! No more!!.” We hope that the reader will also share in this joy, and by reading the Atlas be inspired to embark on their own journey of visual learning.

Part I Basic Science

Neuronal Circuits, Anatomical Substrates, and Mechanism of Sleep-Wakefulness Sudhansu Chokroverty, Sushanth Bhat, and Robert J. Thomas

1 Introduction In addition to the three traditionally described neurobehavioral states of wakefulness, non-rapid eye movement (NREM) sleep and rapid eye movement (REM) sleep [1, 2], the senior author has previously proposed an additional three [3], making for a total of six: • Active Wakefulness (AW) • Relaxed Wakefulness (RW) • Predormitum, denoting a stage between RW and stage 1 of NREM sleep [4] • NREM sleep, in turn consisting of (a) Stage 1 NREM sleep signifying pre-sleep behavior with generators in the brain stem separate from those for other stages of NREM and REM sleep (b) Stage N2 (lighter stage of NREM sleep) with appearance of characteristic sleep spindles and K complexes accompanied by delta waves (0.5–2 Hz) occupying less than 20% of each epoch (c) Stage N3 Slow wave sleep (SWS) or deeper stage of NREM sleep) with delta waves occupying 20% or more in each epoch

S. Chokroverty HMH JFK University Medical Center and Neuroscience Institute, Hackensack Meridian School of Medicine at Seton Hall University, South Orange and Nutley, NJ, USA Rutgers Robert Wood Johnson University Medical Center, New Brunswick, NJ, USA S. Bhat (*) Department of Neurology, Hackensack Meridian School of Medicine, Nutley, NJ, USA JFK Neuroscience Institute, Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA R. J. Thomas Department of Medicine, Division of Pulmonary, Critical Care and Sleep, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA e-mail: [email protected]

• REM sleep with its distinctive signatures of rapid eye movements (REMs), muscle atonia, and low amplitude mixed frequency electroencephalogram (EEG) waves as well as sawtooth waves • Post-dormitum (also known as sleep inertia or sleep drunkenness, a stage between the end of night’s sleep and full wakefulness in the morning). This is a transient physiological state characterized by hypovigilance, grogginess, confusion, and impaired cognitive and behavioral performance. The neuroanatomical substrates controlling various states of wakefulness and sleep are the subject of discussion of this chapter.

2 Neuroanatomical Substrates and Control of Wake-Sleep States The areas of the central nervous system (CNS) that control sleep and wakefulness do not exist as distinct “centers” but rather as a series of interconnected circuits and systems. These are discussed below. The neuroanatomical substrates and neurotransmitters involved in the various stages of the wake-sleep cycle are summarized in Table 1.

2.1 Neuroanatomical Substrates and Control of Wakefulness The principal neuroanatomical substrate responsible for wakefulness, arousal, and vigilance is the ascending reticular activating system (ARAS), a part of the reticular formation (RF) of the brain stem [5, 6] (Fig. 1). The ARAS in the central core of the upper pons and midbrain receives collaterals from the specific afferent sensory systems in the lateral part of the brain stem transmitting environmental stimuli to the thalamus and sensory cerebral cortex. The RF has both ascending and descending projections.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. J. Thomas et al. (eds.), Atlas of Sleep Medicine, https://doi.org/10.1007/978-3-031-34625-5_2

5

S. Chokroverty et al.

6 Table 1  Neuroanatomical substrates and neurotransmitters involved in wakefulness and sleep State Neuroanatomical substrates Wakefulness • Reticular formation (RF), particularly the ascending reticular activating system (ARAS) and arousal • Diffuse thalamocortical projections • Cholinergic projections to the thalamus from the laterodorsal tegmental (LDT) and pedunculopontine (PPT) nuclei • Ascending arousal system originating from locus coeruleus (LC) noradrenergic and dorsal raphe (DR) serotonergic neurons through the basal forebrain via septohippocampal pathway to hippocampus and other parts of cerebral cortex • Lateral hypothalamic orexin/hypocretin neurons with their diffuse ascending and descending projections covering the entire central nervous system • Posterior hypothalamic tuberomammillary histaminergic projection through the forebrain • Cholinergic projections from the basal forebrain nucleus basalis of Meynert diffusely to the cerebral cortex • Glutamatergic projection from precoeruleus—From dorsal part of sublaterodorsal (SLD) tegmental nuclei through parabrachial area (PBA) in pons and through the basal forebrain diffusely to the cerebral cortex • Dopaminergic neurons from ventrolateral periaqueductal gray (VLPAG), ventral tegmental area (VTA) and substantia nigra (SN) pars compacta projecting widely • Parts of prosencephalon (forebrain) NREM sleep • Ventrolateral preoptic (VLPO) (clustered) and median preoptic (MnPO) nuclei of the anterior hypothalamus • Nucleus tractus solitarius (NTS) and parafacial zone (Pz) in the medullary region (minor role)

REM sleep

• Pontine SLD tegmental nucleus in rats (equivalent to perilocus coeruleus alpha in cats and subcoeruleus nucleus in humans): Glutamatergic • LDT and PPT nuclei at pontomesencephalic junction: Cholinergic • Lateral hypothalamic melanin-concentrating hormone (MCH)/GABAergic neurons next to orexin/hypocretin • Anterior hypothalamic extended (diffuse) VLPO neurons • Anterior hypothalamic suprachiasmatic nuclei • Part of limbic cortex including amygdala, hippocampus, anterior cingulate gyrus, bed nucleus of stria terminalis

Neurotransmitters • Acetylcholine • Noradrenaline • Orexin/hypocretin • Serotonin (5-hydroxytryptamine) • Histamine • Dopamine • Glutamate • Aspartate

• Gamma-aminobutyric acid (GABA): Inhibitory • Galanin: Inhibitory • Melanin-­concentrating hormone (MCH/GABA) • Adenosine • Acetylcholine (excitatory) • GABA (inhibitory) • Glycine (inhibitory) • MCH/GABA • GABA/galanin—Extended VLPO

From Chokroverty [3] with permission

The ascending projections of the ARAS consist of dorsal and ventral branches. The dorsal branch arises from the cholinergic laterodorsal tegmental (LDT) and pedunculopontine tegmental (PPT) nuclei and activates the cerebral cortex through its projection to the intralaminar and reticular nuclei of the thalamus and diffuse thalamocortical projections (prominent in the frontal cortex). There is also a glutamatergic projection from the parabrachial nucleus through basal forebrain (BF) taking part in this activation. The ventral branch arises from the noradrenergic locus coeruleus (LC), serotonergic dorsal raphe (DR), dopaminergic (DA), ventrolateral periaqueductal gray (VLPAG), ventral tegmental area (VTA), substantia nigra (SN), and histaminergic tuberomammillary nucleus (TMN) in the posterior hypothalamus as well as the orexin/hypocretin neurons in the lateral hypothalamus/perifornical region and cholinergic neurons in the nucleus basalis Meynert of the basal forebrain. It uses an extrathalamic route through the lateral and posterior hypothalamus to the basal forebrain and cerebral hemispheres. The descending projection of the ARAS to the lower brain stem and spinal cord via the tegmentoreticular (TRT) and

spinoreticular (SRT) tracts is responsible for motor control and maintenance of muscle tone. The other neuroanatomical structures besides the ARAS participating in maintaining wakefulness include the suprachiasmatic nucleus (SCM) of the hypothalamus and the orexin/hypocretin system. Besides its principal function of regulating circadian rhythm, the SCN regulates wakefulness through its anatomical connections (mainly indirect) to the wake-sleep-­promoting neurons, possibly due to the actions of corticotropin-­releasing factor (CRF) neurons in the hypothalamic paraventricular nucleus [7, 8]. Orexin A and B (also known as hypocretin 1 and 2 respectively) [9, 10] (Fig. 2) are neuropeptides exclusively produced in the lateral hypothalamus and are endogenous ligands for two orphan G ­protein-­coupled receptors. They were initially recognized as regulators of feeding behavior [11], but the subsequent discovery that orexin deficiency causes narcolepsy in humans and animals led to the discovery that they also play a critical role in regulating sleep/wake cycle [12]. Orexin A and B have extensive central nervous system (CNS) projections and activate the wake-active monoaminergic and cholinergic

Neuronal Circuits, Anatomical Substrates, and Mechanism of Sleep-Wakefulness Fig. 1  Ascending arousal projections (two branches) form reticular formation (RF), including the ascending reticular activating system, to activate the cerebral cortex (CX) and maintain vigilance, awareness and wakefulness. Th thalamus, BF basal forebrain, PH posterior hypothalamus, POA preoptic area, TM tuberomammillary nucleus, DR dorsal raphe, LDT laterodorsal tegmental nucleus, MES mesencephalon, CB cerebellum. (From Saper et al. [5]; with permission)

7

Cx

Th BF PH POA TM VTA

Mes DR RF LDT LC Pons

Cortical activation (W/REM):

Glu Ach

Cortical de-activation (SWS):

GABA

Behavioral arousal (W):

Glu

CB

RF Medulta

NA/DA

Ser HA Orx Behavioral quiescence (SWS/REM):

neurons in the hypothalamus/brain stem regions to maintain a long, consolidated awake period. They also project to the limbic system and might be important for increasing vigilance during emotional stimuli. The regulation of orexin production by substrates such as ghrelin, leptin, and glucose suggests that they might have important roles as a link between energy homeostasis and vigilance states [13]. Disruption of these wakefulness-promoting neurons will result in clinical sleep-wake disturbance causing either insomnia (a state of hyperarousal) or excessive daytime sleepiness. Animal models have clearly shown that stimulation of ARAS will activate the cerebral cortex producing desynchronized EEG and wakeful behavior, whereas lesion in the ARAS, particularly the portion in the midbrain, will result in sleep, stupor, or coma depending on the extent and severity of the lesion. The wake-promoting neurons including LDT, PPT, LC, DR, TM, or DA and orexin fire maximally during wakefulness and are dysfacilitated during sleep. The LDT and PPT also fire maximally during REM sleep, whereas LC, DR, TMs or DA fall silent in REM sleep and show weak firing intensity in NREM sleep.

GABA

Excitatory amino acids, glutamic and aspartic acids intermingle within the ARAS and are released maximally during wakefulness.

2.2 Neuroanatomical Substrates and Control of NREM Sleep 2.2.1 Hypothalamus NREM sleep-generating neurons are mainly located in the ventrolateral preoptic (VLPO) and the median preoptic (MnPO) nuclei of the anterior hypothalamus. Additionally, melanin-concentrating hormone (MCH)/gamma-­ aminobutyric acid-(GABA)ergic neurons located near wake-­ promoting orexin/hypocretin neurons in the lateral hypothalamus also contribute to a lesser extent in the generation of SWS. These hypnogenic neurons show increased firing rates at sleep onset as evidenced by intracellular recordings and increased expression of c-fos immunoreactivity as well as by optogenetic stimulation of MCH neurons. The VLPO and MnPO neurons both use the inhibitory neu-

8

S. Chokroverty et al.

Cerebral cortex Corpus callosum

Hippocampus

3V

Superior colliculus

Inferior colliculus

Midbrain Thalamus

Olfactory Bulb

4V

Cerebellum

Septum Flocculus

Central Gray Ferria Parabrachial nucleus

Olfactory Tubercle

Amyg LOT/PPT VLPO

Optic Chasm

Medulla oblongata

PVN SCN VMH

DR

LHA OREXIN

TMN Pons

ARC

Spinal cord

Pituitary

Fig. 2  Diffuse ascending and descending projections from lateral hypothalamic area (LHA) of the orexin/hypocretin throughout the central nervous system and the autonomic nervous system. DR dorsal raphe, VLPO ventrolateral preoptic nucleus, TMN tuberomammillary

nuclei, PVN paraventricular nucleus, VMH ventromedial hypothalamic nucleus, SCN suprachiasmatic nucleus, Amyg amygdala, (red arrow just before DR dorsal raphe: Locus coeruleus), ARC arcuate nucleus. (From Chokroverty and Provini [10]; with permission)

rotransmitter and neuromodulator GABA and galanin and actively participate in NREM sleep onset (MnPO) and maintenance as well as consolidation (VLPO).

2.2.3 Role of Thalamus The role of the thalamus in sleep-wake regulation and in the generation of NREM sleep was clearly shown in 1985 by Lugaresi and collaborators [15] after their clinicopathological observation in fatal familial insomnia (FFI) of degeneration of mediodorsal and anteroventral nuclei of the thalamus associated with severe insomnia. This observation was subsequently strengthened by electrophysiological documentation by Steriade and collaborators [16] of a significant role played by the thalamus in NREM sleep generation. Their study demonstrated that the thalamus is the first relay station at which reduction in afferent signals takes place at sleep onset. Their research also showed that sleep spindles are generated in the thalamic reticular nucleus, and delta oscillation of NREM sleep has two generators: one in the thalamus and the other in the cerebral cortex.

2.2.2 Brain Stem In the brain stem, the nucleus tractus solitarius (NTS) and parafacial zone (Pz), a node of sleep active rostral medullary neurons identified in rats, play a minor role in NREM sleep generation. The latter projects to the wake-promoting parabrachial nucleus (PBN) in the pons [14]. The medial PBN expresses c-fos positive immunoreactivity (i.e., active neurons) after sleep but not after wakefulness, and hence these are sleep active neurons. Pz lesion causes increased wakefulness but reduced SWS.  More than 50% of Pz sleep-active neurons use inhibitory transmitter/modulator GABA/ glycine.

Neuronal Circuits, Anatomical Substrates, and Mechanism of Sleep-Wakefulness

2.2.4 Role of Suprachiasmatic Nucleus The SCM located in the anterior hypothalamic region primarily controls the circadian rhythm but also plays a minor role in sleep-wake regulation as evidenced by its anatomical connection to both wake-promoting ARAS and sleep-­ promoting VLPO-MnPO neurons of the anterior hypothalamus [7, 8]. Therefore, SCM in addition to its role in controlling sleep timing (circadian), particularly during REM sleep and NREM-REM cycling, also participates in the homeostatic regulation (minor role) of NREM sleep state.

2.3 Mechanisms of Wake-Sleep Transitions The onset of spontaneous sleep represents the final outcome of a chain of disinhibition and inhibition resulting in deactivation of the brain stem ARAS [14]. The main stimulus for this mechanism includes decrement of tonic activity in the ARAS resulting from various factors (e.g., blocking of environmental sensory afferent stimuli and a cascade of dysfacilitation), which in turn releases suppression in many regions in the forebrain thus reinforcing disinhibition (i.e., excitation) of hypnogenic neurons. At sleep onset, there is an increase in recurrent inhibition at the thalamocortical levels. Relative immobility and closure of the eyelids block the activating sensory inflow and corticofugal projections provide excitation in the brain stem ARAS in case of functional or otherwise suppression of the ARAS tone [17]. At sleep onset VLPO and MnPO (“sleep-on” neurons) fire rapidly, thought to be due to a progressive accumulation of adenosine (assumed to be a physiological sleep factor) in the basal forebrain during prior wakefulness acting through adenosine A1 and A2A receptors [18] (other sleep factors such as prostaglandin D2, the cytokine interleukin- IB (IL-IB),

Fig. 3  Sleep switch (flip-flop switch) between the hypothalamus and the brainstem showing schematically mutual inhibition between sleep-­ promoting ventrolateral preoptic (VLPO), median preoptic (MnPO), and wake-promoting neurons [20] During sleep VLPO-MnPO neurons fire rapidly inhibiting wake-promoting neurons thus causing disinhibition and reinforcement of their own firing. In contrast, during wakefulness there is rapid firing of wake promoting neurons and inhibition of VLPO-MnPO neurons resulting in disinhibition of wake-on neurons.

9

growth hormone releasing factor (GHRF) and muramyl peptides may also play a role). The rapid firing of these “sleep­on” neurons inhibits brain stem and hypothalamic orexinergic and histaminergic waking neurons; at the same time there is reciprocal disinhibition of VLPO and MnPO neurons thus further exciting the “sleep-on” neurons (Fig.  3). Orexin/ hypocretin is thought to stabilize this “switch.” At wake onset, “wake-on” neurons fire rapidly and reciprocally inhibiting “sleep-on” neurons, and simultaneously reinforcing disinhibition of wave promoting neurons. Thus, interlinked excitation and inhibition result in a mutually self-reinforcing or self-sustaining mechanism. There is acceleration of one group of neurons (e.g., “sleep-on”) and deceleration of the other contrasting group (e.g., “wake-on”) and vice versa, fulfilling the principle of reciprocal interaction. Disruption of one end of the switch will destabilize the entire switch causing instability [19]. Similar activation, but to a smaller extent, also occurs at sleep onset of the NTS and Pz in the lower brain stem with reciprocal interaction of the upper brain stem ARAS independent of the hypothalamic brain stem switch described above. MCH neurons in the lateral hypothalamus next to orexin neurons play a minor role in NREM sleep but play a major role (master generator) in REM sleep and REM-­ NREM cycling.

2.4 Neuroanatomical Substrates and Control of REM Sleep 2.4.1 Role of the Pons The dorsal pons is the primary region for the control of REM sleep as clearly shown by Jouvet [20]. In his transection experiment of the brain stem of cats, transection at pontomesencephalic junction disclosed all the REM-generating

There is thus a reciprocal interaction like a “flip-flop” switch between two groups of antagonistic neurons. Orexin (hypocretin) is thought to stabilize the behavior of the switch; the switch will be destabilized if there is instability on either end DR Dorsal raphe, TM uberomammillary, LC locus coeruleus, DA dopaminergic, LDT/PPT laterodorsal tegmental/pedunculopontine tegmental. (From Chokroverty [3]; with permission)

10

neurons below the transection whereas section at pontomedullary junction revealed the REM-generating neurons above the cut. The crucial experiment was a transection at the pontomesencephalic and pontomedullary junctions leaving an isolated pons showing all the signs of REM sleep-generating neurons. The pons is thus shown to be necessary and sufficient for generating REM sleep.

2.4.2 Models for REM Sleep Generation The three models for REM sleep generation, derived from animal studies, are discussed below: 2.4.2.1 McCarley–Hobson Model The McCarley-Hobson model is a reciprocal interaction (or activation synthesis) model based on reciprocal interaction of “REM-on” cholinergic LDT-PPT neurons at pontomesencephalic junction and “REM-off” LC noradrenergic and DR serotonergic neurons. It is the earliest and most well-known model for REM generation and maintenance [21] (Fig. 4). The role of GABAergic neurons (both local and distal) in REM sleep generation has been emphasized in the last modification of the reciprocal interaction model by McCarley. 2.4.2.2 Lu and Co-workers from the Saper Group This model [22] introduced the concept of a “flip-flop” switch interaction in rats to explain REM sleep mechanism (Fig. 5) switching between GABAergic “REM-off” neurons in the deep mesencephalon, ventral periaqueductal gray (VLPAG) and lateral pontine tegmentum (LPT), and GABAergic “REM-on” neurons in the sublaterodorsal ­tegmental nucleus (SLD) in the dorsolateral pons (equivalent to peri-locus coeruleus alpha in the cat and subcoeruleus in humans). Of note, the term “flip-flop” switch may be a misnomer as sleep onset and offset occur more gradually than the fast “flip-flop” switch. The ascending glutamatergic projections from the dorsal extension of SLD, named precoeruleus neurons to medial septum are responsible for hippocampal theta and other EEG rhythms during REM Fig. 4 McCarley–Hobson REM sleep model of reciprocal interaction between cholinergic REM-on and aminergic REM-off neurons. LDT laterodorsal tegmental nucleus, PPT pedunculopontine tegmental nucleus, LC locus coeruleus, DR dorsal raphe, GABA gamma-aminobutyric acid. (Reproduced with permission from Reference [10])

S. Chokroverty et al.

sleep. REM muscle atonia is related to descending glutamatergic projections from ventral SLD directly to spinal interneurons without any apparent relay in the ventromedial medulla (VMM) inhibiting (hyperpolarizing) spinal ventral horn cells by both GABAergic and glycinergic mechanisms (inhibitory). Cholinergic and aminergic neurons do not form part of the “flip-flop” switch and play a modulatory role. Based on recent studies [23], investigators concluded in agreement with Luppi’s group [24] that glutamatergic neurons play a significant role in REM sleep generation. Ventral SLD descending glutamatergic neurons may cause muscle atonia by activating GABA/glycinergic premotor neurons in the VMM (indirect projection) as well as to the spinal cord interneurons (direct projections). VLPAG and the adjacent LPT in the reticular formation containing GABA are the “REM-off” neurons that are silent during REM sleep and exert tonic inhibitory control over glutamatergic “REM-on” or REM executive neuron in SLD during NREM sleep and wakefulness. These “REM-off” neurons thus “gate” the appearance of REM sleep. The extended VLPO region (GABAergic) projections to VLPAG and LPT (REM-off) to cause increased firing of the cells thus indirectly supporting REM sleep generation. MCH neurons have projections to both SLD and VLPAG and LPT for controlling REM sleep. The SLD dorsal precoeruleus glutamatergic neurons project to the basal forebrain through a relay in pontine para-brachial nucleus; thus, this pathway (septo-­hippocampal) may generate hippocampal theta and other REM EEG rhythms. 2.4.2.3 The Model Proposed by Luppi’s Group This model [25] (Fig.  6) proposes that at REM sleep onset there is an activation of “REM-on” glutamatergic neurons in the SLD. These REM-generating neurons would be inhibited (hyperpolarized) during wakefulness and NREM sleep by tonic GABAergic input from REM-off neurons located in the deep mesencephalic, VLPAG and LPT regions, and a few GABAergic neurons in the SLD as well as by active firing of monoaminergic “REM-off” neurons. Muscle atonia would be caused by both indirect and direct projections utilizing glycin-

Neuronal Circuits, Anatomical Substrates, and Mechanism of Sleep-Wakefulness

11

Fig. 5  “Flip-flop” switch original model of Saper’s group schematically shown to explain REM sleep mechanism. SLD sublaterodorsal nucleus, GABA gamma-aminobutyric acid, GLUT glutamatergic neurons, GLYC glycinergic neurons, PC precoeruleus, LDT laterodorsal tegmental nucleus, PPT pedunculopontine tegmental nucleus, LC locus coeruleus, DRN dorsal raphe nucleus. (From Chokroverty and Provini [10]; with permission)

Fig. 6  Luppi’s original model schematically shown to explain REM sleep mechanism in rats. GLUT glutamatergic neurons, GABA gamma-aminobutyric acid, LC locus coeruleus, DRN orsal raphe nucleus, GLYC glycinergic neurons, VLPO Ventrolateral preoptic nuclei of anterior hypothalamus, VLPAG ventrolateral periaqueductal grey region, DpMES deep mesencephalic reticular neurons, MCRF-­ PCRF magnocellular and parvocellular reticular formation, eVLPO hypothalamic extended VLPO nuclei, HYPO lateral hypothalamic orexin/ hypocretin. (From Chokroverty and Provini [10]; with permission)

ergic and GABAergic premotor neurons in the magnocellular and parvocellular reticular nuclei in the VMM as well as in the interneurons in the spinal cord and causing hyperpolarization of the motor neurons. Ascending dorsal SLD glutamatergic neurons are responsible for cerebral ­cortical activation and REM EEG pattern through projections to the thalamus and

thalamocortical pathways along with cholinergic neurons from LDT/PPT to the reticular thalamic nuclei and basal forebrain regions. In this model also, cholinergic neurons do not play a significant role but a modulating and permissible influence. In their latest experiment Luppi et  al. suggested that SLD neurons are mainly involved in REM muscle atonia only

12

S. Chokroverty et al.

but not for other components of REM sleep, thus creating certain contradictions in hypothesis for REM sleep regulation. It is important to note that the conclusion of Luppi and co-investigators regarding SLD glutamatergic neurons is based on solid experimental evidence. Following REM sleep deprivation studies in rats they noted that during REM recovery sleep, c-fos activated SLD neurons were not cholinergic (i.e., there was no increase of choline acetyltransferase, the enzyme synthesizing acetylcholine); also the enzyme synthesizing GABA, glutamic acid decarboxylase (GAD) did not increase in the majority of the SLD neurons, and hence most of them were not GABAergic. But most of the SLD neurons expressed vesicular glutamate transporter 2 (vGLUT2), a specific marker of glutamatergic neurons.

3 Mechanisms Underlying Essential Features of REM Sleep 3.1 REM Motor Atonia REM muscle atonia is believed to be mediated by inhibitory post-synaptic potentials generated by interneurons in the region of peri-LC alpha (ventral to LC) that are transmitted by the lateral tegmento reticular tract (TRT) to the VMM (the inhibitory zone of Magoun and Rhines) in and around the nucleus magnocellularis and gigantocellularis in the paramedianus. Reticulospinal tract (RST) arises from this region and projects to the spinal cord ventral horn cells causing hyperpolarization and muscle hypotonia or atonia (Fig.  7). Experimental lesions in the peri-LC alpha region and in the VMM producing REM sleep without atonia can be cited in support of this hypothesis [26–28].

Fig. 7  REM muscle atonia pathway schematically shown. LC locus coeruleus, Peri-LC perilocus coeruleus alpha, TRT tegmentoreticular tract, RST reticulospinal tract, GC gigantocellularis of medullary reticular formation, MC magnocellularis of medullary reticular formation. (From Chokroverty and Provini [10]; with permission)

3.2 REM Sleep EEG

3.3 Central Pattern Generators (CPGs) or the Mesencephalic and Spinal Locomotor Generators

This is manifested by desynchronized EEG with low amplitude mixed frequency waves (predominantly alpha and beta frequency with some theta rhythms) resembling waking EEG as well as sawtooth waves (2–7 Hz with characteristic morphological appearance seen best in frontocentral derivation) appearing just before phasic eye movements of REM sleep. The ascending glutamatergic projections from the dorsal extension of SLD, named precoeruleus neurons to medial septum, are responsible for hippocampal theta and other EEG rhythms during REM sleep, and other proposed mechanisms are discussed under various models described above.

The existence of CPGs is well known in animals but scantily described in humans [29]. These are inhibited during normal REM sleep but postulated to be activated during REM sleep behavior disorder (RBD), most likely in conjunction with activation of descending glutamatergic corticospinal and corticobulbar tracts to spinal ventral motor and cranial motor neurons, respectively, apparently bypassing the basal ganglia. These projections from activated CPGs and central motor cortex appear to be responsible for generating abnormal, mostly violent movements of RBD. The mechanism of inhibition and activation of CPGs remains to be documented.

Neuronal Circuits, Anatomical Substrates, and Mechanism of Sleep-Wakefulness

References 1. Iber C, American Academy of Sleep Medicine, et al. The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications. Westchester: American Academy of Sleep Medicine; 2007. 2. Lugaresi E, Provini F, Montagna P.  The neuroanatomy of sleep. Considerations on the role of the thalamus in sleep and a proposal for a caudorostral organization. Eur J Anat. 2004;8(2):85–93. 3. Chokroverty S. Functional neuroanatomy and mechanism of sleep. In: Chokroverty S, Cortelli P, editors. Autonomic nervous system and sleep. Order and disorder. Cham: Springer Nature Switzerland AG; 2021. 4. Critchley M. The pre-dormitum. Rev Neurol (Paris). 1955;93:101–6. 5. Saper CB, Scammell TE, Lu J. Hypothalamic regulation of sleep and circadian rhythms. Nature. 2005;437(7063):1257–63. 6. Brodal A. The reticular formation of the brain stem: anatomical and functional considerations. Edinburgh: Oliver and Boyd; 1957. 7. Moore RY.  The suprachiasmatic nucleus and sleep-wake regulation. Postgrad Med. 2004;116(6 Suppl Primary):6–9. 8. Ono D, Mukai Y, Hung CJ, Chowdhury S, Sugiyama T, Yamanaka A. The mammalian circadian pacemaker regulates wakefulness via CRF neurons in the paraventricular nucleus of the hypothalamus. Sci Adv. 2020;6(45):eabd0384. 9. Ohno K, Sakurai T.  Orexin neuronal circuitry: role in the regulation of sleep and wakefulness. Front Neuroendocrinol. 2008;29(1):70–87. 10. Chokroverty S, Provini F. Sleep, breathing, and neurologic disorders. In: Chokroverty S, editor. Sleep disorders medicine: basic science, technical considerations and clinical aspects. 4th ed. New York: Springer; 2017. 11. Willie JT, Chemelli RM, Sinton CM, Yanagisawa M. To eat or to sleep? Orexin in the regulation of feeding and wakefulness. Annu Rev Neurosci. 2001;24:429–58. 12. De la Herrán-Arita AK, Guerra-Crespo M, Drucker-Colín R.  Narcolepsy and orexins: an example of progress in sleep research. Front Neurol. 2011;18(2):26. 13. Sakurai T. Orexin: a link between energy homeostasis and adaptive behaviour. Curr Opin Clin Nutr Metab Care. 2003;6(4):353–60. 14. Jones BE.  Neuroanatomical, neurochemical, and neurophysical bases of waking and sleeping. In: Chokroverty S, Ferini-Strambi L, editors. Oxford textbook of sleep disorders. Oxford: Oxford University Press; 2017. p. 23–32. 15. Lugaresi E, Medori R, Montagna P, Baruzzi A, Cortelli P, Lugaresi A, et  al. Fatal familial insomnia and dysautonomia

13

with selective degeneration of thalamic nuclei. N Engl J Med. 1986;315(16):997–1003. 16. Steriade M. Neurophysiologic mechanisms of non-rapid eye movement (resting sleep). In: Chokroverty S, editor. Sleep disorders medicine. 2nd ed. Philadelphia, Pa: Elsevier/Butterworth; 1999. p. 51–62. 17. Livingston RB, Hernandez-Peon R, French JD. Corticofugal projections to brain stem activating system. Fed Proc. 1953;12:89–90. 18. Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW.  Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science. 1997;276(5316):1265–8. 19. Saper CB, Chou T, Scammell TE.  The sleep switch: hypothalamic control of sleep and wakefulness. Trends Neurosci. 2001;24(12):726–31. 20. Jouvet M. Recherches sur les structures nerveuses et les mécanismes responsible des différentes phases du sommeil physiologique. Arch Ital Biol. 1962;100:125–206. 21. McCarley RW.  Neurobiology of REM and NREM sleep. Sleep Med. 2007;8(4):302–30. 22. Lu J, Sherman D, Devor M, Saper CB. A putative flip-flop switch for control of REM sleep. Nature. 2006;441(7093):589–94. 23. Vetrivelan R, Lu J. Neural circuitry regulating REM sleep and its implication in REM sleep behavior disorder. In: Schenck C, Hogl B, Videnovic A, editors. Rapid eye movement sleep behavior disorder. Cham: Springer Nature; 2019. p. 559–77. 24. Luppi PH, Fort P. Neurochemistry of sleep: an overview of animal experimental work. In: Aminoff MJ, Boller F, Swaab DF, series editors. Handbook of clinical neurology, Vol. 98, 3rd series. Montagna P, Chokroverty, volume editors. Sleep disorders part I.  Elsevier: Amsterdam; 2011. p. 173–90. 25. Luppi PH, Clement O, Peyton C, Fort P.  Neurobiology of REM sleep. In: Chokroverty S, Ferini-Strambi L, editors. Oxford textbook of sleep disorders. Oxford: Oxford University Press; 2017. p. 15–22. 26. Lai YY, Siegel JM. Medullary regions mediating atonia. J Neurosci. 1988;8:4790–6. 27. Jouvet M, Michel F.  Corrélations électromyographiques du sommeil chez le chat décortiqué et mésencéphalique chronique. CR Soc Biol (Paris). 1959;153:422–5. 28. Berger RJ. Tonus of extrinsic laryngeal muscles during sleep and dreaming. Science. 1961;134(3482):840. 29. Tassinari CA, Rubboli G, Gardella E, Cantalupo G, Calandra-­ Buonaura G, Vedovello M, et  al. Central pattern generators for a common semiology in fronto-limbic seizures and in parasomnias. A neuroethologic approach. Neurol Sci. 2005;26(Suppl 3):S225–32.

Part II Laboratory Techniques

Polysomnographic Recording Technique Sudhansu Chokroverty and Sushanth Bhat

1 Introduction After a detailed history and physical examination, the diagnostic polysomnogram (PSG) plays a pivotal role in confirming the clinician’s suspicions of an underlying sleep disorder and helps guide further management. A routine overnight PSG records multiple physiological characteristics simultaneously during one night’s sleep [1–3]. Electroencephalography (EEG), electrooculography (EOG), and chin muscle electromyography (EMG) channels are used to stage sleep. Airflow and respiratory effort channels are used to score sleep disordered breathing. The finger pulse oximetry channel provides additional data in this regard, as well as being helpful in identifying sleep hypoxemia independent of apneic and hypopneic events. In patients undergoing continuous positive airway pressure (CPAP) titration for obstructive sleep apnea (OSA), the C-flow channel provides the airflow signal, and CPAP pressure is continuously adjusted during the night to eliminate respiratory events. Limb EMG channels are typically placed on the legs (usually the tibialis anterior muscle) and aid in the scoring and evaluation of limb movements. Additional limb EMG channels may be used in special montages (see section on “Electromyography” below). A single channel electrocardiography (EKG) channel and a snore channel are part of the typical PSG setup. Video and audio recording are essential

S. Chokroverty HMH JFK University Medical Center and Neuroscience Institute, Hackensack Meridian School of Medicine at Seton Hall University, South Orange and Nutley, NJ, USA Rutgers Robert Wood Johnson University Medical Center, New Brunswick, NJ, USA S. Bhat (*) Department of Neurology, Hackensack Meridian School of Medicine, Nutley, NJ, USA JFK Neuroscience Institute Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA

for recording position and evaluating abnormal movements and behavior in sleep (such as bruxism, catathrenia and various other parasomnias). Special techniques, not part of routine recording in most laboratories but used in selected patients, include measurements of intraesophageal pressure in patients with suspected upper airway resistance syndrome (UARS), which helps distinguish obstructive and central apneas, esophageal pH in patients with nocturnal gastroesophageal reflux disease (GERD), and penile tumescence for assessment of patient with erectile dysfunction. These are described in greater detail in the sections below.

2 Patient Preparation and Laboratory Environment Attempts should be made to recreate a typical night’s sleep for the patient so that recorded parameters are most clinically relevant. Within reason, lights out and lights on times (the beginning and the ending of the recording respectively) should match the patient’s regular bedtime and waking times, so as to prevent falsely shortened or prolonged sleep onset and rapid eye movement (REM) sleep latencies. Caffeine and smoking should be avoided the evening of the sleep study. The patient should be advised to take their usual medications, although those medications that may affect sleep or muscle tone (benzodiazepines, over the counter and prescriptions sleep aids, medications that may suppress REM) should be noted. A low light level camera should be used to obtain good quality video in the dark, and an infrared light source should be available after turning the laboratory lights off. The camera should be mounted on the wall across from the head end of the bed. The monitoring station should have remote control that can zoom, tilt, or pan the camera for adequate viewing, and technologists should be alert to any abnormal movements and adjust the view accordingly. An intercom from a microphone near the patient should be available.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. J. Thomas et al. (eds.), Atlas of Sleep Medicine, https://doi.org/10.1007/978-3-031-34625-5_3

17

18

3 PSG Calibration 3.1 Equipment Calibration Before recording of the PSG is begun, the technologist must perform an all-channel calibration by sending a known signal through all the amplifiers, which are set to the same lowand high-frequency filter settings and sensitivities (usually as set for EEG channel recording), thus testing proper ­functioning of all amplifiers. Appropriate filter settings and sensitivities are then set for each channel recording the various physiological characteristics, documenting individual channel calibration.

3.2 Physiologic Calibration Once equipment calibration is completed, physiological calibration is done. The patient is given a series of instructions chosen so that a comparison is available when similar events, important for staging and scoring, occur in sleep. • • • • • •

Open eyes and look straight for 30 s Close eyes for 30 s Look left, right, up, and down Blink eyes five times Clench teeth Inhale and exhale (noting the polarity of excursion; it is best to keep inhalation upward and exhalation downward for consistency) • Hold breath for 10 s • Extend right hand and then left hand • Dorsiflex right foot and then left foot “Lights out” follows the completion of calibration and recording for scoring begins only then. It is not appropriate to score the patient as being awake at the beginning of the study when they are lying in bed watching TV, using electronic devices or reading rather than attempting to fall asleep as this will lead to a falsely prolonged sleep onset latency.

4 Technologist Education Successful completion of an in-laboratory PSG requires skilled and experienced technologists who are trained to deal with unusual circumstances that may arise. • Technologists should be trained to identify behavior suggesting seizures, postictal confusion, tongue biting, transient paralysis or other neurological dysfunction. • Technologists should be instructed to ensure that the camera (during a video-PSG) is focused on the patient during

S. Chokroverty and S. Bhat

unusual behaviors or movements suspected to be seizures, parasomnias or dream enactment. • In all patients in whom there is suspicion for REM sleep behavior disorder (RBD)or in whom complex movements suggesting dream enactment is noted, technologists must ask the patient about dream recall and document this in the chart. Technologists should also be educated about safety precautions and laboratory protocol in dealing with emergency situations that may arise during the night. These include: • Cardiopulmonary arrest • Seizures during PSG recording • Safety and precautions against injury in appropriate cases.

5 Technical Considerations and Polysomnography Equipment Biological signals recorded during PSG are of very small amplitude (EEG, EOG and EMG activity is in the microvolt range) and need to be recorded and amplified, then passed it through adjustable filters for display at different sensitivity settings. PSG equipment uses differential amplifiers, which amplify the potential difference between the two amplifier inputs while suppressing extraneous signals that are present at both electrodes (e.g., 60 Hz artifact). This ability is measured by common mode rejection. The common mode rejection ratio (CMRR) is defined as the ability of the amplifier to reject in-phase (i.e., common to both inputs) and amplify out-of-phase signals and is expressed as (differential mode gain/common mode gain). CMRR should ideally exceed 1000 to 1, but most contemporary PSG amplifiers use a ratio in excess of 10,000 to 1. The higher the ratio, the better the ability of the amplifier to reject unwanted signals and provide clear and accurate recordings. • The amplifiers used consist of both alternating current (AC) and direct current (DC) amplifiers. • The AC amplifiers are used to record physiological characteristics showing high frequencies such as EEG, EOG, EMG, and EKG.  The AC amplifier contains both highand low-frequency filters. • DC amplifiers have no low-frequency filters and are typically used to record potentials with slow frequency such as the output from the oximeter, the output from the pH meter, CPAP titration pressure changes, and intraesophageal pressure readings. • AC or DC amplifiers may be used to record respiratory flow and effort.

Polysomnographic Recording Technique

19

The following PSG filter settings need to be defined: • High frequency (or low pass) filters attenuate all activity at frequencies higher than the value at which they are set, while allowing lower frequency activity to pass. For example, EEG channels are typically set with a high frequency filter at 70-Hz. Lowering the setting to 35 Hz may filter out high-frequency artifact (such as muscle activity) but allows lower frequency activity to pass through unchanged. • Low frequency (or high pass) filters, on the other hand, attenuate all activity lower than the value at which they are set, while allowing faster frequency activity to pass. EEG channels are typically set at 0.3-Hz. Raising the setting to 1 Hz may filter out low-frequency artifact (such as sweating) but allows higher frequency activity to pass through unchanged. • The 60-Hz notch filter attenuates electrical artifact which has a frequency of 60-Hz artifact (the frequency in other countries is usually 50-Hz), generally seen if the electrode application and impedance are suboptimal. Sensitivity is expressed in microvolts per millimeter or millivolts per centimeter. Sensitivity switches should be adjusted to obtain sufficient amplitude for interpretation. Sensitivity and filter settings vary according to the physiological characteristics recorded (Table 1). The standard speed for recording traditional PSG is 10 mm/s so that each monitor screen is a 30-s epoch, making sleep staging easiest. EEG abnormalities can be better analyzed by slowing the recording down to 30  mm/s (10-s epoch). Respiratory events may be best visualized with a 5 mm/sec speed (60-s epoch).

5.1 Electroencephalography The main purpose of EEG recording performed during PSGs is to distinguish between wakefulness and various stages of sleep. The American Academy of Sleep Medicine (AASM) Manual for the Scoring of Sleep and Associated Events [4] recommends a minimum of three channels (F4-M1, C4-M1, O2-M1) representing the right frontal, central, and occipital electrodes referenced to the contralateral mastoid electrode, with corresponding backup electrodes over the left hemisphere (F3-M2, C3-M2, O1-M2) referenced to the contralat-

eral mastoid electrode, in case of malfunction of the primary electrodes. However, the absence of a temporal lead may result in missing epileptiform activity which is most common in this region. The AASM also recommends an alternative derivation that includes the midline and a central channel (Fz-Cz, Cz-Oz, C4-M1); this montage may result in midline, centrally predominant activity such as sleep spindles, K-complexes and slow wave sleep to be attenuated and easily missed. This would pose a particular problem in elderly patients while scoring Stage N3 sleep, where the slow wave activity must meet particular amplitude criteria. Therefore, we recommend a montage that records over both hemispheres and includes the temporal regions (Table 2), in addition to electrodes recommended by the AASM for the scoring of sleep. For patients in whom nocturnal seizures are suspected or likely to occur, a full seizure montage with parasagittal and temporal chains is recommended (Table 3).

5.2 Electrooculography EOG recording is crucial to staging sleep accurately. The two recommended electrodes are labeled E1 (1 cm below the left outer canthus) and E2 (placed 1 cm above the right outer canthus) both referenced to the right mastoid, this allows simultaneous recording of both vertical eye movements (such as blinking) and horizontal eye movements (both slow and rapid). Gold cup or silver-silver chloride electrodes can be used to monitor the EOG. The underlying concept is that the eye is an electric dipole, with relative positivity at the cornea and a relative negativity at the retina. Any eye movement changes the orientation of the dipole, and it is the movement of the dipole that is recorded as a potential difference between the two electrodes used to record the EOG. In this arrangement, conjugate eye movements produce out-of-phase deflections in the two channels, whereas the EEG slow activities contaminating the eye electrodes are in-phase. The sensitivity and filter settings for EOG are similar to those used for EEG (Table 1). Eye movements are generally characteristic of the sleep stage in which they occur and are an essential part of scoring. Eye blinks, seen only in wakefulness, are conjugate vertical eye movements occurring at 0.5–2 Hz with the eyes open or closed. Rapid eye movements (conjugate, irregular, sharp eye movements with an initial deflection of less than half a

Table 1  Filter and sensitivity settings High frequency characteristics Electroencephalogram Electro-oculogram Electromyogram Electrocardiogram

High frequency filter (Hz) 70 or 35 70 or 35 90 15

Time constant (s) 0.4 0.4 0.04 0.12

Low frequency filter (Hz) 0.3 0.3 5.0 1.0

Airflow and effort

15

1

0.1

Sensitivity 5–7 μV/mm 5–7 μV/mm 2–3 μV/mm 1 mV/cm to start; adjust 5–7 μV/mm; adjust

S. Chokroverty and S. Bhat

20 Table 2  Typical overnight polysomnographic montage used in our laboratory 1. F3-M2 2. C3-M2 3. T3-M2 4. O1-M2 5. F4-M1 6. C4-M1 7. T4-M1 8. O2-M1 9. Left electro-oculogram (E1-M2) 10. Right electro-oculogram (E2-M2) 11. Chin electromyogram (EMG) 12. EKG 13. Heart rate 14. Left gastrocnemius EMG 15. Left tibialis anterior EMG 16. Right gastrocnemius EMG 17. Right tibialis anterior EMG 18. Intercostal EMG 19. Oronasal thermistora 20. Nasal pressure transducera 21. Chest 22. Abdomen 23. Snoring 24. Arterial oxygen saturation Channels 1–8 record electroencephalogram activity from bilateral cerebral hemispheres in a referential chain; electrode designation per the 10–20 International System of electrode placement. M1 and M2, left and right mastoid, respectively. Channel 19 and 20 record airflow (“flow channels”). Channels 21 and 22 record respiratory effort (“effort channels”) a  In a CPAP titration study, flow channels are replaced by a CPAP signal (C-flow signal)

second) occur in wakefulness (wake eye movements, WEMs), along with high chin EMG tone, eye blinks, and a posterior dominant rhythm, but also occur in REM sleep, especially in phasic REM where they occur in bursts seen in all directions (horizontal, oblique and vertical) and are accompanied by low to absent chin tone (interspersed with transient phasic bursting) and a desynchronized, amorphous EEG pattern. In REM sleep, rapid eye movements (REMs) are frequently preceded by sawtooth waves, although both may occur independently. REM density is defined as the number of phasic eye movements per minute in REM sleep. It typically increases in later REM cycles during the course of a normal PSG; this may be reversed in patients with depression; this pattern of evolution may not be true in REM sleep behavior disorder (RBD). Slow lateral eye movements (SLEMs) are seen in drowsiness and light sleep and are defined as conjugate, sinusoidal, regular eye movements with an initial deflection of greater than half a second (Fig. 1). These eye movements cannot be volitionally simulated. In patients who do not generate a posterior dominant rhythm, their appearance heralds Stage N1 sleep. While they may persist into Stage N2 during the early

Table 3  Extended EEG (“seizure”) montage 1. F4-M1 2. C4-M1 3. O2-M1 4. C3-M2 5. Fp1-F7 6. F7-T3 7. T3-T5 8. T5-O1 9. Fp2-F8 10. F8-T4 11. T4-T6 12. T6-O2 13. Fp1-F3 14. F3-C3 15. C3-P3 16. P3-O1 17. Fp2-F4 18. F4-C4 19. C4-P4 20. P4-O2 21. Left electro-oculogram (E1-M2) 22. Right electro-oculogram (E2-M2) 23. Chin electromyogram (EMG) 24. Right masseter EMG 25. Left biceps EMG 26. Right biceps EMG 27. Left tibialis anterior EMG 28. Right tibialis anterior EMG 29. Intercostal EMG 30. Oronasal thermistora 31. Nasal pressure transducera 32. Chest 33. Abdomen 34. Snoring 35. Arterial oxygen saturation 36. EKG 37. Heart rate Channels 1–20 record electroencephalogram activity from bilateral cerebral hemispheres with referential and bipolar montages including both temporal and parasagittal chains; electrode designation per the 10–20 International System of electrode placement. Channel 30 and 31 record airflow (“flow channels”). Channels 32 and 33 record respiratory effort (“effort channels”) a  In a CPAP titration study, flow channels are replaced by a CPAP signal (C-flow signal)

part of the night, they generally disappear in Stage N3 and REM sleep. However, patients on antidepressants such as selective serotonin reuptake inhibitors (SSRIs like fluoxetine and paroxetine), as well as serotonin-norepinephrine reuptake inhibitors (SNRI like duloxetine), may have unusual eye movements that appear to be a mixture of rapid and slow eye movements occurring well into Stage N3 and often into REM sleep (colloquially referred to among polysomnographers as “Prozac eyes”); their presence makes sleep staging difficult (Fig. 2) and can render scoring of a multiple sleep latency testing (MSLT) equally frustrating.

Polysomnographic Recording Technique

21

Fig. 1  A 30-s epoch from the overnight polysomnogram of a 46-year-­ old man with a longstanding history of snoring, witnessed apneas and excessive daytime sleepiness. Note the occurrence of slow lateral eye movements in the electrooculogram (EOG) channels, E1-M1 and E2-M1. A posterior dominant rhythm is still present in more than half this epoch, which is therefore still scored as Stage W. Top eight channels; EEG recording with electrodes placed according to the 10–20

international electrode placement system. Chin1-Chin2 submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. INTRC; intercostal EMG. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. OroNs1-OroNS2; oronasal airflow. Pflw1-Pflw2; nasal pressure transducer recording. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

Fig. 2  A 60-s epoch from the overnight polysomnogram of a 35-year-­ old man complaining of excessive daytime sleepiness and with a prior diagnosis of obstructive sleep apnea. He also has a history of depression and is on citalopram, a selective serotonin reuptake inhibitor (SSRI). This epoch represents Stage N2 sleep, as evidenced by the presence of K-complexes and sleep spindles. Note the presence of excessive eye movements, representing a combination of slow and rapid eye movements. Eye movements generally do not persist into stage N2 and beyond but are often seen in patients on SSRIs (colloquially referred to

as “Prozac eyes”). Top eight channels; EEG recording with electrodes placed according to the 10–20 international electrode placement system. Chin1-Chin2; submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. OroNs1-­ OroNS2; oronasal airflow. Pflw1-Pflw2; nasal pressure transducer recording. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

S. Chokroverty and S. Bhat

22

5.3 Electromyography Surface EMG is recorded using gold cup or silver-silver chloride electrodes applied to a clean surface using a tape or electrode glue. PSG consists of chin EMG channels recording activity from the mentalis and submental muscles (the mylohyoid and anterior belly of the digastric) and bilateral leg EMG channels recording activity from the tibialis anterior muscles. For chin EMG recordings, at least three EMG electrodes are applied so that in the event of a problem with one of the electrodes the additional electrode serves as a backup. The electrode impedance should be less than 5 K. The high- and low-frequency filter settings for the EMG recordings are different from those used for EEG and EOG and are listed in Table 1. The sensitivity should be at least 20 microvolts per millimeter for mental or submental EMG activity. Lower limb EMGs are generally recorded with electrodes placed over the tibialis anterior muscles 2–2.5  cm apart. The main utility of these channels is to record limb movements in patients with periodic limb movements in sleep (PLMS). Many patients with a history of abnormal movements or behavior in sleep require a more extended EMG montage, known as a multiple muscle montage, which includes extra channels that record from additional cranially innervated muscles (such as the sternocleidomastoideus, masseter and orbicularis oris), upper limb muscles (e.g., biceps, triceps, extensor digitorum communis, flexor digitorum subliminus and flexor digitorum profundus), lower limb muscles (e.g., quadriceps, gastrocnemius) and axial muscles (e.g., paraspinals, rectus abdominis, intercostals) (Table 4). This is of particular utility in patients with suspected RBD where REM without atonia (RWA) may be missed if an adequate number of muscles is not sampled. It is often helpful to also include intercostal and diaphragmatic EMG channels to record respiratory muscle activity. The intercostal EMG recorded from the seventh to ninth intercostal space with active electrodes on the anterior axillary line and the reference electrodes on the midaxillary or posterior axillary line may also include some diaphragmatic muscle activity in addition to the intercostal activity. Diaphragmatic activity can be recorded by placing surface electrodes over the right or left side of the umbilicus or over the anterior costal margin, but these are contaminated by a mixture of intercostal and rectus abdominis muscle activity, and such noninvasive techniques are unreliable for quantitative assessment of diaphragmatic EMG. True diaphragmatic activity is typically recorded by intraesophageal recording. Intercostal and diaphragmatic EMG is particularly useful in the differentiation between central and obstructive apneas, especially when the respiratory channels are unreliable; continued bursts of activity in these channels during such an

Table 4  Multiple muscle montage (extended EMG channels for parasomnias and other movement disorders in sleep) 1. F3-M2 2. C3-M2 3. O1-M2 4. F4-M1 5. C4-M1 6. O2-M1 7. Left electro-oculogram (E1-M2) 8. Right electro-oculogram (E2-M2) 9. Chin electromyogram (EMG) 10. Right masseter EMG 11. Right sternomastoid EMG 12. Left biceps brachii EMG 13. Left triceps EMG 14. Right biceps brachii EMG 15. Right triceps EMG 16. Right lower rectus abdominis EMG 17. Right lumbar paraspinals EMG 18. Left quadriceps EMG 19. Left hamstrings EMG 20. Right quadriceps EMG 21. Right hamstrings EMG 22. Left gastrocnemius EMG 23. Left tibialis anterior EMG 24. Right gastrocnemius EMG 25. Right tibialis anterior EMG 26. Intercostal EMG 27. Oronasal thermistora 28. Nasal pressure transducera 29. Chest 30. Abdomen 31. Snoring 32. Arterial oxygen saturation 33. EKG 34. Heart rate Channels 1–6 record electroencephalogram activity from bilateral cerebral hemispheres with referential parasagittal chains. Channels 10–25 record EMG activity, including from additional muscles not routinely studied on typical PSGs. Channels 27 and 28 record airflow (“flow channels”). Channels 29 and 30 record respiratory effort (“effort channels”). Also see Figs. 10 and 11 a  In a CPAP titration study, flow channels are replaced by a CPAP signal (C-flow signal)

event would identify it as obstructive, while the absence of such bursts would implicate a central event (Fig. 3). In our laboratory, we have designed a hybrid seizure/multiple muscle montage for recording patients who have abnormal behaviors and events in sleep and in whom the differential diagnosis includes seizures and parasomnias. While most PSG machines have a limited number of inputs, precluding full seizure and multiple muscle recordings during the same study, we have found this hybrid montage useful in analyzing such events (Table 5). Periorbital integrated potentials (PIPs) [5], one of the cardinal features of REM sleep, are noted during phasic bursts

Polysomnographic Recording Technique

23

Fig. 3  A representative 90-s epoch from the overnight polysomnogram (PSG) of a 46-year-old man with a longstanding history of snoring, witnessed apneas and excessive daytime sleepiness. Three respiratory events are depicted in this epoch. Note that during the second event (b), the chest lead shows artifacts making it difficult to determine whether it is central or obstructive in nature. However, the intercostal EMG channel (INTRC) shows the absence of inspiratory bursts during the event, confirming that it is a central apnea. Also noted is ballistocardiographic artifact in the abdomen (ABD) lead, another indicator of a central event. For comparison, the first and third events (a and c) are clearly obstructive apneas, and intercostal EMG bursts continue (albeit at a lower

amplitude) throughout their duration. The intercostal EMG channel is often very helpful in distinguishing between obstructive and central apneas when effort channels are unreliable or show artifacts. Top eight channels; EEG recording with electrodes placed according to the 10–20 international electrode placement system. Chin1-Chin2; submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. OroNs1-OroNS2; oronasal airflow. Pflw1-Pflw2; nasal pressure transducer recording. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

of REMs in addition to desynchronized EEG, chin muscle atonia/hypotonia and transient muscle bursts in EMG channels (Fig. 4). PIP amplitude (20–100 μv) and duration (15– 100 ms) show considerable variation; they may occasionally be seen in wakefulness and NREM sleep where these are usually isolated and not seen in clusters. Simultaneous occurrence of PIPs and REMs may suggest a common central phasic motor system for their generation similar to the suggestion made for the simultaneous occurrence of ponto-geniculo-­ occipital (PGO) waves, REMs and transient phasic (myoclonic) muscle bursts. The precise neural substrates, functional significance, and relationship with PGO waves (recorded predominantly in rats and cats, rarely in humans), however, remain uncertain.

many patients with OSA.  Gold cup surface electrodes are used to record the EKG, and Table 1 lists the filter settings and sensitivities for such recording. The PSG report should mention cardiac rhythm abnormalities seen during the night. Data from a single channel EKG is limited; hence, abnormalities often need to be followed up with a full 12 lead EKG.

5.4 Electrocardiography The PSG generally includes a single channel of EKG recorded by placing one electrode over the sternum and the other electrode at a lateral chest location. This recording detects brady tachyarrhythmias or other arrhythmias seen in

5.5 Recording of Respiratory Effort The AASM now recommends respiratory inductive plethysmography (RIP) belts (calibrated or uncalibrated) or esophageal manometry or for measurement of respiratory effort. Polyvinylidene fluoride (PVDF) belts are acceptable alternatives.

5.5.1 Respiratory Inductive Plethysmography RIP belts are the mostly commonly used method of measuring respiratory effort. Inductance refers to resistance to current flow. RIP belts contain a loop wire whose cross-sectional area changes with alteration in thoracoabdominal volumes as

24 Table 5  Hybrid EEG/parasomnia montage 1. F3-M2 2. C3-M2 3. O1-M2 4. T3-T5 5. T5-O1 6. T4-T6 7. T6-O2 8. C4-P4 9. Left electro-oculogram (E1-M2) 10. Right electro-oculogram (E2-M2) 11. Chin electromyogram (EMG) 12. Right masseter EMG 13. Right sternocleidomastoideus EMG 14. Right biceps brachii EMG 15. Left biceps brachii EMG 16. Right lower rectus abdominis EMG 17. Right lumbar paraspinals EMG 18. Right quadriceps EMG 19. Left quadriceps EMG 20. Right tibialis EMG 21. Right gastrocnemius EMG 22. Left tibialis EMG 23. Left gastrocnemius EMG 24. Intercostal EMG 25. Oronasal thermistora 26. Nasal pressure transducera 27. Chest 28. Abdomen 29. Snoring 30. Arterial oxygen saturation 31. EKG 32. Heart rate Channels 1–8 record electroencephalogram activity from bilateral cerebral hemispheres in a bipolar montage. Channels 12–23 record EMG activity, including from some additional muscles not routinely studied on typical PSGs. Channels 25 and 26 record airflow (“flow channels”). Channels 27 and 28 record respiratory effort (“effort channels”) a  In a CPAP titration study, flow channels are replaced by a CPAP signal (C-flow signal)

occurs during normal inspiration and expiration. Signals from RIP belts are amplified for display on PSG montages using constant AC current. Electrical changes in the coiled wires are linearly proportional to changes in the cross-­ sectional areas occurring during breathing. RIP belts can therefore be calibrated to tidal volume or (more commonly) may be uncalibrated. RIP belts are prone to dislodgement during the night by patient movement, causing inaccuracy in measurements of the respiratory effort. The sum of the two RIP belts (thorax and abdomen) is known as the RIPsum, and the time derivative of RIPsum is known as the RIPflow. When RIP belts are calibrated, the RIPsum is a measure of tidal volume and RIPflow a measure of airflow; in uncalibrated RIP, deflections in the RIPsum and RIPflow signals still reflect a relative change in tidal volume and airflow, respectively, compared to baseline breath-

S. Chokroverty and S. Bhat

ing. RIPsum and RIPflow signals may reflect thoracoabdominal asymmetry which is commonly seen in upper airway resistance, being diminished during out-of-­ phase thoracoabdominal excursions and markedly attenuated or even absent during complete paradoxical breathing. Thus, during an obstructive apnea, the RIPsum and RIPflow signals show absent or minimal excursions (and always show absent excursions during a central apnea due to complete lack of respiratory effort), and during a hypopnea, the excursions are diminished compared to baseline breathing. For diagnostic studies, the AASM recommends using RIPsum and RIPflow channels as alternative methods of detecting hypopneas and apneas, and dual thoracoabdominal belts as alternate sensors for detection of hypopneas (but not apneas) in both children and adults.

5.5.2 Intraesophageal Pressure Monitoring While considered the gold standard for measuring respiratory effort, intraesophageal pressure monitoring (or esophageal manometry) is semi-invasive, requiring the placement of a nasogastric balloon-tipped catheter into the distal esophagus, which registers pleural pressure changes. Hence, it is not routinely employed. • In normal wakefulness, the pleural pressure change is less than 5 cm of water and in sleep it is between 5 and 10 cm of water. • Inspiration causes more negative pleural (and hence esophageal) pressure than expiration. • Intraesophageal pressure monitoring is used to detect respiratory event-related arousals (RERAs), (seen in upper airway resistance syndrome [UARS]) defined as a series of increasingly negative inspiratory pressures culminating in an arousal and return to normal pressures, without preceding apneas or hypopneas. • Intraesophageal pressure monitoring may also help distinguish between central and obstructive apneas and hypopneas when routine airflow and effort channels are unreliable.

5.5.3 Polyvinylidene Fluoride Belts PVDF belts are peizoelectric strain gauges placed around the chest and abdomen consisting of a fluoropolymer film that emits an electrical signal in response to changes in force, acceleration, pressure, or temperature. Signals produced by PVDF belts are converted to PSG-displayable signals via self-generated DC current. In contrast to RIP, PVDF belts use impedance to measure changes in electrical resistance, which are not linearly related to changes in cross-sectional dimensions. Thus, unlike RIP belts, PVDF belts cannot be calibrated to measure tidal volumes. The sum of the dual PVDF belts provides a PVDFsum, analogous to the RIPsum.

Polysomnographic Recording Technique

25

Fig. 4  A 30-s epoch of REM sleep from the overnight PSG of a 54-year-old man with OSA diagnosis. Note biphasic clustered periorbital integrated potentials (PIPs) (boxed) in the left (Lt) orbicularis (O) oculi  – E2 channel, occurring with phasic rapid eye movements (REMs). Not all REMs are associated with PIPs (oval), and PIPs do not occur during the segment without REMs (diamond). Top eight channels; EEG recording with electrodes placed according to the 10–20 international electrode placement system. Chin1-Chin2; submental

electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. OroNs1-OroNS2; oronasal airflow. Pflw1-Pflw2; nasal pressure transducer recording. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel. (From Chokroverty et  al. [5]; with permission)

PVDF belts are being used increasingly in PSG, and the AASM considers dual PVDF belts acceptable to measure respiratory effort and the PVDFsum as acceptable alternate sensor to detect apneas and hypopneas during diagnostic studies in adults and children.

register changes in voltage that result from this temperature variation. Thermal devices are not as sensitive as nasal pressure transducers for detecting airflow limitation and, hence, may miss hypopneas. For this reason, the nasal pressure technique to detect airflow (described below) should be routinely used in addition during PSG recording. However, thermistors are used to score apneas. The temperature of the thermal device must be below body temperature in order to sense the temperature difference between expired and inspired air. These devices must therefore not be in contact with the skin. Because of this, they are easily displaced, causing false changes in airflow.

5.5.4 Impedance Pneumography This technique may not precisely measure the respiratory pattern and the volume. Furthermore, there may be electrical interference and, therefore, this technique is not generally used. 5.5.5 Respiratory Magnetometers Respiratory magnetometers record the chest and abdominal motions in both the anteroposterior and lateral directions. This was used as a research technique but has not been popular in practical PSG.

5.6 Airflow Measurement 5.6.1 Oronasal Temperature Monitoring An oronasal thermal device (thermistor, thermocouple or PVDF airflow sensor) placed between the nose and mouth is commonly used to monitor airflow by detecting changes in temperature (cool air flows during inspiration and warm air flows during expiration). A thermistor consisting of wires records changes in electrical resistance, and thermocouples consisting of dissimilar metals (e.g., copper and constantan)

5.6.2 Nasal Pressure Monitoring Nasal airway pressure decreases during inspiration and increases during expiration. In nasal pressure monitoring, a nasal cannula is connected to a pressure-sensitive transducer, which measures this pressure difference. This alternating decrement and increment of nasal pressure produce electrical signals, which indirectly register airflow. Nasal pressure monitoring is more sensitive than thermal devices in detecting airflow limitation and hypopneas. With increased upper airway resistance, the nasal pressure monitor will register a plateau indicating a flow limitation. A DC amplifier or an AC amplifier with a long time constant should be used. One disadvantage is that nasal pressure cannula cannot be used to measure airflow in mouth breathers and in patients with nasal obstruction. For this reason, nasal pressure transducers are not used to score apneas.

26

5.6.3 Pneumotachography This is an excellent technique to measure quantitatively the tidal volume and direct airflow measurement. However, this requires a sealed face mask, creating patient discomfort and sleep disturbance. Hence, it is not used in most of the laboratories.

5.7 Detection of Nocturnal Hypoventilation The gold standard for detection of nocturnal hypoventilation is the analysis of the partial pressure of CO2 in arterial blood (PaCO2) by arterial blood gases (ABGs) drawn in the morning after a sleep study to be compared to the patient’s waking PaCO2. According to AASM guidelines [4], adults who have an increase in their PaCO2 in sleep by 10  mm Hg or more compared to an awake supine PaCO2 have sleep-related hypoventilation. In the pediatric population, sleep-related hypoventilation is defined by the presence of a PaCO2 greater than 50 mm Hg for over 25% of the total sleep time. Nocturnal hypoventilation is seen in conditions where airflow, and therefore alveolar ventilation, is globally decreased, such as primary pulmonary pathology (chronic obstructive pulmonary disease, interstitial lung disease) or neuromuscular diseases like amyotrophic lateral sclerosis or muscular dystrophy. While ABG analysis remains the gold standard for CO2 measurement, it is invasive, impractical and costly for routine use in sleep laboratories. Therefore, the AASM allows for the use of noninvasive surrogate markers of CO2 ­measurement; end-tidal CO2 (EtCO2) or capnography for both PSGs and PAP titration studies, and transcutaneous CO2 (tcCO2) for PAP titration studies. Both these techniques have the added advantage that they allow for continuous monitoring of CO2 levels, unlike arterial CO2 measurements which can only be obtained intermittently. EtCO2 uses an infrared analyzer over the nose and mouth to detect CO2 in the expired air, which closely approximates intra-alveolar CO2 and qualitatively measures the airflow, while tcCO2 uses a pH-­sensitive glass electrode to measure pH changes that result from the diffusion of CO2 across the skin. However, there are several limitations to both techniques. EtCO2 may provide falsely low values in patients with marked nasal obstruction and nasal secretions, in those who are obligate mouth breathers and in those on supplemental oxygen. Similarly, tcCO2 may be unreliable in patients with perfusion problems, skin diseases, edema, or hypovolemia, and the reading often lags behind changes in arterial PCO2 by 2 min or more. Therefore, the AASM recommends using clinical judgment when the findings on capnography or tcCO2 do not fit the clinical picture [4].

S. Chokroverty and S. Bhat

5.8 Oxygen Saturation Oxygen saturation during sleep (SaO2) reflecting arterial oxyhemoglobin saturation, the percentage of hemoglobin that is oxygenated is routinely measured on PSGs noninvasively by finger pulse oximetry. The PSG report should mention the SaO2 nadir and the time the patient spent with an SaO2 below 90%. Two patterns of nocturnal hypoxemia may be seen: 1. Respiratory event-related hypoxemia is seen in sleep disordered breathing, with recurrent desaturations and a return of SaO2 to baseline at the termination of the respiratory event (hypopnea or apnea) 2. Sleep hypoxemia due to primary pulmonary, neuromuscular or skeletal pathology, with low baseline oxygen saturation, potentially worsening in the supine position or in REM sleep (as well as during superimposed respiratory events). This requires an SaO2 below 90% for more than 30% of the total sleep time, or for a minimum of 5  min with an SaO2 nadir of 85% or less.

5.9 Esophageal pH Esophageal pH is monitored by asking patients to swallow a pH probe and recording the output using a DC amplifier. In this manner nocturnal gastroesophageal reflux disease can be detected. However, this is not a routine practice in most sleep laboratories.

5.10 Body Position Monitoring The most reliable technique to record body position is by visual observation. However, during home sleep apnea testing (HSAT) studies with no technologist present, position can be monitored by placing sensors over one shoulder and using a DC channel.

5.11 Snoring Snoring can be monitored by placing a miniature microphone on the patient’s neck. However, there is no accepted grading system to quantify the intensity of this parameter.

Polysomnographic Recording Technique

27

5.12 Monitoring of Penile Tumescence

7 Ending the Test

In normal adult men, penile tumescence occurs during REM sleep and this persists in psychogenic but not in organic impotence. Strain gauges are used to measure penile tumescence. This is not used routinely in most sleep laboratories nowadays.

When the patient is awoken in the morning (either spontaneously or at a set time), “lights on” is reported, and equipment and physiologic calibrations should be performed again to ensure that the equipment and all the devices have been continuously functioning properly throughout the night.

6 Tonic and Phasic Events in REM Sleep

8 Artifacts During PSG Recordings

Events occurring in tonic and phasic REM sleep are listed in Table 6.

Artifacts occurring during PSG may be misinterpreted as abnormalities; therefore, familiarity with commonly seen artifacts is important. Artifacts can be divided into three categories: physiologic (arising from the patient), environmental (arising from electromagnetic radiation from power lines and outlets), and instrumental (arising from the equipment).

Table 6  Tonic and phasic events in rapid eye movement (REM) sleep Tonic events • Electromyogram (EMG) atonia or marked hypotonia • Low amplitude mixed frequency electroencephalography (EEG) resembling awake EEG with or without saw tooth waves; desynchonized EEG • Hippocampal theta rhythm (mostly in animal studies but also noted in human during depth electrode recording) • High arousal threshold • Increased brain temperature (but reduced core body temperature) • Poikilothermia (poikilostatic state) • Presence of olfactory bulb activity • Increased beta (13–30 Hz) and gamma (>30 Hz) rhythms in the EEG

Phasic events • Rapid eye movements • Phasic muscle bursts (myoclonic or transient muscle bursts) • Phasic tongue movements • Periorbital integrated potentials (PIPs) • Middle ear muscle activity (MEMA) • Pontogeniculooccipital waves (PGO or P-waves) (seen in rats and cats; also reported in humans during corticography while performing epilepsy surgery) • Phasic blood pressure swings • Phasic heart rate swings (tachy-bradyarrythmias) • Sawtooth waves on EEG • Alpha bursts in the EEG • Phasic increase of brain intracellular firing rates • Penile erections (men) and clitoral tumescence (women) • Phasic increase of myocardial oxygen demands • Phasic vivid dreaming • Phasic suppression of muscle atonia • Phasic pupillary dilation and constriction • Phasic fractionations of diaphragmatic activity (pauses of 40–80 ms occurring in clusters) correlating with PGO waves

8.1 Physiologic Artifacts These arise from the patient. The category includes muscle artifact obscuring non-EMG channels, movement artifact, respiratory artifact, sweat artifact pulse and EKG artifacts, as well as rhythmic tremorogenic artifacts. Muscle artifact originating from the scalp muscles may obscure the EEG activities and may simulate beta rhythms, and obscure low-amplitude cerebral activities. Figure 5 documents muscle activities during PSG recordings. Movements of the head, eyes, tongue, mouth and other body parts will produce movement artifacts (Fig. 6) that sometimes resemble slow waves in the EEG or may obscure the EEG activities, causing difficulty in scoring the different sleep stages. The rhythmic movements sometimes generated by rhythmic movements of the head, tongue and legs, as well as rhythmic movements generated by tremor in a patient, may produce apparent slow waves, thus causing difficulty in scoring the slow-wave sleep. Sweating may cause excessive baseline swaying, producing a very slow-frequency wave lasting for 1–3 s, which is noticed prominently in the frontal electrodes (Fig.  7). The electrical potentials result from salt content of the sweat

28

S. Chokroverty and S. Bhat

Fig. 5  A 30-s excerpt of PSG recording shows spike-like potentials of varying amplitude and frequency denoting muscle activity not only on the chin electromyography channel but also in all electroencephalography (EEG) channels, particularly prominent in T3, T4, A2 and A1 electrode derivations. EEG, Top ten channels; Lt. and Rt. EOG, left and

right electrooculograms; electromyography of chin; Lt. and Rt. Tib. EMG; left and right tibialis anterior electromyography; P. Flow, peak flow; oronasal thermistor; chest and abdomen effort channels; snore monitor; EKG, electrocardiography

glands. This is an important artifact for the technician to be sensitive to, as it can be improved by cooling the room. Another source of slow, rhythmic artifact in EEG and EMG channels is respiratory artifact, which can be distinguished from sweat artifact by its time-locked relationship with breathing (Figs. 8 and 9). Artifacts in the flow and respiratory channels may arise from various physiological sources, including CPAP leak and ventral hernias (Fig. 10). Eye movement artifacts can be confused with actual cerebral activities (Fig.  11). Figure  12 shows unilateral EOG (due to an artificial eye). Pulse artifact occurs when the electrode is placed over the scalp arteries. Slow waves are generated by electrode movement caused by the pulsations. Temporal relation of these waves to the EKG recording helps

identify such extracerebral activity. EKG artifacts can contaminate the EEG recordings, particularly in patients who are obese with short neck (Fig. 13). Tongue movements may produce characteristic glossokinetic potentials that obscure EEG activity.

8.2 Environmental Sources of Electrical Signals The most common of these is the 60-Hz artifact that results from electromagnetic radiation from AC current in power lines in North America (Fig.  14). The main frequency is 50-Hz in many other countries. At higher PSG

Polysomnographic Recording Technique

29

Fig. 6  A 30-s excerpt from an overnight PSG recording shows body movement artifact following an obstructive apnea as demonstrated by variable, high-voltage, asymmetrical and asynchronous slow activity in electroencephalography (EEG) channels, followed by an arousal. There is simultaneous increased electromyography (EMG) activity in the chin and limb channels. EEG, top eight channels; Chin1-Chin2; submental

EMG. EKG; electrocardiogram. HR; heart rate. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. INTRC; intercostal EMG. CFlow; CPAP airflow channel. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

speeds, the artifact manifests as a thick line in the channel of interest; undulations at a frequency of 60/sec can be counted if the epoch speed is changed to 1 s/screen. The PSG technician should identify this problem and locate the source of this artifact to try to eliminate it. It generally results from poor electrode preparation and contact. Most important is keeping the impedance below 5 Kohms. If the artifact cannot be eliminated, replacement of the electrode is warranted. As described above, the 60  Hz notch filter can be used to retrospectively eliminate the artifact; however, this is discouraged due to potential loss of biologically significant activity.

electrode gel causing abrupt changes in impedance. The electrode should be reset and gel applied. If the artifact persists, then electrodes need to be changed. Other sources of artifacts are the electrode wires, the cables and the switches. In the PSG machine, random fluctuations of charges result in some instrumental noise artifacts. If the sensitivity is greater than 2 microvolts per millimeter, which is not generally used in PSG recordings, then these instrument artifacts may interfere with the recording. Loose contacts in switches or wires may also cause sudden changes in voltage or loss of signal. Artifacts due to implanted electrical devices such as pacemakers, deep brain stimulators (Fig. 16) and vagal nerve stimulators may also be encountered. PSG artifacts must be eliminated or reduced to a minimum for proper identification of the EEG, EMG and other bioelectrical potentials for correct staging of the sleep and for diagnosis and classification of sleep disorders. Identification of all these artifacts is best made at the time of the recording by the skilled PSG technologist. Particular attention should be paid

8.3 Instrumental Artifacts These arise from the equipment. Electrode “pops” may produce transient sharp waves or slow waves limited to one electrode (Fig.  15). These artifacts are very common and result from suboptimal electrode placement of insufficient

30 Fig. 7  A 30-s epoch from the polysomnogram study of a 43-year-old woman referred to the sleep laboratory to rule out obstructive sleep apnea. Note the slow, irregular swaying of the baseline EEG channels, representing sweat artifact. Top three channels; EEG. LOC-A2, ROC-A1, electrooculogram channels. CHIN; submental EMG. LT. Tib, RT. Tib, LT. GAS, RT. GAS; left and right tibialis anterior and gastrocnemius muscle EMG. ORONA; oronasal thermistor. ABD; abdominal effort belt. SaO2; arterial oxygen saturation via pulse oximetry. Also included is a snore channel. (From Siddiqui et al. [6]; with permission)

Fig. 8  Another 30-s epoch from the same polysomnogram as depicted in Fig. 7, recorded later in the night. The technician eliminated the sweat artifact noted in the previous epoch by cooling the room, but a new, slow artifact is noted in the EEG leads. Note that it is regular and time-locked to breathing as recorded in the flow and effort channels. This allows its identification as respiratory artifact. (From Siddiqui et al. [6]; with permission)

S. Chokroverty and S. Bhat

Polysomnographic Recording Technique

31

Fig. 9  A 60-s epoch from the overnight CPAP titration study of a 54-year-old man with mild obstructive sleep apnea. Note the rhythmic, slow artifact occurring in the left electrooculogram lead (E1-M1), which is time-locked to respirations as noted in the flow and effort channels. Unlike in Fig. 8, this artifact is restricted to a single electrode (E1). Top eight channels; EEG recording with electrodes placed accord-

ing to the 10–20 international electrode placement system. Chin1-­ Chin2; submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. CFlow; CPAP airflow channel. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

Fig. 10  A 30-s epoch from the polysomnogram of a 64-year-old man referred for evaluation of upper airway obstructive sleep apnea syndrome. The abdomen effort channel reveals a double-peaked waveform in the post-apnea recovery breaths. This may be related to the clinical finding of a large protuberant abdomen with a ventral midline hernia.

EEG, Top 10 channels; Lt. and Rt. EOG, left and right electrooculograms; electromyography of chin; Lt./Rt. Tib. EMG; left and right tibialis anterior electromyography; oronasal thermistor; chest and abdomen effort channels; snore monitor; EKG, electrocardiography; SaO2, oxygen saturation by finger oximetry

32

Fig. 11  A 30-s epoch from the overnight PSG of a 64-year-old woman referred for possible sleep apnea. In the beginning of the epoch there are several eye blink artifacts, best noted in the electrooculogram channels (E1-M1, E2-M2) and in the frontal EEG channels (F3-M2, F4-M1). This is rapid eye flutter (a series of well-defined eye blinks occurring in rapid succession), best seen in the same channels. These artifacts may be misinterpreted for abnormal cerebral activity. EEG, Top 8 channels;

S. Chokroverty and S. Bhat

Chin1-Chin2; submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. OroNs1-OroNS2; oronasal airflow. Pflow; nasal pressure transducer. Chest and ABD; effort belts. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

Polysomnographic Recording Technique

Fig. 12  A 30-s epoch in REM sleep from the polysomnogram of a 50-year-old man referred for exclusion of upper airway obstructive sleep apnea syndrome. Past medical history is significant for an artificial left eye and limited vision in the right eye secondary to severe uveitis. Note the presence of unilateral eye movement artifact in the right electrooculogram and anterior temporal EEG electrodes but not in the corresponding channels on the left side. Myoclonic bursts of electromyographic activity are recorded in the tibialis electromyography chan-

33

nels and a mixed apnea is recorded on the oronasal thermistor and effort channels. A superimposed electrocardiography artifact is also noted. EEG, Top 10 channels; Lt. and Rt. EOG, left and right electrooculograms; electromyography of chin; Lt. and Rt. Tib. EMG, left and right tibialis anterior electromyography; P. flow, peak flow; oronasal thermistor; chest and abdomen effort channels; snore monitor; EKG, electrocardiography

34

Fig. 13  This is a 30-s excerpt from an overnight PSG selected to show electrocardiography (EKG)artifact, which is characterized by its morphology, rhythm and synchrony with EKG recording. It is seen throughout the epoch in the electroencephalography (EEG) and electrooculography (EOG) channels, as well as the bilateral gastrocnemius EMG channels. Top eight channels; EEG recording with electrodes placed according to the 10–20 international electrode placement

S. Chokroverty and S. Bhat

system. E1-M2 and E2-M2; EOG channels. Chin1-Chin2; submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. OroNs1-OroNS2; oronasal airflow. Pflow; nasal pressure transducer. Chest and ABD; effort belts. LTIB, RTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

Polysomnographic Recording Technique

35

a

b

Fig. 14 (a) A 60-s epoch from the polysomnogram of a 46-year-old man referred to evaluate for sleep apnea. Note the dark band-like appearance of the chin (Chin1-Chin2) and snore channels. (b) The same epoch viewed at a 1  s/epoch speed. There is an artifact occurring at 60 Hz in the above channels. This artifact is due to electrical interference from power lines and equipment occurring at a frequency of 60 Hz in North America (but at 50  Hz in many other countries). Maximum

interference is seen in the presence of poor electrode contact. EEG, top 8 channels; E1-M1, E2-M1, left and right electrooculograms; EKG, electrocardiography; HR, heart rate; LTIB, RTIB, LGAST, RGAST, left and right tibialis anterior and gastrocnemius EMGs; INTRC, intercostal EMG; oronasal thermistor; chest and abdomen effort channels; snore monitor; SaO2, oxygen saturation

36

S. Chokroverty and S. Bhat

a

Fig. 15 (a) 10 s/page epoch from the polysomnogram of a 24-year-old man referred to evaluate for sleep apnea. Note the occurrence of a sharp artifact in all channels referenced to the left mastoid (M1) electrode. This is an example of an electrode pop, due to intermittently poor electrode contact. Such artifact may be mistaken for epileptiform activity. Incidental note is made of ventricular bigeminy in the EKG channel. (b)

b

The same epoch viewed at a 30  s/page speed. EEG, top 8 channels; E1-M1, E2-M1, left and right electrooculograms; EKG, electrocardiography; HR, heart rate; LTIB, RTIB, LGAST, RGAST, left and right tibialis anterior and gastrocnemius EMGs; INTRC, intercostal EMG; oronasal thermistor; nasal pressure transducer; chest and abdomen effort channels; snore monitor; SaO2, oxygen saturation

Polysomnographic Recording Technique

37

a

b

Fig. 16  Representative epochs from the polysomnogram of a 68-year-­ old woman referred to the sleep laboratory for excessive daytime sleepiness and snoring. She has a history of Parkinson’s disease with gait instability and frequent falls leading to wheelchair dependence, micrographia and trouble turning in bed. Due to a lack of response to medical therapy, she underwent bilateral subthalamic deep brain stimulator (DBS) placement. (a) A 30-s epoch. (b) The same epoch viewed at a 1 s speed. Note the DBS artifact at 20 Hz obscuring all the recording channels except the flow and effort belts and the EKG channel. Top six channels; EEG recording with electrodes placed according to the 10–20 international electrode placement system. E1-M2 and E2-M2; electrooculogram channels. EKG; electrocardiogram. HR; heart rate.

Masseter-REF; masseter EMG. Chin1-Chin2; submental electromyogram (EMG). OroNs1-OroNS2; oronasal airflow. Pflow; nasal pressure transducer. Chest and ABD; effort belts. RSterno1-RSterno2; right sternocleidomastoideus EMG. LBiceps1-LtElbow, RBiceps1-RtElbow; left and right biceps brachii EMG. LTriceps1-LtElbow, RTriceps1-­ RElbow; left and right triceps EMG. RQuad-iliac crest, LtQuad-iliac crest; right and left quadriceps femoris EMG. LTIB, RTIB; left and right tibialis anterior EMG. RGastr1-RtGastr-2, LtGastr1-LtGastr2; right and left gastrocnemius EMG. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel. EKG artifact is also noted to occur in the quadriceps EMG channel

38

to the electrodes, electrode gel application and impedance. Preparation of the patient and electrode application is fundamental for recording a good quality PSG.

9 Digital Polysomnography Digital PSG has now completely replaced analog recording. The following are the advantages of digital recording: • Previously recorded data can be manipulated retrospectively, and changes can be applied to the filter settings, sensitivities and monitor speeds, thereby minimizing and eliminating artifacts. • Data can be analyzed in multiple ways, and areas of interest can be more easily pinpointed and logged for future reference. • Autoscoring capabilities allow both the polysomnographer to be more efficient in evaluating and reporting data. • Computerized PSG recording is paperless, thus conserving space and being environmentally friendly. • Digital information is stored on inexpensive media such as DVDs, and the digital format translates more easily into databases. On the other hand, some disadvantages of digital PSG include: • Incompatibility of recordings created by software produced by different manufacturers, making sharing of data between laboratories cumbersome. Steps are being taken to address this problem. • Hardware issues may result in breach or loss of data. There are certain minimal requirements for digital PSG. • The sampling rate, which refers to the frequency with which a signal is converted to a digital format, should ideally be 200 Hz for EEG, EOG and EMG channels. • A 12-bit analog-digital conversion is a suggested minimum acceptable for the digital amplitude resolution. • It is important to have scroll-back mode without interrupting data acquisition so that a particular change during recording can be compared with the previous signals. • The computer screen must be of sufficient size and have high resolution.

10 Home Sleep Apnea Testing (Portable Monitoring/Oligosomnography) Due to the expense and labor-intensive nature of in-­laboratory sleep studies, and the convenience afforded to patients, the use of portable monitoring devices for home sleep apnea

S. Chokroverty and S. Bhat

studies (HSAT) is rapidly increasing. Sleep studies may be classified into four types: 1. Type I devices refer to traditional attended in-laboratory PSGs. 2. Type II devices require a minimum of seven channels, including EEG/EOG, chin EMG, EKG, oximetry, airflow and respiratory effort channels. The presence of EEG/ EOG and EMG channels permits sleep staging and reporting of a traditional AHI. 3. Type III studies (also called “cardiopulmonary studies”) have a minimum of four channels (airflow, respiratory effort, pulse oximetry and EKG); these studies can be attended or unattended. Lack of EEG/EOG and EMG channels results in inability to perform sleep staging or detection of sleep/wake state; therefore, these studies report a “respiratory event index” (REI), based on recording time and not including events associated with cortical arousals in the absence of EEG channels, rather than a traditional AHI based on sleep time (Fig. 17). 4. Type IV devices may consist of a single channel such as pulse oximetry alone (monosomnography), but to be approved by the Centers for Medicare & Medicaid Services (CMS), these devices require a minimum of three channels measuring airflow or thoracoabdominal movement and calculating an index to quantify OSA severity. The above classification system does not include other technologies, including those measuring peripheral arterial tonometry (PAT) for diagnosing OSA.  In 2008, CMS updated their policies to allow coverage for CPAP devices for patients diagnosed with OSA based on HSAT performed using unattended Type II and Type III devices, as well as Type IV devices measuring at least three channels and those measuring actigraphy, oximetry and PAT [7]. The AASM does not recommend that Type IV devices be used in the diagnosis of OSA due to lack of sensitivity or specificity (but has approved the use of PAT technology with actigraphy and oximetry as technically adequate for HSAT [8]. In 2018, the AASM issued updated guidelines for the clinical use of HSAT [9]. They observed that although less sensitive than PSG in the detection of OSA, HSAT can be ordered by a physician for the diagnosis of suspected moderate-tosevere OSA where the patient lacks conditions that would preclude the use of an HSAT (significant cardiorespiratory disease including chronic obstructive pulmonary disease or congestive heart failure, potential respiratory muscle weakness due to neuromuscular condition, awake hypoventilation or suspicion of sleep related hypoventilation, chronic opioid medication use, history of stroke or severe insomnia, or those with disorders of central hypersomnolence, parasomnias, and sleep related movement disorders). Their recommendations were:

Polysomnographic Recording Technique

39

Fig. 17  Representative 90-s epoch of a Type III home sleep apnea test (HSAT) showing recurrent obstructive apneas (pink boxes) with a hypopnea partially seen at the end (green box). SpO2 desaturations of up to 11% are noted in the oximetry channel (Ox%, blue boxes). These

devices have flow and effort channels (labeled), as well as pulse and snore channels and body position notations, but lack EEG channels; therefore, sleep staging is not possible and respiratory indices are reported over recording rather than sleep time

• Only a physician can diagnose medical conditions such as OSA and primary snoring. • The need for, and appropriateness of, an HSAT must be based on the patient’s medical history and a face-to-face evaluation by a physician, either in person or via telemedicine. • An HSAT is a medical assessment that must be ordered by a physician to diagnose OSA or evaluate treatment efficacy. • An HSAT should not be used for general screening of asymptomatic populations. • Diagnosis, assessment of treatment efficacy and treatment decisions must not be based solely on automatically scored HSAT data, which could lead to sub-optimal care that jeopardizes patient health and safety. • The raw data from the HSAT device must be reviewed and interpreted by a physician who is either board-­ certified in sleep medicine or overseen by a board-­certified sleep medicine physician.

• If a single HSAT is negative, inconclusive or technically inadequate, full in-laboratory PSG should be performed for the diagnosis of OSA. The practice of sleep medicine is increasingly moving toward a completely out-of-center workflow, with patients with uncomplicated OSA undergoing HSAT, being prescribed autotitrating devices and following up with their clinician for adherence visits, potentially via telemedicine. Indeed, a recent randomized, unblinded, open-label multicenter trial compared an entirely in-laboratory approach (with both diagnostic PSG and CPAP titration studies being performed in-laboratory) with an entirely ambulatory approach (HSAT followed by prescription of CPAP based on readings from an auto-CPAP device) and found noninferiority of the ambulatory approach in terms of acceptance, adherence and patient outcomes in patients with moderate-to-severe OSA and no medical comorbidities [10]. This management paradigm was further accelerated by the COVID-19 pandemic [11], and is expected to constitute an ever-expanding proportion of

40

patients in most sleep medicine practices in the coming years.

11 Indications for Polysomnography and HSAT The following are clinical indications for PSG (HSAT may be substituted only where indicated) [12]: • PSG or HSAT is routinely indicated for the diagnosis of sleep-related breathing disorders. • PSG or HSAT is indicated pre-operatively in patients who are being evaluated for upper airway surgery for snoring or OSA. • For patients with suspected narcolepsy, an in-laboratory PSG is mandatory before a multiple sleep latency test (MSLT). However, an overnight PSG before a maintenance of wakefulness test (MWT) is at the discretion of the polysomnographer • PSG is indicated in patients with parasomnias that are unusual or atypical, or behaviors that are violent or injurious to the patient or others, especially those not responding to standard treatment or with forensic implications. • PSG is indicated for patients with excessive leg movements in sleep, reported by either a patient or bedpartner, both to diagnose periodic limb movement disorder (PLMD) or where there is suspicion that the movements are triggered by sleep disordered breathing. The patient should also have complaints of fragmented sleep and excessive daytime sleepiness. Additionally, follow-up studies are indicated for the following [13]:

S. Chokroverty and S. Bhat

• Routine reassessment of asymptomatic patients with obstructive sleep apnea on PAP therapy, however, ­follow-­up PSG or HSAT can be used to reassess patients with recurrent or persistent symptoms, despite good PAP adherence. • Patients with uncomplicated, nonviolent parasomnias. • Patients with nocturnal seizures in whom the diagnosis has been established by clinical evaluation and routine awake/asleep EEGs. • Patients with RLS (which is a clinical diagnosis) or uncomplicated PLMS (no complaints of fragmented sleep/sleep disordered breathing or excessive daytime sleepiness). • In patients with suspected depression. • In patients with suspected circadian rhythm disorders • In the management of insomnia

12 Pitfalls of PSG and HSAT • Poor sleep efficiency, failure to capture supine or REM sleep and failure to account for factors that would affect muscle tone (such as benzodiazepines or alcohol use) may lead to underestimation of the degree of OSA or false negative studies • Standard PSG and HSAT cannot diagnose upper airway resistance syndrome definitively due to lack of esophageal manometry • Standard PSG and HSAT may miss hypoventilation (due to lack of pCO2 monitoring), which is an early abnormality (particularly REM-related hypoventilation) in neuromuscular disorders. • Standard PSG and HSAT do not include autonomic monitoring, which may be important for assessment of autonomic arousal (alterations in heart rate and blood pressure) as well as for assessing autonomic changes which are intense during sleep.

• Follow-up PSG or HSAT is recommended to assess response to treatment with non-PAP interventions. • Follow-up PSG or HSAT may be used if clinically significant weight gain or loss has occurred since diagnosis of OSA or initiation of its treatment. • Follow-up PSG may be used for reassessment of sleep-­ References related hypoxemia and/or sleep-related hypoventilation 1. Butkov N, Keenan SA.  An overview of polysomnographic techfollowing initiation of treatment for OSA. nique. In: Chokroverty S, editor. Sleep disorders medicine: basic • Follow-up PSG or HSAT may be used in patients being science, technical considerations, and clinical aspects. 4th ed. treated for OSA who develop or have a change in cardioNew York, NY: Springer; 2017. 2. Chokroverty S.  Polysomnography and related procedures. In: vascular disease. Hallett M, editor. Movement disorders. Handbook of clinical neu• Follow-up PSG may be used in patients with unexplained rophysiology, vol. 1. Amsterdam: Elsevier; 2003. p. 139–51. PAP device-generated data. 3. Chokroverty S, Vertugno R. Polysomnography: technical and cliniIn the absence of signs and symptoms suggestive of sleep-­ disordered breathing, neither PSG nor HSAT is recommended for:

cal aspects. In: Schomer DL, Da Silva L, editors. Niedermeyer’s electroencephalography: basic principles, clinical science, and related fields. 7th ed. New  York, NY: Oxford University Press; 2018.

Polysomnographic Recording Technique 4. Berry RB, Brooks R, Gamaldo CE, et al. The AASM manual for the scoring of sleep and associated events: rules, terminology, and technical specifications, version 2.5. Darien, IL: American Academy of Sleep Medicine; 2018. 5. Chokroverty S, Bhat S, Rubinstein M. Periorbital integrated potentials: useful phasic REM sleep markers. Sleep Med. 2017;37:74–6. 6. Siddiqui F, Osuna E, Walters AS, Chokroverty S. Sweat artifact and respiratory artifact occurring simultaneously in polysomnogram. Sleep Med. 2006;7(2):197–9. 7. CMS Manual System, Department of Health & Human Services (DHHS), Pub 100–03 Medicare National Coverage Determinations, Centers for Medicare & Medicaid Services (CMS),Transmittal 86, Change Request 6048 (2008) 8. Kapur VK, Auckley DH, Chowdhuri S, Kuhlmann DC, Mehra R, Ramar K, Harrod CG.  Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea: an American academy of sleep medicine clinical practice guideline. J Clin Sleep Med. 2017;13(3):479–504. 9. Rosen IM, Kirsch DB, Carden KA, Malhotra RK, Ramar K, Aurora RN, Kristo DA, Martin JL, Olson EJ, Rosen CL, Rowley JA, Shelgikar AV, American Academy of Sleep Medicine Board

41 of Directors. Clinical use of a home sleep apnea test: an updated American academy of sleep medicine position statement. J Clin Sleep Med. 2018;14(12):2075–7. 10. Rosen CL, Auckley D, Benca R, et al. A multisite randomized trial of portable sleep studies and positive airway pressure autotitration versus laboratory-based polysomnography for the diagnosis and treatment of obstructive sleep apnea: the HomePAP study. Sleep. 2012;35(6):757–67. 11. Kole A. Home sleep apnea testing in the era of COVID-19: a community perspective. J Clin Sleep Med. 2020;16(9):1633. 12. Kushida CA, Littner MR, Morgenthaler T, Alessi CA, Bailey D, Coleman J Jr, Friedman L, Hirshkowitz M, Kapen S, Kramer M, Lee-Chiong T, Loube DL, Owens J, Pancer JP, Wise M.  Practice parameters for the indications for polysomnography and related procedures: an update for 2005. Sleep. 2005;28(4):499–521. 13. Caples SM, Anderson WM, Calero K, Howell M, Hashmi SD. Use of polysomnography and home sleep apnea tests for the longitudinal management of obstructive sleep apnea in adults: an American academy of sleep medicine clinical guidance statement. J Clin Sleep Med. 2021;17(6):1287–93.

Hypnogram and Compliance Graph Analysis Robert J. Thomas, Sudhansu Chokroverty, and Sushanth Bhat

1 Introduction

even resolving in REM sleep [3]. A prolonged sleep onset latency followed by consolidated sleep may suggest delayed The hypnogram is a compressed graphical summary of a circadian phase. Fragmentation of the sleep cycle can be secwhole-night sleep study. It allows, on a single page, a repre- ondary to a severe first night effect or poor sleep hygiene. sentation of multiple variables including sleep stages, respi- Worsening of periodic limb movements in sleep (PLMS) in ratory events, positive airway pressure (PAP) (if used), motor the first third to half of the night [4] is also a commonly seen movements, oximetry, end-tidal or transcutaneous CO2, heart pattern. rate variability measures, electroencephalographic (EEG) Hypnograms are available for review in both home sleep power spectrum, and body position [1]. apnea testing (HSAT) and in-laboratory polysomnograms Information in a hypnogram has two dimensions—hori- (PSGs). With HSAT, the signals are relatively simple, but the zontal and vertical. Examples of data displayed on the hori- number of variations nearly infinite. Figures 1, 2, 3, 4, 5, 6, zontal dimension are the flow of sleep stages across the and 7 provide some examples, with patterns discernable “at night, the oximetry profile, and the occurrence of respiratory a glance.” These include classic obstructive sleep apnea, high events. Examples of data displayed on the vertical dimension loop gain apnea, possible hypoventilation, informative heart are the effect of body position and sleep stage on respiration rate kinetics, and even the quality of sleep. The full PSG has and oxygenation. Some patterns are common and character- far richer information across multiple physiological systems, istic, such as stage-related sleep disordered breathing (rapid resulting in more complex hypnograms. Some illustrative eye movement [REM]-predominant obstructive sleep apnea examples are highlighted in Figs. 8, 9, 10, 11, 12, 13, 14, 15, [OSA]), positional sleep disordered breathing (supine-­ 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, and 26. Additionally, predominant OSA) [2], central sleep apnea including stacked therapy-use hypnograms (compliance graphs), both Cheyne–Stokes respirations (associated with minimal dis- for continuous positive airway pressure (CPAP) use and for ease in REM sleep and smooth-symmetrical moderate oxy- newer modalities such as hypoglossal nerve stimulation [5], gen desaturations in non-REM [NREM] sleep) and are routinely evaluated in the clinical setting and provide Cheyne–Stokes respirations which worsens in the latter half important clinical information (Figs. 27 and 28). of the night and is more severe in NREM sleep, improving or

R. J. Thomas Department of Medicine, Division of Pulmonary, Critical Care and Sleep, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA e-mail: [email protected] S. Chokroverty HMH JFK University Medical Center and Neuroscience Institute, Hackensack Meridian School of Medicine at Seton Hall University, South Orange and Nutley, NJ, USA

Rutgers Robert Wood Johnson University Medical Center, New Brunswick, NJ, USA S. Bhat (*) Department of Neurology, Hackensack Meridian School of Medicine, Nutley, NJ, USA JFK Neuroscience Institute, Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. J. Thomas et al. (eds.), Atlas of Sleep Medicine, https://doi.org/10.1007/978-3-031-34625-5_4

43

44

R. J. Thomas et al.

Fig. 1  Home sleep apnea testing (HSAT) hypnogram showing severe obstructive sleep apnea and possible hypoventilation. Key features are severe hypoxia with a lack of return to baseline suggesting hypoventila-

tion, V-shaped oxygen desaturations at the nadirs, loss of heart rate dipping with mostly a reverse increase, and non-stop respiratory events of relatively similar duration

Fig. 2  Home sleep apnea testing (HSAT) hypnogram showing severe OSA with likely hypoventilation. Note the sagging oximetry trace during a period of likely stable NREM sleep in the first sleep cycle, and the

persistently low baseline during periods of deep/stable sleep either supine (snoring recorded) or lateral (no snoring)

Hypnogram and Compliance Graph Analysis

45

Fig. 3  Home sleep apnea testing (HSAT) hypnogram showing mixed physiology apnea with supine dominance. Blue events are central apneas. Note clear worsening during supine periods. The conversion to

largely central events late in the recording from an earlier mixed even profile is well described in the literature

Fig. 4  Home sleep apnea testing (HSAT) hypnogram showing abnormal heart rate profile. During the first third of the night, when snoring is continuous and saturations stable (suggesting N3 or stable NREM

sleep), heart rate increases. This may be from obstructive hypoventilation and hypercapnia-driven tachycardia, as hypoxia is absent

46

R. J. Thomas et al.

a

b

Fig. 5 (a) Home sleep apnea testing (HSAT) hypnogram showing supine-dominant sleep apnea with high loop gain. The oxygen desaturation pattern is band-like, from self-similar respiratory events, which reflects high loop gain. (b) A close-up of the respiratory events

Hypnogram and Compliance Graph Analysis

47

Fig. 6  Home sleep apnea testing (HSAT) hypnogram showing high loop gain apnea. Band-like oxygen desaturation is the clue, even if the majority of individual events are scored as “obstructive apneas.” The respiratory event duration is also relatively even (self-similar)

Fig. 7  Home sleep apnea testing (HSAT) hypnogram showing mixed physiology apnea. Band-like oxygen desaturation of high loop gain, alternating with V-shaped oxygen desaturation of REM sleep

48

R. J. Thomas et al.

Fig. 8  High-quality sleep detected from a home sleep apnea test (HSAT). The continued snoring which is periodically interrupted by likely REM sleep (fractionation or inhibition of snoring, respiratory

events, and heart rate fluctuations) suggests well-maintained macro-­ architecture of sleep, and consolidated NREM sleep

Fig. 9  An average polysomnogram (PSG) hypnogram. It is rare to have a normal study in a clinical sleep laboratory, but this sample is close. There is the normal distribution of N3 (blue), cycling of sleep, though the normal progression of REM sleep is abnormal with a larger middle

period. There is REM-dominant apnea, and thus, chronic REM sleep fragmentation, which can increase REM pressure and may explain the prominent middle of sleep REM period

Hypnogram and Compliance Graph Analysis

Fig. 10  Polysomnogram (PSG) hypnogram showing possible idiopathic hypersomnia. The large increase in N3 even to the edge of morning awakening raises this possibility. This 24-year-old female indeed

49

sleeps over 15 h in any 24-h period. The study was ordered as an unconstrained PSG (to allow the patient to sleep as long as possible) but was inadvertently terminated at a conventional time (6 AM)

50

Fig. 11  Polysomnogram (PSG) hypnogram showing “sleep failure.” The exact reverse of idiopathic hypersomnia with severely fragmented sleep, sleep apnea, and substantial periodic limb movements during a

R. J. Thomas et al.

CPAP titration. In the second half of the study, PLM events are occurring even when awake (red; PLMS, during sleep, are in blue)

Hypnogram and Compliance Graph Analysis

Fig. 12  Polysomnogram (PSG) hypnogram showing high loop gain, periodic limb movements, and severe sleep microfragmentation. Note continuous N2-N1 transitions, lack of clear cycling of sleep, reduced

51

REM sleep, band-like oxygen desaturation (high loop gain), and periodic limb movements across the whole night, not just the first half, a more typical distribution

52

Fig. 13  Polysomnogram (PSG) hypnogram showing rapid eye movement (REM) sleep rebound during first night of CPAP titration. “Rebound” may be reflected in increased percentage, larger initial REM periods, particularly long for time of time periods, increased

R. J. Thomas et al.

REM density, or increase in the number of REM periods. Here, the dissipating REM drive results in an “inverse progression” of REM sleep distribution

Hypnogram and Compliance Graph Analysis

Fig. 14  Polysomnogram (PSG) hypnogram showing rapid eye movement (REM) and non-REM (NREM), specifically N3, rebound. A commonly seen effect of successful CPAP, this degree of N3 is unusual

53

(across the whole night) and may reflect other influences on slow wave amplitude and occurrence, such as certain medications

54

Fig. 15  Polysomnogram (PSG) hypnogram showing excessive N3 for age. This 66-year-old female with multiple sclerosis treated with dimethyl fumarate demonstrates increased (percentage and dispersion) of conventional slow wave sleep. Both sleep fragmentation and distor-

R. J. Thomas et al.

tions of sleep architecture may be seen in multiple sclerosis. In this instance, the patient was also using baclofen for spasticity, which likely contributed to the increased N3

Hypnogram and Compliance Graph Analysis

Fig. 16  Polysomnogram (PSG) hypnogram showing circadian profile of periodic limb movements of sleep (PLMS). PLMS is usually maximal in the first half of the night. Also note the increased frequency of PLMS early with gradual dissipation and final near dissolution. The inter-movement interval as a consequence lengthens during the course

55

of the night. Let movement trace (red) are movements scored regardless of sleep, and thus show periodic limb movements while awake (PLMWA), a typical feature of restless legs which can be quantified by the Suggested Immobilization test

56

R. J. Thomas et al.

Fig. 17  Polysomnogram (PSG) hypnogram showing rapid eye movement (REM) and position-dominant sleep apnea. Note that non-supine REM sleep shows minimal oxygen saturation fluctuations

Hypnogram and Compliance Graph Analysis

Fig. 18  Polysomnogram (PSG) hypnogram showing supine-specific obstructive sleep apnea (OSA). A 28-year-old man presenting with loud snoring, witnessed apneas and excessive daytime sleepiness, with bilaterally enlarged tonsils. His BMI was 33. The hypnogram shows clustering of frequent obstructive apneas and hypopneas in the supine position. The apnea-hypopnea index (AHI) was 62.3/h in the supine position with oxygen desaturations (O2 saturation less than 90% for 2.1% of the

57

total sleep time); AHI was 2.4/h in the lateral positions. Note that events depend only on position and appear unaffected by stage of sleep (being present both in supine N1/N2 sleep and supine REM sleep, although improved in N3 sleep which is a well-known pattern), making this purely positional OSA. HR heart rate, SaO2 oxygen saturation by finger oximetry, LM leg movements

58

R. J. Thomas et al.

Fig. 19 Polysomnogram (PSG) hypnogram showing prone-­ tive sleep-apnea-hypopnea in the prone position (AHI 47/h), indepenpredominant obstructive sleep apnea. A 61-year-old woman presented dent of stage of sleep. Prone position induced more severe apnea than with excessive daytime sleepiness, snoring, and frequent nasal conges- in supine or lateral position; an unusual finding that may have been tion and stuffiness. Her BMI is 22. Hypnogram shows mild obstructive related to nasal obstruction in this patient. HR heart rate, PLM periodic sleep apnea (AHI of 11.5/h) but recurrent episodes of moderate obstruc- limb movements, SaO2 oxygen saturation by finger oximetry

Hypnogram and Compliance Graph Analysis

Fig. 20  Polysomnogram (PSG) hypnogram showing Cheyne–Stokes respiration worsening throughout the night. Progressive worsening of respiratory events as the night progresses, with events maximal in the last third of the night independent of the body position and sleep stage (although even in this section, the events in REM sleep appear to be more obstructive hypopneas labelled in pink as opposed to mixed and central apneas in NREM sleep, labelled in green and blue, respectively).

59

The patient is an 81-year-old man with a history of excessive daytime sleepiness, disturbed nocturnal sleep, hypertension, and coronary artery disease. This pattern strongly suggests occult congestive cardiac failure. His overnight PSG showed mixed and central apneas, including Cheyne–Stokes breathing. HR heart rate, PLM periodic limb movement, SaO2 oxygen saturation by finger oximetry

60

R. J. Thomas et al.

Fig. 21  Polysomnogram (PSG) hypnogram showing a perfect night of continuous positive airway (CPAP) titration. Severe apnea (evident at the lower CPAP pressures) with subsequent normalization of sleep architecture and oxygenation, associated with both N3 and REM rebound

Hypnogram and Compliance Graph Analysis

Fig. 22 Polysomnogram (PSG) hypnogram showing imperfect response of sleep to CPAP. Compared to Fig. 18, in this instance, while respiratory targets are achieved, sleep remains poor, with both micro (stage transitions) and macro (interspersed wake after sleep onset blocks, poor sleep cycle structure, minimal REM sleep) architectural

61

levels. The causes are numerous, include a lab night effect, excessive daytime napping, and conditions with increased sleep vulnerability especially when associated with insomnia. Plausibly, COMISA (Co-Morbid Insomnia and Sleep Apnea) may also be associated with this pattern during CPAP titration

62

Fig. 23  Polysomnogram (PSG) hypnogram showing difficult positive airway pressure (PAP) titration night. Both CPAP and bilevel ventilation are only partially effective, as there is hypoxia, hypercapnia (not

R. J. Thomas et al.

shown in this graph), micro and macro sleep fragmentation, and severe desaturations in REM sleep despite reasonable pressures. Note the persistent fluctuations is respiratory rate (lowest trace) across the night

Hypnogram and Compliance Graph Analysis

Fig. 24  Polysomnogram (PSG) hypnogram showing periodic limb movements while awake. A 42-year-old with severe sleep onset insomnia and restless legs. Note the predominance of PLMS while awake.

63

Current scoring and computational approaches may readily exclude these movements

64

Fig. 25  Polysomnogram (PSG) hypnogram showing interaction of REM sleep and PLMS. The first REM sleep is unable to completely suppress PLMS, while the second REM period completely silences the motor system. This reflects the relative “power” of competing systems.

R. J. Thomas et al.

Though typically REM sleep suppresses PLMS, the latter drivers can be powerful enough to overcome. Continuation of NREM PLMS into REM sleep is not REM sleep without atonia

Hypnogram and Compliance Graph Analysis

65

a

b

Fig. 26 (a) Polysomnogram (PSG) hypnogram showing failed hypoglossal nerve stimulation. There is persistent sleep fragmentation and severe distortion of overall sleep architecture and ongoing respiratory events The oxygen desaturation profile is band-like, suggesting unrecognized high loop gain, which predicts failure of this approach to treatment of sleep apnea. (b) Close-up of failing hypoglossal nerve

stimulation from Fig. 23. Ongoing respiratory events despite maximal tolerable stimulation. Sleep fragmentation is severe, respiratory events are short cycle (less than 30  s and self-similar, features of high loop gain), and the duration of inspiration is also fluctuating (the periods of the stimulation artifact, pink chin EMG trace)

66

Fig. 27 Stacked PAP use hypnograms (compliance graphs) in a 60-year-old male with a complicated circadian disorder. Each horizontal line is 24 h, mid-night is at left edge. Stacked hypnograms are typical of sleep apnea therapy device graphical representations of use. Stacking can be vertical or horizontal. This case shows hyper-

R. J. Thomas et al.

somnia, unstable sleep-wake times (durations), and suggestion of progression (non-24). In fact, this patient converted from severe stable delay into a non-24 during the pandemic. The use of CPAP, which he never sleeps without, captured the dynamics of circadian dysregulation

Hypnogram and Compliance Graph Analysis

Fig. 28  Horizontal stacking of apnea-therapy use hypnograms. Inspire hypoglossal nerve stimulator. This is data from the adjustment period before titration in the sleep laboratory. There is overall progressive

References 1. Swihart BJ, Caffo B, Bandeen-Roche K, Punjabi NM. Characterizing sleep structure using the hypnogram. J Clin Sleep Med. 2008;4(4):349–55. 2. Eiseman NA, Westover MB, Ellenbogen JM, Bianchi MT.  The impact of body posture and sleep stages on sleep apnea severity in adults. J Clin Sleep Med. 2012;8(6):655–6A.

67

increase in use but sleep timings are unstable, which may in part reflect the novelty of the experience

3. Orr JE, Malhotra A, Sands SA. Pathogenesis of central and complex sleep apnoea. Respirology. 2017;22(1):43–52. 4. Drakatos P, Olaithe M, Verma D, et al. Periodic limb movements during sleep: a narrative review. J Thorac Dis. 2021;13(11):6476–94. 5. Mashaqi S, Patel SI, Combs D, et al. The hypoglossal nerve stimulation as a novel therapy for treating obstructive sleep apnea-a literature review. Int J Environ Res Public Health. 2021;18(4):1642.

Electroencephalography for the Sleep Specialist Sudhansu Chokroverty, Eli S. Neiman, and Sushanth Bhat

1 Introduction An understanding of the technical aspects of electroencephalography (EEG) and recognition of basic normal and abnormal EEG patterns in wakefulness and sleep are essential skills for a polysomnographer. EEG records the potential difference between two electrodes placed over the scalp. This scalp EEG activity results from extracerebral current flow due to summated excitatory postsynaptic potentials (EPSP) and inhibitory postsynaptic potentials (IPSP). Rhythmic oscillations of the thalamocortical neuronal projections cause synchronous synaptic EPSP and IPSP over areas of the cortex [1–3]. It is important to remember that the voltages recorded by scalp EEG are attenuated by the skull and intervening tissues and reflect about one tenth of the voltage recorded directly over the cortical surface.

2 Method of EEG Recording Electrical signals are recorded by the electrodes (secured by paste or conducting gel) and transmitted through electrode wires, which connect the electrodes to the headbox of the

S. Chokroverty HMH JFK University Medical Center and Neuroscience Institute, Hackensack Meridian School of Medicine at Seton Hall University, South Orange and Nutley, NJ, USA Rutgers Robert Wood Johnson University Medical Center, New Brunswick, NJ, USA E. S. Neiman St. Francis Medical Center, Boca Raton, FL, USA S. Bhat (*) Department of Neurology, Hackensack Meridian School of Medicine, Nutley, NJ, USA JFK Neuroscience Institute, Hackensack Meridian-Health JFK University Medical Center, Edison, NJ, USA

polysomnography (PSG) equipment via numbered pins known as jacks. The electrodes used are usually gold cup electrodes with holes in the center and the silver-silver chloride electrodes, which need repeated chloriding for proper maintenance. Positive and negative charges are generated between the scalp and recording electrode because of ionic dissociation. The electrode–electrolyte interface is the most critical link in the PSG machine, as most artifacts originate at this site; careful preparation is therefore very important. The impedance in a pair of electrodes should be measured by an impedance meter and should not be greater than 5000 (5 K) ohms [4, 5]. High impedance impairs the ability of the electrical signal to reach the amplifier and interferes with the capacity of the amplifier to eliminate environmental noise, thus increasing artifacts. All the wires from the head are gathered, tied together into a “ponytail” in the back of the head and secured by wrapping. A shielded conductor cable to the electrode montage selector containing rows of switches in pairs (active and reference electrodes) corresponding to the inputs of the amplifier. The placement of the electrodes is determined by the 10–20 electrode placement system which is recommended by the International Federation of Societies for EEG and Clinical Neurophysiology and was published by Jasper in 1958 [6]. The 10–20 system is based on definable anatomical landmarks (Fig. 1). The system consists of letters denoting regions of the brain underneath the area of the scalp (FP  =  frontopolar, F  =  frontal, C  =  central, P  =  parietal, T = temporal, O = occipital) and numbers denoting specific locations (odd numbers refer to the left side of the head, and the even numbers refer to the right side). Electrode placement is determined by measuring important landmarks including the inion, the nasion, and the right and left preauricular points. The distance from the nasion to inion along the midline through the vertex should be measured. FPz in the midline is 10% above the nasion of the total distance between the inion and nasion. The electrodes marked FP1 and FP2 are located laterally 10% above the nasion of the total distance between the inion and nasion measured along

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. J. Thomas et al. (eds.), Atlas of Sleep Medicine, https://doi.org/10.1007/978-3-031-34625-5_5

69

70

S. Chokroverty et al.

3 Normal Waking and Sleep EEG Rhythms in Adults

Fig. 1  The international 10–20 electrode placement system

the temporal regions through the preauricular points. Oz denotes an electrode placed at a distance of 10% above the inion of the total distance between the nasion and inion. T3 and T4 electrodes are placed in a location 10% above the preauricular points of a total distance between the two preauricular points. The rest of the electrodes are located at a distance of 20% measured from inion to nasion anteroposteriorly or laterally through the ears as well as transversely between the ears. A montage (derivation) refers to an arrangement of two electrodes. Both bipolar montages (connection of the electrodes between two relatively active sites over the scalp) and referential montages (connection of the electrodes between an active and relatively inactive site, e.g., M1, M2, Cz, Pz) are recommended. The nomenclature was recently changed to rename T3, T4, T5 and T6 to T7, T8, P7 and P8, respectively, in a modified 10–20 electrode placement system. The American Academy of Sleep Medicine (AASM) Manual for the Scoring of Sleep and Associated Events [7] recommends three channels; this results in significant limitations. Most important, such a limited montage would result in the inability to capture focal epileptiform activity arising from the temporal lobes, the most common location for such discharges, as well as missing focal slowing from the hemisphere not being recorded. In our own laboratory, we use between 4 and 8 EEG channels including temporal leads from both hemispheres; this increases the yield of capturing focal or diffuse slow waves or epileptiform activities. If possible nocturnal seizures are suspected, an extended EEG montage (Fig. 2), covering the bilateral temporal and parasagittal regions and including both bipolar and referential channels, is recommended.

The dominant rhythm in adults during wakefulness is the alpha rhythm, consisting of 8–13 hertz (Hz) activity distributed synchronously and symmetrically over the parieto-occipital regions (“posterior dominant rhythm,” see Fig.  2). The frequency of the rhythm between the two hemispheres should not vary by more than 1 Hz, and the amplitude should not vary by more than 50%. This alpha rhythm is best seen during quiet wakefulness with eyes closed and is significantly attenuated by eye opening or mental concentration. A small percentage of normal adults show no alpha rhythm, and their EEG is characterized by a dominant rhythm of low-­amplitude fast frequency EEG in the beta frequencies (greater than 13  Hz) with amplitudes varying between 10 and 25 microvolts (Fig.  3), which is a normal finding. In most adults, the beta rhythms are seen predominately in the frontal and central regions intermixed with the posterior alpha rhythms. Certain medications such as benzodiazepines and barbiturates may enhance this activity. There are characteristic changes in the background EEG rhythms as an individual progresses through the three stages of non-rapid eye movement (NREM) sleep and the tonic and phasic stages of REM sleep. EEG in elderly subjects shows a progressive slowing of the alpha frequency during wakefulness and diminution of alpha blocking and photic driving responses. Focal temporal slow waves, particularly in the left temporal region, often called benign temporal delta transients (Fig. 4) of the elderly, are seen in many apparently normal elderly individuals and are sometimes associated with sharp transients. Transient bursts of anteriorly dominant rhythmic delta waves may also be seen in some elderly subjects in the early stage of sleep. Other changes in older adults consist of sleep fragmentation with frequent awakenings, including early morning awakening and multiple sleep stage shifts. Another important finding in the sleep EEG of older adults is the reduction in amplitude of the slow waves. Due to this, many slow waves do not meet AASM scoring guidelines; the percentage of slow-wave sleep is therefore often reduced in these subjects.

4 Abnormal EEG Patterns The importance of an extended EEG montage during PSG covering the parasagittal and temporal regions bilaterally is to document abnormal EEG patterns which may include focal (Fig. 5) or diffuse (Figs. 6 and 7) slowing and epileptiform discharges (see examples below). However, epilepsy is a clinical diagnosis; therefore, history must be obtained

Electroencephalography for the Sleep Specialist

71

a

b

Fig. 2 (a) A 10-s epoch from the polysomnogram of a 32-year-old man referred to the sleep laboratory for possible OSA. He has a history of epilepsy; hence, an extended EEG (“seizure”) montage was requested. Normal awake EEG pattern characterized by symmetrical, sinusoidal posterior dominant alpha rhythm at 9–10  Hz is noted. The top four channels are referential channels connected to the left mastoid. The next 16 electrodes are bipolar channels arranged in a double banana montage as per the International 10–20 electrode placement system. (b) Same subject data as in A viewed at a 30-s epoch. E1-M1 and E2-M1;

electrooculogram (EOG) channels. Masseter1-Masseter2, Chin1-­ Chin2; masseter and submental electromyogram (EMG). EKG; electrocardiogram. HR; heart rate. Int1-Int2; intercostal EMG. LtArm1-LtArm2, RtArm1-RtArm2; left and right biceps EMG; LTIB, RTIB; left and right tibialis anterior EMG. OroNs1-OroNS2; oronasal airflow. PFlow; nasal pressure transducer. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel. EKG artifact is noted in the referential leads and EOG leads

72

S. Chokroverty et al.

a

b

Fig. 3 (a) Normal awake EEG pattern characterized by diffuse low-voltage beta (greater than 13 Hz) rhythm in a 23-year-old man at a conventional EEG paper speed of 10-s per page. (b) Same subject data as in a viewed at a 30-s epoch

Electroencephalography for the Sleep Specialist

73

a

b

Fig. 4 (a) A 10-s EEG shows transient burst of delta activity in the left temporal, and occasionally also in the right temporal, region in a 94-year-­ old woman with history of syncope. (b) Same subject data as in a viewed at a 30-s epoch

74

S. Chokroverty et al.

a

b

Fig. 5 (a) A 10-s EEG epoch shows focal slowing at 3–6 Hz over the left temporoparietal region in an 85-year-old man with history of a stroke. Normal alpha rhythm at 10–11 Hz is noted on the right side. (b) Same subject data as in a viewed at a 30-s epoch

Electroencephalography for the Sleep Specialist

75

a

b

Fig. 6 (a) A 10-s EEG epoch shows mild diffuse slowing of the background rhythm, dominated by 6- to 7-Hz theta activity in an 80-year-old woman with history of hypertension, diabetes mellitus, and bacteremia. (b) Same subject data as in A viewed at a 30-s epoch. Additionally, brief bursts of rhythmic muscle artifact are noted on the left side toward the end of the epoch. Mild diffuse slowing of the background rhythm may be missed during conventional polysomnography scoring. The clue is in that the epoch looks like an awake recording without the characteristic alpha rhythm. This necessitates viewing the segment as a 10-s epoch

76

S. Chokroverty et al.

a

b

Fig. 7 (a) A 27-year-old man with a 3-week history of sudden severe headaches diagnosed as pituitary adenoma. 10 s EEG epoch shows evidence of a brief burst of frontal intermittent rhythmic delta activity (FIRDA), synchronously recorded over both hemispheres. FIRDA was originally believed to arise from deep midline pathology, which is demonstrated in this case. However, FIRDA is now believed to be anatomically non-­ localizing and indicates diffuse cerebral dysfunction secondary to metabolic or structural involvement. (b) Same subject data as in a viewed at a 30-s epoch

Electroencephalography for the Sleep Specialist

from patients, relatives, witnesses, or medical personnel if a patient is referred to the sleep laboratory with a suspected diagnosis of nocturnal seizures. The history must include descriptions of ictal, preictal, postictal, and interictal phenomena, family and drug histories, as well as history of any significant medical or surgical illnesses that might be responsible for triggering the seizures. Physical examination must be conducted to find any evidence of neurological or other medical disorders before PSG recording in the laboratory.

4.1 EEG Signs of Epilepsy EEG is the single most important diagnostic laboratory test for patients with suspected seizure disorders. Certain characteristic EEG waveforms correlate with a high percentage of patients with clinical seizures and therefore can be considered of potentially epileptogenic significance. These epileptiform patterns consist of spikes, sharp waves, spike and waves, sharp and slow-wave complexes, as well as evolving pattern of rhythmic focal activities, particularly in neonatal seizures. In addition, a pattern that correlates highly with complex partial seizure is the temporal intermittent rhythmic delta activity (TIRDA) (Fig. 8). Another pattern that is considered a marker of the seizure onset zone is interictal scalp high frequency oscillations (HFOs) consisting of gamma frequency activity (30–80  Hz), ripples (80–250  Hz), and fast ripples (250–1000  Hz) [8], recorded noninvasively using

77

amplifiers with appropriate filter settings (Fig. 9). Therefore, HFOs are considered significant biomarkers for epileptogenic zones (EZ). However, it remains unclear how to differentiate pathological HFOs (pHFOs) from physiological HFOs and other artifacts. Using high-density EEG and electrophysiological source imaging it has been shown that pHFOs by showing consistent concurrence of morphologically distinct epileptiform spikes are prominent biomarkers of EZ and sleep-onset zone in presurgical evaluation of patients with epilepsy [9]. It is important for the polysomnographer to be able to distinguish a true epileptiform discharge from sharply contoured normal background activity, such as wicket rhythms. Recognition of a spike or a sharp wave depends on the presence of certain characteristics (Fig. 10). A spike is defined as a waveform which [10]: • Is generally biphasic or triphasic-appearing • Disrupts the background rhythm • Has a sharp ascending limb followed by a slow descending limb • Has a brief duration of 20–70 ms • Shows a field of distribution in neighboring channels (on a bipolar montage, the area of greatest cortical negativity can be identified by the channel that shows a phase-­ reversal; on a referential montage, it can be identified by the channel where the waveform shows the highest amplitude) (Fig. 11) • Is often followed by an after-going slow wave

Fig. 8  10-s EEG epoch shows left temporal intermittent rhythmic delta activity (TIRDA) during wakefulness in a 30-year-old woman with focal-­ onset impaired awareness seizure (previously known as partial complex seizures). TIRDA correlates highly with history of seizures

78

S. Chokroverty et al.

a

b

Fig. 9  (1) Short EMG burst, which could be mistaken for ripples, but are actually artifactual oscillations. (2) Ripples co-occurring with sharp wave. (a) Raw EEG. (b) EEG filtered with high-pass filter of 80 Hz. Gray section in A is expanded in time and amplitude in b. Note that for this figure the calibration is different in the left and right part of the figure but is the same for the top and bottom parts. Ripple oscillations

are underlined. The waveform morphology of non-artifactual fast oscillations is more rhythmic and regular in amplitude and frequency than artifactual oscillations. (Reproduced from Andrade-Valencia et al. [8]; with permission). See also recent controversy about pathological HFOs and their differentiation from physiological HFOs and artifacts described in the text

In contrast, augmented sharply contoured background rhythms are generally monophasic, have uniform ascending and descending limbs and arise as part of the background rhythm rather than disrupting it. Generally, spikes or sharp waves are surface negative and have an amplitude at least 30% higher than the background activity. A sharp wave fulfills all the criteria described for a spike except that the duration is 70–200 ms. Spike and waves, and sharp- and slow-wave complexes may be isolated or may repeat in trains for

several seconds or longer (Fig. 12), and in some patients these repeat in a rhythmic manner, such as may be seen in patients with absence seizure showing 3-Hz spike and wave complexes in the EEG (Fig.  13). Patients with tonic-clonic seizure may show a sudden burst of spike and slow waves beginning at a rate of 4–6  Hz and then gradually slowing down and stopping before the postictal period. These discharges are seen in a bilaterally symmetrical and synchronous fashion. Postictal slow waves are followed by gradual recovery to the preictal normal

Electroencephalography for the Sleep Specialist

79

a

b

Fig. 10  Diagram of (a) an epileptic spike and (b) an augmented background sharp rhythm. (Reproduced from Ajmone-Marsan [10]; with permission)

background rhythm after several hours. Interictal EEG may show generalized synchronous and symmetrical 4to 6-Hz spike- and slow-wave discharges or multiple spike and waves. An EEG showing all the characteristic features of epileptiform activity and accompanied by the behavioral correlates simultaneously during the recording in the laboratory con-

Fig. 11  Schematic representation of an area of cortical negativity at the T4 electrode in a bipolar (left) and referential (right) montage. On the bipolar montage, there is a phase reversal at T4, the area of maximal cortical negativity, noted in channels F8-T4 and T4-T6. Note that adjacent channels (FP2-F8 and T6-O2) also show defections predicted by the relative negativity of the individual electrodes, thereby producing a “field.” The referential montage has been created by connecting electrodes Fp2, F8, T4, T6, and O2 to the ipsilateral mastoid (M2). All channels demonstrate negative (upward) deflections, with the waveform being of highest amplitude in the channel containing the electrode with the highest degree of cortical negativity (T4-M2). Adjacent channels demonstrate negative deflections of lower amplitudes, thereby producing a “field”

firms the diagnosis of epilepsy. However, it is rare to observe the occurrence of a clinical seizure during an EEG recording (Fig. 14). Therefore, a definitive statement about the diagnosis of epilepsy in a particular patient cannot be made even in the presence of the characteristic EEG signs of epilepsy. In patients suspected of seizure disorder, it is advantageous to include activation procedures such as sleep, hyperventilation, and photic stimulation during routine EEG recording in the daytime in order to bring out the ictal or interictal epileptiform patterns. Special basal temporal electrodes (T1, T2, and sphenoidal electrodes) should be used in addition to the routine electrode placement in patients suspected of partial complex seizure.

80

S. Chokroverty et al.

a

b

Fig. 12 (a) A 10-s epoch of daytime EEG from a 50-year-old man with a history of generalized tonic-clonic (grand mal) seizure disorder showing frontally predominant, symmetric, synchronous paroxysmal interictal bursts of 5 Hz spike-wave discharges occurring during drowsiness.

Top 20 channels: EEG with channels applied per International Nomenclature viewed in a Double Banana montage (sensitivity 7 microvolts/mm). (b) The same epoch displayed at a 30-s speed (sensitivity 15 microvolts/mm)

Electroencephalography for the Sleep Specialist

81

a

b

Fig. 13  Polysomnogram (PSG) epochs from an eight-year-old girl referred to the sleep laboratory with complaints of insomnia, frequent nocturnal awakenings, and declining school performance. She did not have a history of convulsions or a previous diagnosis of epilepsy, but the family reported episodes of “daydreaming” and staring. Her neurological examination was normal with no history of developmental delay. The PSG showed an apnea-hypopnea index of 1/h, mainly due to physiological central apneas. However, the EEG channels showed frequent bursts of frontally predominant, generalized 3 Hz spike and wave discharges in drowsiness and NREM sleep, less well seen in REM sleep. (a) A 10-s epoch showing these discharges. (b) same epoch as viewed at a 30 s speed. Top eight channels; EEG recording with electrodes placed according to the 10–20 international electrode placement system. E1-M1 and E2-M1; electrooculogram channels. EKG; electrocardiogram. HR; heart rate. Chin1-­ Chin2; submental electromyogram (EMG). LTIB, RTIB; left and right tibialis anterior EMG.  LGAST, RGAST; left and right gastrocnemius EMG. OroNs1-OroNS2; oronasal airflow. Chest and ABD; effort belts. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel

82

S. Chokroverty et al.

a

b

Fig. 14 (a) A 61-year-old woman with a history of insomnia and one episode of screaming at night. The EEG shows four sequential epochs at 15 s per page at sensitivities ranging from 20 microvolts to 50 microvolts per millimeter. The figure shows onset of left temporofrontal rhythmic spike and wave activity during photic stimulation at 17 Hz. The spike and wave activity is incremental in nature with increasing frequency and amplitude noted. (b) The figure shows bilateral generalized spread of the spike and wave activity now being contaminated by muscle activity. (c) The figure shows generalized spike and wave activity contaminated by rhythmic generalized muscle activity. (d) The figure now shows continuation of generalized spike and wave activity contaminated by muscle activity followed by postictal EEG depression of the amplitudes and slowing of the background. The EEG development of the ictus was clinically accompanied by a loud scream and sudden turning of the head to the right, followed by shaking of both arms, frothing at the mouth, and fixed and dilated pupils. The entire episode lasted for about 60 s. The patient was in a postictal state for approximately 10 min. (e–g) The same subject data as in a through d viewed over three 30-s epochs

Electroencephalography for the Sleep Specialist

c

d

Fig. 14 (continued)

83

84

e

f

Fig. 14 (continued)

S. Chokroverty et al.

Electroencephalography for the Sleep Specialist

85

g

Fig. 5.14 (continued)

5 Nonepileptiform Patterns Mimicking Epileptiform Discharges There are several non-epileptiform EEG patterns that may mimic epileptiform discharges but are not of epileptogenic significance. It is crucial that the polysomnographer be familiar with such normal variants to avoid an erroneous report of epileptiform activity, which may lead to unnecessary further testing and unwarranted medication administration. Some commonly seen normal variants are described below.

5.1 Sharp and Slow Artifacts Sharp artifacts include activity that is generally easy to identify from its distribution, temporal relationship to other phenomena, and morphologic appearance. Electrocardiogram (EKG) artifacts (Fig. 13) can be identified by correlation to a one-lead EKG.  Unlike true epileptiform activity, electrode pops are limited to one channel (Fig. 15) and are eliminated after the electrode problems are resolved. Muscle artifacts can be reduced by having the patient relax and, where necessary, reducing the high frequency filter settings.

86

S. Chokroverty et al.

Fig. 15  10-s EEG epoch. Pulse artifact is noted to occur in the right frontal EEG electrode (F4). This artifact occurs as a result of placement of electrodes over scalp arteries. Note the clear time-locked relationship

between the EKG and the artifact. The recognition of this relationship helps distinguish the artifact from focal slowing

Slow artifacts include sweat artifact, movement artifact, and respiratory artifact. Pulse artifact (Fig. 15) occurs due to placement of an electrode over a major scalp artery and is time-locked to EKG; it is generally seen in temporal or frontal channels.

active or passive movements of the contralateral limbs or even an intention to move the limbs. Similar rhythms located in the temporal regions are called wicket rhythms (Fig. 24), and these have no specific significance but may sometimes be mistaken for epileptiform discharges (especially when the rhythm occurs as a fragment, so-called “wicket spike”). The Ciganek rhythm is a 4–7  Hz, midline, somewhat sharply contoured waveform seen in drowsiness and early NREM sleep (Fig.  25), but occasionally in REM sleep as well (Fig. 26). Its field is generally maximal at Cz, and while it may have a waxing and waning amplitude, it generally does not evolve or spread to other locations. While in the past it was thought to have epileptogenic significance, it is now considered a normal variant. Lambda waves (Fig. 27), also known as “scanning waves,” are surface-positive, occipitally predominant waveforms that resemble POSTs but occur in wakefulness and are associated with reading and scanning.

5.1.1 Sharp Transients Seen During Sleep and Wakefulness Sharp transients seen during sleep and wakefulness may include positive occipital sharp transients (POSTs) (Fig. 16), mu rhythms (Fig. 17) posterior slow waves of youth (Fig. 18), particularly sharp appearing vertex sharp waves (Fig.  19) and K -complexes (Fig.  20), hypnagogic or hypnopompic hypersynchrony resembling spike-wave complexes (Fig. 21), benign epileptiform transients of sleep (BETS) (Fig.  22), photic driving responses that may resemble spikes or sharp waves (Fig. 23), and augmentation of ongoing background rhythm assuming a sharp or spike-like appearance (Fig. 24). These sharp transients can be differentiated from true epileptiform activity by studying their distribution and morphology, the underlying background, and the state of the patient’s alertness. Vertex sharp waves, sharp-appearing K complexes, spindles, and POSTs are stage dependent and have characteristic distributions. In addition, they tend to occur in younger subjects. Hypnagogic hypersynchrony and posterior slow waves of youth are seen in children and adolescents. Mu rhythms (Fig.  17) consist of brief bursts of 7–11  Hz activity over the central regions, which show attenuation on

5.1.2 Sharp Transients of Uncertain Significance These transients include small sharp spikes (Fig.  22) also known as benign epileptiform transients of sleep (BETS), 6-Hz spike and waves or phantom spike-waves (Fig. 28), and psychomotor variant pattern or rhythmic midtemporal discharges (Fig. 29), which occur in drowsiness and light sleep and have characteristic square-topped appearance that may occasionally be sharply contoured and a source of confusion with epileptiform activity.

Electroencephalography for the Sleep Specialist

87

a

b

Fig. 16 (a) A 30-s epoch from nocturnal polysomnography recording. Positive occipital sharp transients of sleep are seen bilaterally in posterior temporal and occipital electrodes recorded during stage II non–rapid eye movement sleep. (b) Same subject data as in a viewed at a 10-s epoch

88

S. Chokroverty et al.

a

b

Fig. 17 (a) A 40-year-old man referred for evaluation of difficulty sleeping at night, snoring, and excessive daytime sleepiness. The epoch is taken from multiple sleep latency test in stage I non–rapid eye move-

ment sleep. Mu rhythm is seen bilaterally over the central regions (C3-­ C4 electrodes), but it is better defined on the left than on the right central regions. (b) Same subject data as in a viewed at a 30-s epoch

Electroencephalography for the Sleep Specialist

89

Fig. 18  A 15-year-old female patient with history of new onset of seizures. The EEG in wakefulness shows slow waves of youth (underlined in red) at 2–3 Hz superimposed intermittently on occipital alpha rhythm

at 9–10  Hz bilaterally. This is a normal finding between ages 2 and 21 years of age, most commonly occurring between 8 and 14 years of age and attenuates with eye opening

5.2 Epileptiform-Like Patterns Without Epileptogenic Significance

epileptiform discharge of adults, and burst suppression patterns (Fig. 33). Triphasic waves have an initial positive configuration and characteristic distribution. These are seen synchronously and symmetrically with frontal dominance of the amplitude with anteroposterior phase shift (“A-P lag”). Triphasic waves are not of epileptiform significance but may be seen in metabolic or toxic encephalopathies (e.g., hepatic, renal, or respiratory failure). Sometimes, these are seen in anoxic encephalopathies and advanced neurodegenerative processes also.

These patterns include the following: Triphasic waves (Fig.  30), lateralized periodic discharges (LPDs) (Fig.  31), generalized periodic discharges (GPDs) which may occur intermittently (Fig.  32) (LPDs and GPDs were formerly known as periodic lateralized epileptiform discharges [PLEDs] and generalized periodic epileptiform discharges [GPEDs] but recently renamed as the pattern is only potentially, not definitively, epileptogenic), subclinical rhythmic

90

S. Chokroverty et al.

a

b

Fig. 19 (a) This 10-s epoch shows vertex sharp waves seen as negative potentials with a wide distribution around the vertex. In this segment, V waves are seen at C3, C4, and CZ with spread of activities to F3, F4, P3,

P4, PZ, and FZ. (b) Same subject as in a with data viewed over a 30-s epoch

Electroencephalography for the Sleep Specialist

91

a

b

Fig. 20 (a) The tracing shows the presence of a sharp appearing K complex in the middle of a 10-s epoch characterized by a high-voltage negative-­positive potential followed by sleep spindles. (b) Same subject as in a with data viewed over a 30-s epoch with several K complexes

92

S. Chokroverty et al.

a

b

Fig. 21 (a) This 10-s EEG epoch was recorded in a 9-year-old girl during stage II non–rapid eye movement sleep showing hypnagogic hypersynchrony characterized by a brief burst of synchronous high-voltage

slow activity at 4- to 5-Hz activity. (b) Same subject data as in a viewed over a 30-s epoch. At end of the epoch, a sleep spindle burst is seen in all channels referred to the CZ electrode

Electroencephalography for the Sleep Specialist

93

a

b

Fig. 22 (a) Benign epileptiform transients of sleep or small sharp spikes (SSS) are seen bilaterally, more prominent on the left side in channels 1 to 5 and 9 to 11 in a 67-year-old man. SSS are best recorded

in ipsilateral ear montage during stage I non–rapid eye movement sleep as shown in this epoch. SSS may be seen asynchronously on either side as well. (b) Same subject as in a viewed over a 30-s epoch

94

S. Chokroverty et al.

Fig. 23  Photic driving responses seen diffusely at 7 Hz resembling spike and wave bursts. (From Chokroverty [14]; with permission) Fig. 24  Several spike-like waves (wicket spikes) repeating rhythmically at T3-T5 (arrows) as well as in T4 with reduced amplitude in an adult without any history of seizures. (From Chokroverty [14]; with permission)

Electroencephalography for the Sleep Specialist

Fig. 25  10-s (left) and 30-s epochs of Stage N1 sleep from the overnight PSG of a 35-year-old woman referred to the sleep laboratory for abnormal movements in sleep, later diagnosed to be intensified hypnic jerks. Note the occurrence of a burst of 6–7 Hz theta range slowing in the bilateral parasagittal leads (F3-M1, C3-M1; F4-M2, C4-M2), not evident in the bitemporal leads (F7-T3, T3-T5, T5-O1; F8-T4, T4-T6, T6-O2). This rhythm, known as a midline theta of drowsiness or Ciganek’s rhythm, is a variant usually occurring in drowsiness or light sleep. It is considered to be benign but may be mistaken for epilepti-

95

form discharges. Chin1-Chin2; submental electromyogram (EMG). RT Stma1; right sternomastoideus EMG.  RT Bicep, LT Bicep; right and left biceps brachii EMG. RLRA; right lateral rectus EMG. RT Quad, LT Quad; right and left quadratus femoris EMG. RTIB, LTIB; left and right tibialis anterior EMG. LGAST, RGAST; left and right gastrocnemius EMG. EKG; electrocardiogram. HR; heart rate. OroNasal; oronasal airflow. Pflow; nasal pressure transducer recording. SaO2; arterial oxygen saturation by finger oximetry. Also included is a snore channel. (From Bhat and Chokroverty [1]; with permission)

96

Fig. 26  A 30-s epoch from a PSG segment in an adult man referred to the sleep laboratory for abnormal movements in sleep. The EEG channels show bilaterally symmetric and synchronous centrally predominant 4 Hz paroxysmal bursts of midline theta activity (best seen in the C3-A2 and C4-A1 derivations) in REM sleep (tonic phase just before phasic eye movements and transient muscle bursts) consistent with Ciganek’s rhythm. It is of uncertain clinical significance but thought to be a benign pattern, although it may be mistaken for epileptiform spikes or sharp waves. Top ten channels; EEG recording with electrodes

S. Chokroverty et al.

placed according to the 10–20 international electrode placement system. Left and right EOG: electrooculogram. Lt/Rt Tib EMG: electromyograms of left and right tibialis anterior muscles connected together in a single channel. Oronasal: airflow thermistor. Chest and abdomen: respiratory effort channels. Snoring: snore channel. EKG: electrocardiogram (note artifact in this channel). SaO2: arterial oxygen saturation by finger oximetry. Also recorded is heart rate. Sensitivity 7 microvolts/ mm; high frequency filter 15 Hz; low frequency filter 1 Hz

Electroencephalography for the Sleep Specialist

Fig. 27  10 s EEG epoch showing lambda waves (in red box) recorded from a 16-year-old female with a history of a syncopal spell. She was scanning ceiling tiles at the time of the recording. Note the surface posi-

Fig. 28  Six-hertz phantom spike and wave pattern seen in the last four channels. (From Bhat and Chokroverty [1]; with permission)

97

tive, occipitally predominant activity occurring simultaneously with rapid eye movements recorded in the frontal channels

98

S. Chokroverty et al.

a

b

Fig. 29 (a) A 40-year-old man referred for evaluation of difficulty sleeping at night, snoring, and excessive daytime sleepiness. This 10-s epoch is taken from nocturnal polysomnography in stage I non–rapid eye movement sleep. Rhythmic mid-temporal theta of drowsiness

(RMTD) is seen as a brief burst of rhythmic theta activity at 6–7 Hz recorded asynchronously over left and right temporal electrodes. (b) Same subject data as in a viewed over 30-s epoch. It shows two brief bursts of RMTD, one at each end of the epoch

Electroencephalography for the Sleep Specialist

99

a

b

Fig. 30 (a) 10-s EEG epoch showing triphasic waves at 1–2  Hz recorded synchronously, with maximal amplitude frontally in a 67-year-old woman with a history of end-stage renal disease and hepatitis. (b) Same subject data as in a viewed over a 30-s epoch

100

S. Chokroverty et al.

a

b

Fig. 31 (a) A 49-year-old woman with a history of right frontal subdural hematoma, ovarian cancer, colon cancer, and seizures. Lateralized periodic discharges (LPDs, formerly known as periodic lateralizing epileptiform discharges [PLEDs]) are seen in this 10-s EEG epoch, charac-

terized by spike and wave activity recurring at 1.5–2 s over the right central region and spreading to the right temporal region. The background shows diffuse slowing. (b) Same subject data as in a viewed over 30-s epochs

Electroencephalography for the Sleep Specialist

101

a

b

Fig. 32 (a) An 81-year-old man with history of dementia and transient ischemic attacks. 10-s EEG epoch shows the presence of generalized, intermittent, periodic or pseudo-periodic sharp- and slow-wave dis-

charges at 1–2 Hz, representing periodic lateralized discharges (GPDs, formerly known as generalized periodic discharges [GPEDs]). (b) Same subject data as in a viewed at a 30-s epoch

102

S. Chokroverty et al.

a

b

Fig. 33 (a) The patient is a 6-month-old girl with a history of hypoxic ischemic encephalopathy and recurrent seizures. The recording shows a burst suppression pattern characterized by bursts of generalized spike

and wave and multi-spike and wave discharges lasting 2–3 s with intervening periods of relative voltage suppression (quiescence). (b) Same subject data as in a viewed at a 30-s epoch

Electroencephalography for the Sleep Specialist

103

6 Epileptiform Patterns in Sleep Patients with focal or generalized seizures may sometimes present only during sleep. On the one hand, such nocturnal seizures may be mistaken for other motor disorders during sleep such as parasomnias. On the other hand, patients with a variety of sleep-related movement disorders (severe sleep apnea, periodic limb movements of sleep, hypnic jerks, confusional arousals, night terrors) are misdiagnosed as having nocturnal seizures. It is, therefore, important for the sleep specialists to be familiar with the clinical and EEG patterns seen in primary generalized seizure (e.g., absence spells or generalized tonic-clonic seizures) and partial complex seizures of temporal or extratemporal (frontal) origin. Some patients with generalized tonic-clonic and partial complex seizure may have predominantly nocturnal seizures. It is well known that sleep deprivation can lower seizure threshold; indeed, sleep-deprived EEGs increase the yield of identifying epileptiform activity. Tonic seizures are typically activated by sleep, occur frequently during NREM sleep, and are rarely seen during REM sleep. Additionally, certain types of seizures are characteristically observed during sleep. Benign focal epilepsy of childhood with centrotemporal spikes (BECTS, also known as benign rolandic epilepsy) is characterized by the presence of centrotemporal or rolandic spikes or sharp waves (Fig. 34). Seizures generally resolve by the age of 15–20 years without any neurologic sequelae. Juvenile myoclonic epilepsy (Fig. 35) generally presents in late childhood and early adolescence with early morning myoclonic jerking or tonic-clonic seizures; the EEG shows synchronous and symmetrical polyspike-wave discharges, often precipitated by photic stimulation (photoparoxysmal or

Fig. 34  EEG showing left centrotemporal spikes and sharp waves in a patient with benign rolandic epilepsy (sylvian seizures). Benign rolandic seizures commonly present in childhood as unilateral tonic or clonic seizures in face or arm, speech arrest, and paresthesias in the mouth or tongue. (Reproduced from Chokroverty and Nobili [13]; with permission)

104

S. Chokroverty et al.

Fig. 35  Interictal generalized multiple spike and wave discharges in the EEG of a patient with juvenile myoclonic epilepsy. Note the recording at (a) 30 mm/s (10-s epoch, conventional EEG speed) and (b) at 10 mm/s (30-s epoch, conventional PSG speed). (Reproduced from Chokroverty and Nobili [13]; with permission)

photoconvulsive response). With continuous spike and wave in slow wave sleep (CSWS) [11, 12], which may be misdiagnosed as autism, there is developmental and language regression, and the EEG consists of generalized 2- to 2.5-Hz spike

and wave discharges occupying 85% of NREM sleep, suppressed during REM sleep (Fig.  36); the EEG may show minimal findings in wakefulness and illustrates the importance of capturing sleep even on daytime EEGs.

Electroencephalography for the Sleep Specialist

105

a

b

Fig. 36  EEG tracings from a 9-year-old boy with developmental delay and poor speech output, and nocturnal events of uncertain etiology. (a) 30 mm/s epoch of wakefulness showing eye blinks in the bilateral frontopolar leads with no evidence of epileptiform activity. (b) 30  mm/s EEG epoch of stage N3 sleep. The patient had 2–3.5  Hz continuous generalized spike and wave discharges with 1–2  s of electrocerebral

decrement (continuous spike and wave in slow wave sleep, CSWS). (c) Same epoch as b, displayed at a speed of 10 mm/s (the usual speed at which PSGs are reviewed). (d) 30 mm/s EEG epoch of rapid eye movement (REM) sleep in the same patient showing absence of epileptiform activity. (From Neiman et al. [12]; with permission)

106

c

d

Fig. 36 (continued)

S. Chokroverty et al.

Electroencephalography for the Sleep Specialist

7 Diagnostic Approach to Patients with Suspected Nocturnal Seizures In cases of suspected nocturnal seizures, a routine daytime EEG recording with activation procedures (hyperventilation and photic stimulation), and attempts to capture drowsiness and sleep, is the initial recommended diagnostic step [13]. If no epileptiform discharges are noted, then the EEG should be repeated with partial or total sleep deprivation [14]. If this is also nondiagnostic, an overnight video-PSG study should be obtained for electroclinical correlation. An appropriately devised seizure montage with full complement of electrodes (see above) or special electrode placements (e.g., T1 and T2, sphenoidal electrodes, and nasopharyngeal electrodes) should be used. During interpretation, the PSG montage can be reformatted to analyze any abnormal behavior or suspected clinical seizures at 30 mm/s (the conventional EEG speed) instead of the usual 10 mm/s (the conventional PSG speed); this makes it easier to recognize abnormal EEG patterns.

References 1. Bhat S, Chokroverty S.  Electroencephalography, electromyography, and electro-oculography: general principles and basic technology. In: Chokroverty S, editor. Sleep disorders medicine: basic science, technical considerations, and clinical aspects. 4th ed. New York, NY: Springer; 2017. 2. Butkov N, Keenan SA.  An overview of polysomnographic technique. In: Chokroverty S, editor. Sleep disorders medicine: basic

107 science, technical considerations, and clinical aspects. 4th ed. New York, NY: Springer; 2017. 3. Schomer DL, Da Silva L, editors. Niedermeyer’s electroencephalography: basic principles, clinical science, and related fields. 7th ed. New York, NY: Oxford University Press; 2018. 4. Ebersole J, Pedley T.  Current practice of clinical electroencephalography, vol. 10. 3rd ed. Philadelphia, PA: Lippincott, Williams and Wilkins; 2003. p. 604–5. 5. Libenson MH.  Practical approach to electroencephalography. Philadelphia, PA: Saunders/Elsevier; 2010. 6. Jasper HH.  The ten-twenty electrode system of the international federation. Electroencephalogr Clin Neurophysiol. 1958;10:371–5. 7. Berry RB, Brooks R, Gamaldo CE, et al. The AASM manual for the scoring of sleep and associated events: rules, terminology, and technical specifications, version 2.5. Darien, IL: American Academy of Sleep Medicine; 2018. 8. Andrade-Valencia LP, Dubeau F, Mari F, et  al. Interictal scalp fast oscillations as a marker of the seizure onset zone. Neurology. 2011;77:524–31. 9. Cai Z, Sohrabpour A, Jiang H, Ye S, Joseph B, Brinkmann BH, Worrell GA, He B. Noninvasive high-frequency oscillations riding spikes delineates epileptogenic sources. Proc Natl Acad Sci U S A. 2021;118(17):e2011130118. 10. Ajmone-Marsan C. Encephalographic studies in seizure disorders: additional considerations. J Clin Neurophysiol. 1984;1:143–58. 11. Nickels K, Wirrell E. Electrical status epilepticus in sleep. Semin Pediatr Neurol. 2008;15:50–60. 12. Neiman ES, Seyffert M, Richards A, Gupta D, Chokroverty S.  Epilepsy with continuous spikes and waves during slow wave sleep in a child diagnosed with pervasive developmental disorder-­ not otherwise specified. Sleep Med. 2010;11(8):799–802. 13. Chokroverty S, Nobili L.  Sleep and epilepsy. In: Chokroverty S, editor. Sleep disorders medicine: basic science, technical considerations, and clinical aspects. 4th ed. New York, NY: Springer; 2017. 14. Chokroverty S.  Role of electroencephalography in epilepsy. In: Chokroverty S, editor. Management of Epilepsy. Boston, MA: Butterworth-Heinemann; 1996. p. 67–112.

Odds Ratio Product (ORP): Description and Implications to Understanding and Management of Sleep Disorders Magdy Younes

There is substantial evidence that short sleep duration and circadian misalignment of sleep contribute to poor health outcomes [1, 2]. By contrast, evidence that sleep depth is relevant to health outcomes has been little explored. The main difficulty has been the lack of a practical and reliable method for measuring sleep depth, particularly since sleep depth changes within and between sleep cycles [3, 4]. Apart from addressing the issue of the impact of sleep depth on health outcomes, measurement of sleep depth could help us understand the underlying mechanisms of sleep symptoms and, by extension, improve the therapy of sleep disorders. Sleep depth reflects the difficulty of being aroused from sleep or the arousal threshold. The tradition of measuring the arousal threshold is to measure the intensity of a specified stimulus, most commonly acoustic stimuli [5] or negative pharyngeal pressure [6], required to induce a cortical arousal or awakening. This approach is not practical for measuring the instantaneous arousal threshold since it requires multiple stimulus applications of different intensities over a relatively long time to determine one arousal threshold value. During this time, sleep depth would have varied considerably [3, 4], and the result would at best be an average threshold over the duration of the measurement. Additional difficulties include the disrupting effect of the stimuli on sleep depth [5], the fact that sleep depth estimates may be specific to the stimulus used, and the fact that the stimuli will tend to cause arousals more readily when the arousal threshold is low, therefore resulting in underestimation of the average arousability of the subject across the night. EEG delta power is often used as a measure of sleep depth. However, until recently, not much was known about the relation between delta power and sleep depth, as defined by ease of being aroused. Berry et al. determined the relation

M. Younes (*) Sleep Disorders Centre, University of Manitoba, Winnipeg, MB, Canada

between delta power and the magnitude of negative pharyngeal pressure just before arousal in patients with OSA [4]. The correlation coefficient within participants was low (average r = 0.47 ± 0.03). More recently, the relation between delta power and arousability was determined using the approach for determining arousability at different levels of odds ratio product (ORP) (see next paragraph) [7]. Delta power was found to be sensitive to arousability only up to 300  μV2, which represents about one tenth of the full range of delta power in PSGs [7]. Beyond this range, further increases in delta power as the frequency of the large delta waves increases are not associated with further reduction in arousability. The odds ratio product (ORP) is a recently introduced measure derived from the relation of EEG power in different frequencies to each other [8]. ORP is a continuous metric ranging from 0 (very deep sleep) to 2.5 (full wakefulness). Considerable evidence supports its use as an index of sleep depth. Qualitatively, ORP varies predictably with conventional sleep stages [8–11] and decreases progressively as EEG transitions from full wakefulness to sleep onset and from light to deep stage N2 [8]. It also increases following sleep deprivation [12] and restriction [7] and decreases across the night in response to declining homeostatic drive [13]. Importantly, it transiently increases in a graded manner in response to noise stimuli of different intensities [14]. To determine ORP’s quantitative relation with ease of arousal, and to avoid the pitfalls of conventional methods of measuring the arousal threshold (see above), ease of arousal (arousability) was assessed by the relation between ORP in any given 30-s epoch and probability of a spontaneous cortical arousal (as defined by AASM [15]) or awakening occurring within the next 30 s [8]. This approach takes advantage of the various natural arousal stimuli, of different intensities, impinging continuously on the brain. Thus, in deep sleep only the strongest natural stimuli would be able to induce arousal while in very light sleep stimuli of all intensities would do so. As a result, the frequency of new spontaneous arousals should reflect arousability at the time ORP is measured [8]. Thus, this approach

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 R. J. Thomas et al. (eds.), Atlas of Sleep Medicine, https://doi.org/10.1007/978-3-031-34625-5_6

109

110

M. Younes

can be used to determine arousability at all levels of ORP and the results reflect ­arousability to all types of stimuli arising from the patient’s body or his environment. When so measured, the relation between the current ORP and the likelihood of an arousal/awakening spontaneously in the next epoch was almost perfect (r2 = 0.98) [8].

1 Why Do We Need a Continuous New Index of Sleep Depth? Sleep depth covers a continuous spectrum that does not change in a stepwise fashion as implied by the Rechtschaffen and Kales rules currently used (Figs. 1 and 2).

Why Do We Need A Continuous New Index of Sleep Depth? 1) R&K Metrics of Sleep Depth are Seriously flawed

a

Different Levels of Stage Wake

ORP

-60µV

2.50

C3/M2

b

Different Sleep Depth in Stage N2

-60µV

ORP

C3/M2

1.80

60 µV

60µV

C3/M2

1.55

C3/M2

1.17

C3/M2

0.91

C3/M2

0.63

2.25

C3/M2

1.97

C3/M2

Micro sleep 1.69

C3/M2

5 Seconds From: Younes M, Azarbarzin A, Reid M, Mazzotti DR, Redline S. Sleep. 2021;44(10):zsab145.

Fig. 1  Rechtschaffen and Kales rules assume that stage wake is a uniform stage. Panel a shows that the EEG pattern varies widely as the state progresses from full wakefulness, with predominant beta/alpha activity (top tracing) to epochs with microsleep sections that do not meet the required 15 s to score sleep. Note that ORP reflects these progressive changes in stage wake. Differences in ORP during stage wake are associated with sleepiness [16]. Likewise, R&K rules do not dif-

5 Seconds From: Younes M, Azarbarzin A, Reid M, Mazzotti DR, Redline S. Sleep. 2021;44(10):zsab145.

ferentiate different levels of depth in stage 2. Panel b shows that the EEG pattern in stage N2 can range from one that is a little different from stage N1, or even wake, except for the presence of spindles (top tracing), to a pattern that is much like stage N3 except for the fact that the total duration of delta waves does not add up to 6 s. Again, ORP reflects these different levels. (From Younes et  al. [17]; with permission)

Why Do We Need A Continuous New Index of Sleep Depth? a)

Fraction Awake within REM Periods

1) R&K Metrics of Sleep Depth are Seriously flawed Different Levels of Stage REM

SPO 2 Stage

REM

REM

ORP

CPAP

b) More REM Fragmentation rSHHS1=0.27; p2.25 they almost invariably agree the epoch is wake (black zones). As ORP increases beyond 0.75, disagreements (gray zones) become more frequent. Disagreement decreases beyond ORP of 1.75 such that between 1.75 and 2.00 at least one scorer scores the epoch as wake 90% of the time. These findings can be generalized as follows: Epochs with ORP 1.75 represent wake epochs with decreasing sleep

4 Validation There is an excellent correlation between current ORP and the likelihood of arousal or awakening occurring in the next 30-s epoch (Figs. 5 and 6).

propensity. The green columns show that the range of ORP found in NREM and REM sleep in different individuals is remarkably wide. (From Younes et al. [8]; with permission). Panel b shows the range of ORP values in different stages in individual subjects (each dot is a subject) of five different studies. Average ORP in all studies progressively decreases as the stage progresses from stage wake to stage N3 but there is a wide range of ORP within each stage. The range in REM sleep is widest (ORP ranging from 0.3 to nearly 2.0  in different individuals (see also Fig.  1a, compare horizontal red lines in the upper and lower panels). (From Younes et al. [18]; Creative Commons Attribution License; https:// creativecommons.org/licenses/by/4.0/)

113

Odds Ratio Product (ORP): Description and Implications to Understanding and Management of Sleep Disorders Probability of A/AW in next 30 seconds (%)

b VALIDATION a

Direct Relation with Arousability 8213 7814 5574 3893 2973 1925 1408 605 318

y =23.4x + 7.7 r 2 = 0.98

Odds Ratio Product During Sleep (ORP)

Probability of A/AW in next 30 seconds (%)

Arousability Index (%)

56 PSGs

Relation with Arousability: ORP vs. Delta Power 36 30 24 18 12

6 0

0

0.2

0.4

0.6

0.8

1.0

1.2

Odds Ratio Product

36

1.4

1.6

1.8

30 24 18

12

6 0

3.1

2.9

2.7

2.5

2.3

2.1

1.9

1.7

1.5

1.3

1.1

Log Delta Power Decreasing Delta Power

Fig. 5  In a variety of studies, the change in ORP was consistent with what is expected for sleep depth [7–14]. However, the most convincing validation is the relation between the current ORP in 30-s epochs and the probability of an arousal or awakening occurring in the next epoch (Arousability). Panel a shows this relation in 56 clinical PSGs with various disorders. The numbers at the top are the numbers of 30-s epochs examined at each decile. The average relation is almost perfectly linear. (From Younes et al. [8]; with permission). An important question is whether this relation is superior to the relation between delta power, which is frequently used to assess sleep depth, and arousability measured in the same way. Panel b compares the relation obtained in a different study on normal subjects obtained with ORP and with delta power (From Younes et al. [7]; with permission). As in the earlier study (Panel a), the relation between current ORP and arousability was linear while with delta power arousability reached a minimum at a log delta of Fig. 6  Figure shows the response of ORP to brief noise applications during sleep. It shows that a corresponding dose-dependent transient increase in ORP occurs that can discriminate 5 dB differences in noise amplitude. These transient ORP peaks are associated with reduced reaction time the next day. (From Smith et al. [14]; with permission)

2.5, corresponding to delta power of 300 ΗV2. This value is reached at the earliest appearance of slow wave sleep and increases further, up to a few thousand, as delta waves increase in frequency. It follows that changes in delta power are sensitive to sleep depth only over a small part of its range. In the same study, changes in ORP over 4 nights of sleep restriction matched the changes in spontaneous arousal index, with both decreasing progressively with the number of restriction nights, while changes in delta power were inconsistent and did not follow the pattern of change in arousal index. To this must be added the difficulty of comparing delta power across individuals because absolute delta is sensitive to age and a variety of technical factors (distance between recording and reference electrodes, signal gain, etc.), whereas ORP reflects the gains in different frequencies relative to each other, thereby countering differences between individuals and technical aspects of measurement

VALIDATION ORP increases incrementally with noise intensity ORP 1.37

1.25

1.12

1.00

0.87

0.75

Peak noise

114

M. Younes

5 Main Clinical Advantage of ORP It provides information that helps understand the mechanism of sleep disorders (Figs. 7 and 8).

Main Clinical Advantage of ORP: It Can be er explain underlying mechanism(s) of sleep disorders PaŒent 1:TST=453 min; SE, 95%; N1%=5.8%; N3%=14.1%; REM %, 25.6%; Arousal index=13/hr. Deep sleep 7% TRT Full Wake 2.5% TRT

0

1

2

3

4

ORPWAKE , 2.24; ORP NREM ,1.34; ORPREM , 1.35, ORP TRT , 1.39

5

a

b

6

c

Dee Deep D De eep slee sle sleep l ep 6 66% 6% TRT TRT TR Wake Fulll Wak ke 3.2% 3 % TRTT 3.2

7

PaŒent 2: TST=425 min; SE, 89%; N1%=6.8%; N3%=33.8%; REM %, 16.2%; Arousal index=19/hr. Deep sleep 67% TRT Full Wake 3.5% TRT

0

1

2

3

4

ORPWAKE , 2.02; ORPNREM , 0.40; ORPREM , 0.45, ORPTRT , 0.58

5

Fig. 7  Two patients with complaints of non-restorative sleep. Sleep hypnogram was normal in both cases (values on top of each panel) and there was no OSA or PLMs. The ORP values are shown, epoch by epoch, below the hypnogram. It is visually clear that sleep is mostly light (ORP > 1.0) in patient 1 and mostly deep in patient 2. However, such visual displays need to be converted to numbers for ease of use and comprehension. The figure shows two ways of doing that. In one, the average ORP in all epochs scored as belonging to specified sleep stages is calculated. These are shown for the two patients below each panel. Here, we see that ORP in all stages is higher in patient 1, indicating mostly transitional sleep (ORP between 1.0 and 1.75), whereas in patient 2 ORP was mostly