Intelligence-Based Cardiology and Cardiac Surgery: Artificial Intelligence and Human Cognition in Cardiovascular Medicine

By Anthony C Chang

Intelligence-Based Cardiology and Cardiac Surgery: Artificial Intelligence and Human Cognition in Cardiovascular Medicine provides a comprehensive survey of artificial intelligence concepts and methodologies with real-life applications in cardiovascular medicine. Authored by a senior physician-data scientist, the book presents an intellectual and academic interface between the medical and data science domains. The book's content consists of basic concepts of artificial intelligence and human cognition applications in cardiology and cardiac surgery. This portfolio ranges from big data, machine and deep learning, cognitive computing and natural language processing in cardiac disease states such as heart failure, hypertension and pediatric heart care.

The book narrows the knowledge and expertise chasm between the data scientists, cardiologists and cardiac surgeons, inspiring clinicians to embrace artificial intelligence methodologies, educate data scientists about the medical ecosystem, and create a transformational paradigm for healthcare and medicine.

• Covers a wide range of relevant topics from real-world data, large language models, and supervised machine learning to deep reinforcement and federated learning
• Presents artificial intelligence concepts and their applications in many areas in an easy-to-understand format accessible to clinicians and data scientists
• Discusses using artificial intelligence and related technologies with cardiology and cardiac surgery in a myriad of venues and situations
• Delineates the necessary elements for successfully implementing artificial intelligence in cardiovascular medicine for improved patient outcomes
• Presents the regulatory, ethical, legal, and financial issues embedded in artificial intelligence applications in cardiology

• Language: English
• Publisher: Academic Press
• Release date: Sep 6, 2023
• ISBN: 9780323906296

Book preview

Intelligence-Based Cardiology and Cardiac Surgery

Edited by

Anthony C. Chang

Alfonso Limon

Section Editors

Robert Brisk, Francisco Lopez-Jimenez, and Louise Y. Sun

Table of Contents

Cover image

Title page

Copyright

Dedication

Contributors

About the editors

Foreword by Eric Topol

Foreword by Ami Bhatt

Preface

Acknowledgments

Section I. Basic concepts of data science and artificial intelligence

Chapter 1. Introduction to artificial intelligence for cardiovascular clinicians

Basic concepts of artificial intelligence

History of artificial intelligence

History of artificial intelligence in medicine

Healthcare data and databases

Machine and deep learning

Assessment of model performance

Fundamental issues in machine and deep learning

Other key concepts and technologies in artificial intelligence

Human cognition and artificial intelligence in cardiology

Current status of AI in medicine and relevance to cardiovascular medicine

Artificial intelligence in cardiovascular medicine

Subsection A. Basic concepts of artificial intelligence in cardiology and cardiac surgery

Chapter 2. Application of artificial intelligence in cardiovascular medicine and cardiac surgery

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 3. Data and databases in cardiovascular medicine and surgery

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 4. Data and databases for pediatric and adult congenital cardiac care

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 5. Cognitive biases and heuristics in human cognition

Introduction

Major Takeaways

Chapter 6. Spectrum bias in algorithms and artificial intelligence

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 7. Medical visual question answering

Introduction

Current state of the art

Future directions

Main takeaways

Subsection B. Artificial intelligence in cardiovascular areas

Chapter 8. Artificial intelligence and the electrocardiogram

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 9. Artificial intelligence in electrophysiology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 10. Artificial intelligence in echocardiography

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 11. Artificial intelligence in cardiac CT

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 12. Artificial intelligence in cardiac MRI

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 13. Artificial intelligence in pediatric and congenital cardiac magnetic resonance imaging

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 14. Artificial intelligence in three-dimensional and fetal echocardiography

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 15. Artificial intelligence in nuclear cardiology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 16. Artificial intelligence and in situ exercise monitoring, modeling, and guidance

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 17. Artificial intelligence in the catheterization laboratory

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 18. Artificial intelligence in the cardiology clinic

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 19. Artificial intelligence in cardiac surgery

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 20. Congenital cardiac surgery and artificial intelligence

Introduction

Current state of the art

Future directions

Major takeaways

Subsection C. Clinical applications of artificial intelligence in cardiovascular medicine

Chapter 21. Artificial intelligence in heart failure

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 22. Big data in cardiovascular population health research

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 23. Intelligence-based cardiovascular disease prevention

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 24. Artificial intelligence in cardiovascular genetics

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 25. Artificial intelligence in congenital heart disease

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 26. Artificial intelligence and cardiovascular disease in women

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 27. Artificial intelligence and COVID-19 in children with heart disease

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 28. Artificial intelligence in cardiac critical care

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 29. Artificial intelligence in cardio-oncology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 30. Artificial intelligence in adult congenital heart disease

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 31. Artificial intelligence for quality improvement

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 32. Clinical safety in cardiology and artificial intelligence

Introduction

Current state of the art

Future directions

Major takeaways

Subsection D. Artificial intelligence and related technologies in cardiovascular medicine

Chapter 33. Data sharing principles

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 34. Natural language processing in cardiovascular medicine

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 35. Artificial intelligence and wearable technology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 36. Digital twin in cardiovascular medicine and surgery

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 37. Artificial intelligence and extended reality in cardiology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 38. Cybersecurity and blockchain in cardiology

Introduction

Current state of the art

Future directions

Major takeaways

Subsection E. Artificial intelligence and special topics in cardiovascular medicine

Chapter 39. Starting an artificial intelligence program in cardiology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 40. Artificial intelligence for cardiac care: a view from the top

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 41. Strategy of artificial intelligence in cardiology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 42. Education of artificial intelligence for cardiovascular clinicians

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 43. Artificial intelligence in cardiology: the trainee perspective

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 44. Artificial intelligence and agile project management

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 45. Synthetic data in cardiovascular health research

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 46. Ethical and legal issues in artificial intelligence-based cardiology

Introduction

Current state of the art

Major takeaways

Chapter 47. Regulatory frameworks for artificial intelligence in cardiovascular medicine and surgery

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 48. Regulatory issues of artificial intelligence in cardiology – international perspective

Introduction

Current state of the art

Future Directions

Major Takeaways

Chapter 49. Industry perspective of artificial intelligence in medicine and surgery

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 50. Entrepreneurship lessons from artificial intelligence in cardiology

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 51. Global cardiac network: an innovative artificial intelligence-enabled learning system

Introduction

Current state of the art

Future directions

Major takeaways

Chapter 52. The future of artificial intelligence in cardiology and cardiac surgery

Introduction

The future of artificial intelligence in cardiology and cardiac surgery

The future of artificial intelligence in cardiovascular medicine—stakeholders

The future of dyads in artificial intelligence in cardiovascular medicine

Major takeaways

Suggested readings

Section III Artificial intelligence in medicine compendium

Appendix. Compendium

Recommended resources

Glossary

Index

Copyright

Academic Press is an imprint of Elsevier

125 London Wall, London EC2Y 5AS, United Kingdom

525 B Street, Suite 1650, San Diego, CA 92101, United States

50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom

Copyright © 2024 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

ISBN: 978-0-323-90534-3

For information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Stacy Masucci

Acquisitions Editor: Tracey Lange

Editorial Project Manager: Timothy J. Bennett

Production Project Manager: Punithavathy Govindaradjane

Cover Designer: Greg Harris

Typeset by TNQ Technologies

Dedication

First and foremost, this book is dedicated to the many thousands of children and adults with heart disease whom I have had the privilege to serve as their cardiologist (including a very special nine-year-old girl named Ilsa, from Myanmar, with rheumatic heart disease and whose passing inspired me to pursue this domain and make artificial intelligence in medicine and cardiology available to anyone around the world). Their supreme fortitude and will to survive in their long and complex medical journeys will continually motivate me to maintain my ardent passion for learning and teaching this new and wondrous world of artificial intelligence in clinical cardiovascular medicine and healthcare.

This book is also dedicated to my countless mentors in cardiology and cardiac surgery, particularly several special ones: first, Dr. William Norwood, who not only taught me the nuances of congenital heart disease and cardiac surgery but also the amazing world of complex mathematics in biomedicine (he was decades ahead of his time); second, Dr. Richard van Praagh, whose passion for learning and dedication to teaching all aspects of cardiology (including its mathematical dimensions) have inspired me even to this day, and finally, Dr. Peter Holbrook, who encouraged me to pursue my dream of becoming a cardiac intensive care clinician and taught me about chaos theory. I would also like to pay a special tribute to the too-many-to-individually-name clinicians in cardiology and cardiac surgery, as well as data scientists and AI experts, who are open-minded in taking on this new domain of artificial intelligence as it is applied to cardiovascular medicine, as difficult as it is and as challenging as it can be. The synergy between these two groups of dedicated advocates will define and foster the field of AI in cardiology and cardiac surgery. One of the best rewards of putting this book together has been knowing the special people behind the chapters and convening this special group to become the community of AI advocates for cardiology and cardiac surgery for decades to come.

Finally, I look forward to sharing this work with my two beautiful daughters, Emma, who has complex congenital heart disease, and Olivia, who yearns to be a pediatric cardiologist someday with a data science education to care for her beloved sister after daddy d - - s (she doesn't like that word). I like to thank them not only for their pure love and innocent joy for their far-from-perfect father but also for their utter affinity for sleep and reading that allowed me just sufficient quiet time to write most of this work (in front of the serene immensity of the Pacific Ocean while sublime pieces by Mozart and Satie were playing in the background, which I wholeheartedly recommend to any author).

Anthony C. Chang

Spring, 2023

Contributors

Michael J. Ackerman

Department of Pediatric and Adolescent Medicine, Division of Pediatric Cardiology, Mayo Clinic, Rochester, MN, United States

Department of Molecular Pharmacology & Experimental Therapeutics, Windland Smith Rice Sudden Death Genomics Laboratory, Mayo Clinic, Rochester, MN, United States

Department of Cardiovascular Medicine, Division of Heart Rhythm Services, Windland Smith Rice Genetic Heart Rhythm Clinic, Mayo Clinic, Rochester, MN, United States

Demilade A. Adedinsewo,     Department of Cardiovascular Medicine, Mayo Clinic, Jacksonville, FL, United States

Oguz Akbilgic,     Cardiovascular Section, Internal Medicine, Wake Forest School of Medicine, Winston–Salem, NC, United States

Zaidon Al-Falahi,     Cardiology Department, Campbelltown Hospital, South West Sydney Local Health District, Sydney, NSW, Australia

Deya Alkhatib,     Division of Cardiovascular Diseases, University of Tennessee Health Science Center, Memphis, TN, United States

Mohamad Alkhouli,     Department of Cardiovascular Medicine, Mayo Clinic School of Medicine, Rochester, MN, United States

Jordan D. Awerbach

Center for Heart Care, Division of Cardiology, Phoenix Children's, Phoenix, AZ, United States

Department of Child Health, University of Arizona-College of Medicine-Phoenix, Phoenix, AZ, United States

David M. Axelrod,     Department of Pediatrics (Cardiology), Stanford University, Palo Alto, CA, United States

G. Hamilton Baker,     CU-MUSC Artificial Intelligence Hub, Medical University of South Carolina, Charleston, SC, United States

Mohamad Bashir,     Vascular and Endovascular Surgery, Velindre University NHS Trust, Health Education and Improvement, Wales, United Kingdom

Raymond Bond,     Faculty of Computing, Engineering & the Built Environment, Ulster University, Coleraine, United Kingdom

J. Martijn Bos

Department of Pediatric and Adolescent Medicine, Division of Pediatric Cardiology, Mayo Clinic, Rochester, MN, United States

Department of Molecular Pharmacology & Experimental Therapeutics, Windland Smith Rice Sudden Death Genomics Laboratory, Mayo Clinic, Rochester, MN, United States

Robert Brisk

Department of Cardiology, Southern Health & Social Care Trust, Portadown, United Kingdom

Faculty of Computing, Engineering & the Built Environment, Ulster University, Coleraine, United Kingdom

Sherry-Ann Brown,     Cardio-Oncology Program, Division of Cardiovascular Medicine, Medical College of Wisconsin, Milwaukee, WI, United States

Liam Butler,     Cardiovascular Section, Internal Medicine, Wake Forest School of Medicine, Winston–Salem, NC, United States

Anthony C. Chang

Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children’s Health of Orange County, Orange, CA, United States

Heart Failure Program, Heart Institute, Children’s Health of Orange County, Orange, CA, United States

Chapman University, Orange, CA, United States

Oscar Camara,     Physense, BCN Medtech, Department of Information and Communications Technologies, University Pompeu Fabra, Barcelona, Spain

Timothy Chou,     Department of Computer Science, Stanford University, Stanford, CA, United States

Matthew Davis,     CU-MUSC Artificial Intelligence Hub, Medical University of South Carolina, Charleston, SC, United States

Joseph A. Dearani

Division of Cardiovascular Surgery, Mayo Clinic-Children’s Minnesota Cardiovascular Collaborative, Minneapolis, MN, United States

Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, United States

Division of Cardiovascular Surgery, The Children’s Heart Clinic, Children’s Hospitals and Clinics of Minnesota, Minneapolis, MN, United States

Mohsen Dorraki

South Australian Health and Medical Research Institute (SAHMRI), Adelaide, SA, Australia

Australian Institute for Machine Learning (AIML), The University of Adelaide, Adelaide, SA, Australia

Rami Doukky,     Division of Cardiology, Cook County Health, Chicago, IL, United States

Regina Druz,     Holistic Heart Centers, Miami, FL, United States

Louis Ehwerhemuepha,     Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children’s Health of Orange County, Orange, CA, United States

Zachary Ernst

Ainthoven, Inc., Orlando, FL, United States

Who We Play for (501(c)3 Organization), Orlando, FL, United States

William Feaster

CHOC Children's, Orange, CA, United States

Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children’s Health of Orange County, Orange, CA, United States

Albert K. Feeny,     Department of Internal Medicine, University of California, San Francisco, CA, United States

Beatriz A. Fernandez-Campos,     Division of Cardiology, University of Toronto, University Health Network and Mount Sinai Hospital, Toronto, ON, Canada

Michael Fisher

Department of Cardiology, Liverpool University Hospital NHS FT, Liverpool, United Kingdom

Liverpool Centre for Cardiovascular Science, University of Liverpool, Liverpool, United Kingdom

Mona G. Flores,     Healthcare, NVIDIA, Santa Clara, CA, United States

Wayne J. Franklin

Center for Heart Care, Division of Cardiology, Phoenix Children's, Phoenix, AZ, United States

Department of Child Health, University of Arizona-College of Medicine-Phoenix, Phoenix, AZ, United States

Abby Frederickson,     Phoenix Children’s/Banner University Medical Center, Medicine-Pediatrics Residency Program, Phoenix, AZ, United States

Sharib Gaffar,     Department of Pediatric Cardiology, UCLA Mattel Children's Hospital, Los Angeles, CA, United States

Darren Gates,     Alder Hey Children’s Hospital, Liverpool, United Kingdom

Addison Gearhart,     Department of Cardiology, Boston Children's Hospital, Department of Pediatrics, Harvard Medical School, Boston, MA, United States

Sara Gerke,     Penn State Dickinson Law, Carlisle, PA, United States

Javier Gomez,     Division of Cardiology, Cook County Health, Chicago, IL, United States

Shinichi Goto,     Division of Cardiovascular Medicine, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, United States

Giorgia Grutter,     Heart Failure and Transplant Unit, Bambino Gesù Pediatric Hospital, Rome, Italy

E. Kevin Hall,     Division of Pediatric Cardiology, Department of Pediatrics, Yale University School of Medicine and Division of Biostatistics (Health Informatics), Yale University School of Public Health, New Haven, CT, United States

Jeffrey P. Jacobs,     Congenital Heart Center, Division of Cardiovascular Surgery, Departments of Surgery and Pediatrics, University of Florida, Gainesville, FL, United States

John L. Jefferies,     Division of Cardiovascular Diseases, University of Tennessee Health Science Center, Memphis, TN, United States

Anusha Jegatheeswaran

Department of Cardiothoracic Surgery, Great Ormond Street Hospital for Children, Great Ormond Street Hospital, London, United Kingdom

Children's Cardiovascular Disease, Institute of Cardiovascular Sciences, University College London, London, United Kingdom

Kathy Jenkins,     Center for Applied Pediatric Quality Analytics, Boston Children's Hospital, Boston, MA, United States

Pei-Ni Jone,     Lurie Children's Hospital, Northwestern University Feinberg School of Medicine, Chicago, IL, United States

Vahid Ghodrati Kouzehkonan,     UCLA Graduate Program in Biomedical Physics, Los Angeles, CA, United States

S. Ram Kumar

Division of Pediatric Cardiothoracic Surgery, University of Nebraska Medical Center, Criss Heart Center, Children’s Hospital and Medical Center, Omaha, NE, United States

Department of Surgery, University of Southern California, Los Angeles, CA, United States

Wyman Lai

CHOC Children's, Orange, CA, United States

Department of Pediatrics, UC Irvine School of Medicine, Irvine, CA, United States

Peter C. Laussen

Health Affairs, Boston Children's Hospital, Boston, MA, United States

Department of Anaesthesia, Harvard Medical School, Boston, MA, United States

Howard Lei

CHOC Children's, Orange, CA, United States

Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children’s Health of Orange County, Orange, CA, United States

Ruben Leta,     Cardiac Imaging Unit, Cardiology Department, Hospital de La Santa Creu I Sant Pau, Barcelona, Spain

Zhibin Liao,     Australian Institute for Machine Learning (AIML), The University of Adelaide, Adelaide, SA, Australia

Alfonso Limon,     Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children's Health of Orange County, Orange, CA, United States

Francisco Lopez-Jimenez,     Department of Cardiovascular Medicine, Mayo Clinic, Rochester, MN, United States

Aleksandra Lopuszko,     Barts and The London School of Medicine and Dentistry, Queen Mary University of London, London, United Kingdom

Kevin Maher,     Cardiac Intensive Care Unit, Children's Hospital of Atlanta, Pediatrics, Emory School of Medicine, Atlanta, GA, United States

Rachel Marano,     Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children’s Health of Orange County, Orange, CA, United States

Donald Mattia,     Phoenix Children's, Pediatric Residency Program, Phoenix, AZ, United States

David McEneaney

Department of Cardiology, Southern Health & Social Care Trust, Portadown, United Kingdom

Faculty of Computing, Engineering & the Built Environment, Ulster University, Coleraine, United Kingdom

Colin J. McMahon

Department of Paediatric Cardiology, Children's Health Ireland at Crumlin, Dublin, Ireland

School of Medicine, University College Dublin, Dublin, Ireland

School of Health Professions Education, Maastricht University, Maastricht, The Netherlands

Thierry Mesana,     University of Ottawa Heart Institute, Ottawa, ON, Canada

Patrizio Moras,     Department of Perinatal Cardiology, Bambino Gesù Children's Pediatric Hospital IRCCS, Rome, Italy

Tatiana Moreno,     Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children’s Health of Orange County, Orange, CA, United States

Abdel Hakim Moustafa,     Cardiac Imaging Unit, Cardiology Department, Hospital de La Santa Creu I Sant Pau, Barcelona, Spain

Tarek Nafee,     Department of Medicine, Roger Williams Medical Center, Providence, RI, United States

Nitish Nag,     University of California, Irvine, CA, United States

Jai Nahar,     Department of Pediatrics, Division of Cardiology, Children's National Heart Institute, George Washington University School of Medicine and Health Sciences, Children's National Hospital, Washington, DC, United States

Hoang H. Nguyen,     Department of Pediatrics, University of Texas Southwestern, Dallas, TX, United States

Olufemi Olajide,     Alder Hey Children’s Hospital, Liverpool, United Kingdom

David M. Overman

Division of Cardiovascular Surgery, Mayo Clinic-Children’s Minnesota Cardiovascular Collaborative, Minneapolis, MN, United States

Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, United States

Division of Cardiovascular Surgery, The Children’s Heart Clinic, Children’s Hospitals and Clinics of Minnesota, Minneapolis, MN, United States

J. Paul Finn,     Department of Radiological Sciences, David Geffen School of Medicine at UCLA, Los Angeles, CA, United States

Jessily P. Ramirez,     Boston Children’s Hospital, Boston, MA, United States

David Rayan,     Department of Medicine, Medical College of Wisconsin, Milwaukee, WI, United States

Mitch Recto,     Cardiac Catheterization Laboratory, Heart Institute, Children's Health of Orange County, Orange, CA, United States

Charitha D. Reddy,     Stanford University, Lucile Packard Children's Hospital, Department of Pediatrics, Division of Pediatric Cardiology, Palo Alto, CA, United States

Partho P. Sengupta,     Robert Wood Johnson University Hospital, New Brunswick, NJ, United States

Ketemwabi Y. Shamavu,     Biomedical Informatics, Center for Intelligent Health Care, University of Nebraska Medical Center, Omaha, NE, United States

Candice K. Silversides,     Division of Cardiology, University of Toronto, University Health Network and Mount Sinai Hospital, Toronto, ON, Canada

Elsayed Z. Soliman,     Cardiovascular Section, Internal Medicine, Wake Forest School of Medicine, Winston–Salem, NC, United States

James D. St Louis

Department of Cardiac Surgery, Inova Fairfax Hospital and Inova L.J Murphy Children’s Hospital, Fairfax, VA, United States

Departments of Surgery and Pediatrics, Children’s Hospital of Georgia, Augusta University, Augusta, GA, United States

Louise Y. Sun,     Division of Cardiothoracic Anesthesiology, Stanford University School of Medicine, Stanford, CA, United States

Sven Tan,     Barts and The London School of Medicine and Dentistry, Queen Mary University of London, London, United Kingdom

Animesh Tandon

Department of Pediatric Cardiology, Children's Institute, Cleveland Clinic, Cleveland, OH, United States

Department of Pediatrics, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland, OH, United States

Department of Biomedical Engineering, Cleveland Clinic Lerner Research Institute and Case School of Engineering at Case Western Reserve University, Cleveland, OH, United States

James E. Tcheng,     Family Medicine and Community Health (Informatics), Duke University School of Medicine, Durham, NC, United States

Alessandra Toscano,     Department of Perinatal Cardiology, Bambino Gesù Children's Pediatric Hospital IRCCS, Rome, Italy

Alberto Eugenio Tozzi,     Multifactorial and Complex Diseases Research Area, Bambino Gesù Pediatric Hospital IRCSS, Rome, Italy

Tu Hao Tran,     Cardiology Department, Liverpool Hospital, South West Sydney Local Health District, Sydney, NSW, Australia

Bhavya Trivedi

Pediatric Cardiology and Electrophysiology, Children's Cardiology Clinic, LLC, Orlando, FL, United States

Ainthoven, Inc., Orlando, FL, United States

Wendy Tsang,     Division of Cardiology, University of Toronto, University Health Network and Mount Sinai Hospital, Toronto, ON, Canada

Andrew S. Tseng,     Department of Cardiovascular Medicine, Mayo Clinic, Rochester, MN, United States

Anton van den Hengel,     Australian Institute for Machine Learning (AIML), The University of Adelaide, Adelaide, SA, Australia

Johan W. Verjans

Australian Institute for Machine Learning (AIML), The University of Adelaide, Adelaide, SA, Australia

South Australian Health and Medical Research Institute (SAHMRI), Adelaide, SA, Australia

Royal Adelaide Hospital, Adelaide, SA, Australia

Krista Young,     University of Iowa Hospitals and Clinics, Iowa City, IA, United States

About the editors

Anthony C. Chang, MD, MBA, MPH, MS

Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children's Health of Orange County, Orange, CA, United States; Heart Failure Program, Heart Institute, Children's Health of Orange County, Orange, CA, United States; Chapman University, Orange, CA, United States

Dr. Chang attended Johns Hopkins University for his BA in molecular biology prior to entering Georgetown University School of Medicine for his MD. He then completed his pediatric residency at the Children's Hospital National Medical Center and his pediatric cardiology fellowship at the Children's Hospital of Philadelphia. He then accepted a position as an attending cardiologist in the cardiovascular intensive care unit of Boston Children's Hospital and as an assistant professor at Harvard Medical School. He has been the medical director of several pediatric cardiac intensive care programs (including the Children's Hospital of Los Angeles, Miami Children's Hospital, and Texas Children's Hospital). He served as the medical director of the Heart Institute at the Children's Health of Orange County.

He is currently the chief intelligence and innovation officer and medical director of the Heart Failure Program at Children's Health of Orange County. He has also been named a Physician of Excellence by the Orange County Medical Association and a Top Cardiologist, Top Doctor for many years, as well as one of the nation's Top Innovators in Healthcare.

He completed his Master in Business Administration (MBA) in health care administration at the University of Miami School of Business and graduated with the McCaw Award of Academic Excellence. He also completed a Master in Public Health in health care policy at the Jonathan Fielding School of Public Health of the University of California, Los Angeles, graduating with the Dean's Award for Academic Excellence. Finally, he graduated with his Master of Science (MS) in biomedical data science with a subarea focus in artificial intelligence from the Stanford School of Medicine. He has completed a certification in artificial intelligence from both MIT and the University of California, Berkeley. He is an associate scholar at Stanford Artificial Intelligence and Medicine and Imaging, as well as a computer scientist-in-residence and a member of the Dean's Scientific Council at Chapman University.

He has helped build a successful cardiology practice as a startup company and completed a deal on Wall Street. He is known for several innovations in pediatric cardiac care, including introducing the cardiac drug milrinone and codesigning (with Dr. Michael DeBakey) an axial-type ventricular assist device in children. He has been a committee member of the National Institute of Health pediatric grant review committee. He is the editor of several textbooks in pediatric cardiology and intensive care, including Pediatric Cardiac Intensive Care, Heart Failure in Children and Young Adults, and Pediatric Cardiology Board Review. He founded the Pediatric Cardiac Intensive Care Society, which launched a multidisciplinary focus on cardiac intensive care for children. He is also the founder of the Asia-Pacific Pediatric Cardiac Society, which united pediatric cardiologists and cardiac surgeons from 24 Asian countries and launched a biennial meeting in Asia that now draws over 1000 attendees.

He is the founder and medical director of the Medical Intelligence and Innovation Institute, which is supported by the Sharon Disney Lund Foundation (since 2015). The institute is dedicated to implementing data science and artificial intelligence in medicine and is the first institute of its kind within a hospital. The new institute is concomitantly dedicated to facilitating innovation in children and health care worldwide. He is the former organizing chair for the biennial Pediatrics2040: Emerging Trends and Future Innovations meeting, as well as the founder and codirector of the Medical Intelligence and Innovation Summer Internship Program, which mentors close to 100 young physicians-to-be every summer. He is also the founder and chairman of the board of a pediatric innovation leadership group called the International Society for Pediatric Innovation.

He intends to build a clinician–computer scientist interface to enhance all aspects of data science and artificial intelligence in health and medicine. He lectures widely on artificial intelligence in medicine (he has been called Dr. A.I. by the Chicago Tribune) and has been named an AI Influential Thinker. He has given a TEDx talk and is a regularly featured speaker at Singularity University's Exponential Medicine. He has published numerous review papers on big data and predictive analytics as well as machine learning and artificial intelligence in medicine. He is on the editorial board of the Journal of Medical Artificial Intelligence. He is the editor of a book with Elsevier, Intelligence-Based Medicine: Principles and Applications of Data Science, Artificial Intelligence, and Human Cognition in Medicine and Healthcare, and the editor-in-chief of the journal of the same name. He is the founder and organizing chair of several artificial intelligence in medicine meetings in the United States and abroad (Europe and Asia) that will focus on artificial intelligence in healthcare and medicine. He is also the founding chair of a new group for clinicians with a special focus on data science and artificial intelligence as a nascent society (Medical Intelligence Society, or MIS). He is the cofounder and chair of the American Board of AI in Medicine, which holds monthly courses at all levels for everyone in AI in medicine, as well as the cofounder and cochair of the nascent Alliance of Centers of AI in Medicine and its pediatric subgroup.

He is the founder of three startup companies in the artificial-intelligence-in-medicine domain:

1) Medical Intelligence 1, LLC is a startup focused on using graph databases and knowledge graphs for rare disease detection and other applications.

2) Artificial Intelligence in Medicine, LLC is a multimedia and events company that organizes meetings and educational programs in artificial intelligence in medicine in local and global venues and across subspecialties.

3) Medical Intelligence 10 (MI10), LLC is an education and consulting/advising conglomerate for clinicians, executives, and leaders of healthcare organizations and companies, as well as investors, for evaluating and implementing AI strategies in healthcare organizations and companies and assessing and implementing education for health professional schools. The proprietary MI10 assessment tool (MIQ™) can evaluate any organization or company for AI readiness and quality and uses both deep learning and cognitive architecture for its AI-enabled strategy recommendations.

Alfonso Limon, PhD

Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children's Health of Orange County, Orange, CA, United States

Alfonso Limon, PhD, is an expert in numerical analysis and frequently speaks about the intersection of predictive analytics and healthcare. He is a principal at Oneirix, a consulting company developing market-leading technologies in computational intelligence for med-tech. Before joining Oneirix, Dr. Limon served as the director of research at Intersection Medical (I-Med), leading the development of algorithms for decision support systems to manage congestive heart failure. Prior to his work at I-Med, he was a senior research scientist at Impedance Cardiology Systems. He is now the Senior data scientist at the Medical Intelligence, Information, Investigation, and Innovation (Mi4) Institute at Children's Health of Orange County (CHOC).

Before his work in industry, Dr. Limon was a visiting professor of mathematics at Pomona College and held a postdoctoral fellowship at Harvey Mudd College in the math department. He earned his BS in mechanical engineering at San Diego State University, an MS in mathematics from Claremont Graduate University, and a PhD in computational science from CGU/SDSU. Dr. Limon holds several impedance spectroscopy patents and has published on a wide range of technical topics, ranging from computational finance to optimal bean design to transfer energy from a satellite.

Dr. Limon sits on the American Board of Artificial Intelligence in Medicine, with duties including teaching machine learning to medical professionals. He also teaches deep learning to graduate students of the Computational Research Center at SDSU. Dr. Limon advises several AI startups on technical development and strategy, including Medical Intelligence One, Waya Health, Data to Decision, and a few others in stealth mode.

Dr. Limon serves as an associate editor of the journal Intelligence-Based Medicine, the board chair of the Computational Science Research Center at SDSU, an advisory board member to the UCSD Extension for Machine Learning and Engineering, and an advisory board member of the San Diego Information Technology Roundtable and is a former chair of the IEEE Consultants' Network of San Diego.

Section editors

Robert Brisk, MBBCh, MRCP, PhD

Department of Cardiology, Southern Health & Social Care Trust, Portadown, United Kingdom; Faculty of Computing, Engineering & the Built Environment, Ulster University, Coleraine, United Kingdom

Robert spent the first decade of his career as a full-time medical doctor in the UK's National Health Service. Today, he continues to practice general adult cardiology but dedicates a substantial portion of his professional life to research and innovation in bioinformatics. He is particularly passionate about translational deep-learning research in the clinical arena. In his role as chief scientific officer at Eolas Medical, his primary focus is on biomedical language processing as a means of driving evidence-based clinical practice.

Francisco Lopez-Jimenez, MD, MSc, MBA

Department of Cardiovascular Medicine, Mayo Clinic, Rochester, MN, United States

Dr. Lopez-Jimenez is a professor of medicine at Mayo College of Medicine, the chair of the Division of Preventive Cardiology at Mayo Clinic, and codirector of artificial intelligence in cardiology in the Department of Cardiovascular Medicine. He is the editor-in-chief of Mayo Clinic Proceedings: Digital Health and a cochair of the Advanced Healthcare Analytics workgroup, American College of Cardiology. Dr. Lopez-Jimenez did his cardiology fellowship at Mount Sinai Medical Center in Miami, Florida, and Brigham and Women's Hospital, Harvard Medical School. He holds an MS degree from Harvard School of Public Health and an MBA degree from Augsburg University. The artificial intelligence program in cardiology at Mayo Clinic that he coleads has been quite productive, publishing numerous articles in the last 5 years, primarily using ECG signals, which has been extended to outcome-based machine learning and image processing. Dr. Lopez-Jimenez has published more than 350 scientific publications, and his scientific work has been cited more than 12,000 times.

Louise Y. Sun, MD, SM, FRCPC, FAHA

Division of Cardiothoracic Anesthesiology, Stanford University School of Medicine, Stanford, CA, United States

Dr. Louise Y. Sun recently joined the Stanford University School of Medicine as the chief of cardiothoracic anesthesiology and professor of anesthesiology, perioperative and pain medicine. She is an adjunct scientist at the Institute for Clinical Evaluative Sciences (ICES) in Toronto. Prior to this, she was an associate professor of anesthesiology and epidemiology and director of big data and health bioinformatics research at the University of Ottawa Heart Institute and a clinical research chair in big data and cardiovascular outcomes at the University of Ottawa.

Dr. Sun received her medical degree from McMaster University. She completed her anesthesiology residency at the University of Ottawa and her MS in epidemiology at the Harvard School of Public Health, followed by a clinical and research fellowship in cardiac anesthesia at the University of Toronto. She then joined the Division of Cardiac Anesthesiology at the University of Ottawa Heart Institute and was cross-appointed as an ICES faculty member.

Dr. Sun's areas of clinical focus are hemodynamic monitoring and heart failure. Her methodologic areas of focus are the conduct of population-based cohort studies using large healthcare databases; predictive analytics; sex and gender epidemiology; patient engagement; innovative methods for data processing and warehousing; and software and applications development. Her research leverages big data and digital technology to bridge key gaps in care delivery and outcomes for patients with heart failure or undergoing cardiovascular interventions, zooming in on sex/gender and personalized care. She holds several patents and collaborates with health authorities and policymakers to evaluate and report on models of cardiac healthcare delivery.

Dr. Sun is active in the scientific community. She sits on a number of US, Canadian, and international editorial boards and scientific and grant review committees. She collaborates nationally and internationally on various population health and data science initiatives. Her patient-centered research program aims to improve access to care and outcomes, focusing on personalized risk stratification and long-term, patient-defined outcomes. She has authored over 100 peer-reviewed papers and published in leading clinical journals, including JAMA, JAMA Cardiology, JAMA Internal Medicine, Circulation, JACC, Diabetes Care, and Anesthesiology. Her research program has been well funded by the Canadian Institutes of Health Research, the Heart and Stroke Foundation of Canada, and the Ontario Ministry of Health.

Foreword by Eric Topol

The first automated, machine-read electrocardiograms (ECGs) appeared in the 1970s, nearly 50 years ago. I remember looking at automated 12-lead interpretations as a medical resident at UCSF in the early 1980s and even as a novice reader realizing how they were checkered with blatant errors. This early foray into artificial intelligence for cardiology, using rules-based heuristics, was rife with mistakes. That was highlighted in 1991 when a large international assessment determined the accuracy of this technology was only 69%.

Fast-forward to the era of deep learning in recent years that has brought us the extraordinary capability for machines to ingest massive datasets, particularly input of images but also speech and text, passing through innumerable layers of artificial, hidden neurons and obtaining outputs of interpretation that can be highly accurate. These deep neural networks are setting the foundation for revolutionary changes in cardiovascular medicine.

Take the 12-lead ECG, for example. Following a randomized controlled trial conducted at Mayo Clinic provided an interpretation for low ejection fraction for primary care physicians, this highly revered health system now routinely provides interpretations that suggest low ejection fraction, possible pulmonary hypertension, hypertrophic cardiomyopathy, and other diagnoses to consider. This goes well beyond improving the accuracy of basal readings such as rhythm, axis, conduction disturbances, and prior infarction. But we're also learning that there's much more information on the ECG that we, as cardiologists, cannot see, but machines can. That includes a reasonable estimate of the age and sex of the patient, the detection of and severity of anemia, and the likelihood of developing atrial fibrillation, heart failure, or valvular heart disease within a person's lifetime. By providing inputs of millions of ECGs with ground truths, supervised learning has taught machine eyes to pick up features and information that previously would not have been possible.

Echocardiography is similarly going through a major AI transformation for both acquisition and interpretation. Multiple algorithms have been developed for uninitiated, untrained people to obtain high-quality echocardiograms. By having the individual place an ultrasound probe on the chest, the AI can direct the movement of the probe in any direction on the chest and clockwise/counterclockwise. When the algorithm detects the desired window in the plane, a video loop is automatically captured—just as simple as when a check is deposited through a smartphone app to the bank. The acquisition can lead to the autolabeling of heart chambers, valves, and other structures, along with calculating the left ventricular ejection fraction. There are even AI deep neural networks that will make all the measurements, such as the size of the left atrium, thickness of the myocardium, or dimension of the proximal aorta, and create a full report. Just think of how the combination of acquisition and interpretation will further develop in the years ahead, such that an echocardiogram could be obtained by anyone, such as a patient with heart failure, and be automatically interpreted on a preliminary basis with input by a physician as needed. Smartphone echocardiography and ultrasonography are already being performed in places throughout the world that previously had little or no access; further progress has the potential to reduce inequities in access to and interpretation of echocardiograms. And there seem to be almost no limits to complexity; at UCSF, the accurate interpretation of fetal echocardiograms for the whole gamut of common congenital abnormalities has already been demonstrated.

Of course, as aptly reviewed in the chapters of this book, medical images as inputs for convolutional neural networks are not at all restricted to ECGs and echocardiograms. MRIs, CT scans, and PET and nuclear scans are part of the wavefront of cardiovascular AI. And relevant images for cardiologists are not only derived from heart and vessel scans. It turns out that the retina is rich with information that has been extracted via deep learning. The calcium score of the coronary arteries can be predicted by the retinal vessels; initial studies support the ability to detect blood pressure control and glucose regulation in people with diabetes via retinal photos. The retina may also prove to be a gateway for early diagnoses of kidney, hepatobiliary, and neurodegenerative diseases.

So far, I've touched on images for clinicians, but AI can also help patients who are increasingly gaining the capability of capturing their data. In fact, the very first AI algorithm cleared for consumers by the Food and Drug Administration was for the detection of atrial fibrillation from a smartwatch. When used by the right people with an increased risk, such as prior rhythm disturbances or symptoms, an abnormally fast resting heart rate warns the individual to capture a one-lead rhythm strip that is automatically interpreted. Countless emergency room or urgent care visits have been avoided by the use of such apps, although false-positive alerts have occurred at a high rate, predominantly in people at low risk who probably should not be using the apps.

Early in the pandemic, our team at Scripps Research developed a smartphone app that captured resting heart rate, along with physical activity (steps) and sleep metrics from the fitness wristbands or smartwatches of nearly all manufacturers to help make an early diagnosis of COVID-19, identify a signature that indicates a high risk of long COVID, and detects the physiological response to vaccination even when a person experienced no symptoms. The use of recurrent neural networks to analyze sensor data is nascent, but as all vital signs can be continuously monitored, the foundation will eventually be laid for safe and inexpensive monitoring of patients in their homes, obviating the need for admission or a longer stay in the hospital.

Of course, AI inputs extend well beyond images and sensor data. They include speech, and already, we are seeing many efforts to transform the conversation between a patient and physician during a clinic visit into a high-quality note that is at least as good, if not better, than typically found in Epic or Cerner electronic medical records. This is the beginning of the liberation of physicians from keyboards, since these conversations can eventually include orders for lab tests, scans, consults, prescriptions, and follow-up appointments, as well as automated coding and billing. Perhaps nothing would be more meaningful for improving the daily lives of clinicians than not having to spend so much time working on keyboards.

The other major substrate is text, which lags behind images and voice, especially when unstructured. However, considerable effort is ongoing to crack that barrier, with the potential in the near term to fully review a patient's chart entirely and synthesize the major problems and trends. This will lead to someday being able to fully review the medical literature pertaining to a patient's conditions.

All this progress and what lies on the horizon are exciting, but it's imperative to keep in mind the substantial liabilities of AI. These include not only concerns about privacy and security, the embedded bias of input data that can compromise fairness and unwittingly promote discrimination, the potential to worsen health inequities, algorithm explainability, the chasm of algorithm validation to clinical implementation, the fulfillment of regulatory clearance for a deep neural network without transparency to the clinical community, the need for surveillance of AI after implementation, and a long list of issues that must be fully addressed and grappled with for AI's potential to be actualized.

For the first time, Intelligenge-Based Cardiology comprehensively pulls the AI field together, and the authors and editors deserve considerable praise for their efforts. I suspect many share my vision that the most far-reaching aspect of AI and medicine, beyond improved accuracy or streamlined workflow, will be profound improvements in the patient–doctor relationship. That was the premise of my Deep Medicine book, and in the years since it was published, my sense that this should be considered the overarching goal—to use AI technology to improve human connections—has only been strengthened. We have such a remarkable potential to enhance humanity in medicine going forward, and I truly hope that cardiologists will be a leading force in making that happen.

Foreword by Ami Bhatt

For decades, the cardiovascular (CV) field has been at the forefront of innovation. Following the pharmaceutical and device industries, the field of digital health and data analytics is now the newest essential part of the foundation of CV innovation. The sheer amount of well-organized data available for deeper assessment than the human brain can achieve is a gold mine for better understanding CV disease, from population-level data to personalized individual insights. Anthony C. Chang and the authors herein have unfolded the complex world of CV care into the basic building blocks we have long relied on and then advanced our ability to offer care by demonstrating the superior power of data analytics-infused clinical acumen at every stage.

A few examples of the actionable strengths of CV AI include the role of pattern recognition, the ability to triage the level of disease, and the unveiling of systems laden with implicit bias. Pattern recognition allows for earlier diagnosis and the dissemination of new medical knowledge to all clinical caregivers and the community. In some cases, pattern recognition contributes to hypothesis generation for testing and assessment. Triage mechanisms integrating multiple data streams can ensure timely care with appropriate resource utilization. As access to care remains a major challenge for the healthcare industry, allowing sick patients to receive care earlier to minimize emergency visits is essential. The functional issue of long wait times also conjures up concerns in the minimally ill and worried well, and triage mechanisms promote a positive patient experience and decrease unnecessary anxiety. Lastly, the application of AI can lessen the digital divide by helping us see patterns of bias we may have overlooked and ensuring that more diverse contingents are offered up for analysis.

This book is an ideal introduction for those practicing medicine or with a healthcare career adjacent to cardiology who want a picture of where the field is headed. It is also a tome of exquisite insightful detail from CV innovators in AI/ML/NLP and associated fields of study who understand and beautifully convey the importance of pursuing this field and its direct applicability to improving CV care.

Preface

It is not the strongest of the species that survives, not the most intelligent that survives. It is the one that is the most adaptable to change. Often attributed to Charles Darwin

Why a book on artificial intelligence in cardiology and cardiac surgery, and why now?

I sincerely believe that cardiovascular medicine and cardiac surgery are amongst the subspecialties that can benefit most from everything artificial intelligence and related technologies offer. Cardiology, with its breadth and depth, from congenital heart disease to coronary artery disease, the complexity of patients across the age spectrum, and the diversity of medical images and other types of data, can fully take advantage of the panoply of AI technologies to improve the care of these patients.

Artificial intelligence, especially machine and deep learning, has established itself as a valuable resource not only in myriad sectors of our society but also, to some extent, in healthcare and clinical medicine. More recently, large language models such as ChatGPT and GPT-4 have very much captured the interest and fulfilled both the excitement and concern of artificial intelligence. Over the past decade, the more mature areas of artificial intelligence in biomedicine have included medical image interpretation, machine learning in various aspects of clinical workflow and applications, protein genomic sequence-to-structure determination, and robotic process automation for healthcare administration. During this early era of modern AI in healthcare, the disappointments were innumerable. Particularly noteworthy have been continual challenges with data access and accuracy, the publication-to-practice schism, and the lack of clear success with real-time decision support. All of these deficiencies were made even more obvious during the COVID-19 pandemic. With the advent of AI technologies such as federated and swarm learning, transformers for language and imaging, digital twins, synthetic data, foundation models, and edge computing, the future of artificial intelligence in cardiovascular medicine is very promising. We are in the Medieval musical period of AI in healthcare, with hauntingly beautiful musical passages—but the music needs to mature into the more sophisticated concertos and symphonies of the later Baroque, Classical, and Romantic periods.

I have rewritten and updated (with ample help from my good friend and AI muse Dr. Alfonso Limon) many parts of the original book Intelligence-Based Medicine (a distillation of 4 years of class notes from the Stanford program in biomedical data science) for this book's introduction section; this work serves as a concise but comprehensive primer of the basic tenets of artificial intelligence in healthcare. Even though the original text is just 2 years old, that time duration is a lifetime in AI, with new topics such as synthetic data, federated learning, and transformers recently surfacing. The remainder of the book (an amazing 50 chapters) is on AI in various subareas (such as echocardiography, electrophysiology, and cardiac MRI), as well as AI methodologies as utilized in cardiology and cardiac surgery (such as data sharing, natural language processing, and digital twins). Our section editors, Louise Y. Sun, Robert Brisk, and Francisco Lopez-Jimenez, worked diligently on editing these chapters. I am particularly proud that each chapter in this first edition of the book has at least one cardiologist authoring the work. Therefore, I owe a special debt of gratitude to our many authors; I am simply in awe of their pioneering work in these burgeoning areas of AI in cardiovascular medicine. Finally, the compendium at the end of the work compiles many useful references for readers (lists, books and articles, and a glossary).

The book is aptly designed for anyone interested in a comprehensive primer and more on the principles and applications of data science, artificial intelligence, and human cognition toward intelligence-based cardiology and cardiac surgery. The readers that this book is written for include the curious but busy clinician, the interested data/computer scientist, the astute investor, the inquisitive hospital administrator and leader, and any knowledge-seeking patient or family member. This book is not full of mathematical formulae and esoteric data science topics, nor is it filled with medical jargon and superficial descriptions of AI concepts. This book is a hybrid between the clinical cardiology domain and the relevant and relatable AI and data science (in short, a convolution of these two disciplines). In short, it is written with everyone in mind while defining the rapidly growing interface between cardiovascular medicine and artificial intelligence.

Both my daughter Emma and I are direct beneficiaries of excellent cardiac care. With excellent multidisciplinary teams taking extraordinary care of both of us, it was obvious that we often still make many decisions without sufficient data or enough certainty and do not naturally seek out data science or AI as a resource in the present-day imbroglio of healthcare. This theme of uncertainty in clinical medicine is pervasive, even at premier heart programs and medical centers, but can be reduced greatly if some or all of the methodologies described in this book are not only published in elite journals but also put into practice in all heart centers around the world.

The AI portfolio of tools can provide a sanctuary for all those who seek refuge from the escalating burden of biomedical sciences and clinical medicine, including cardiovascular medicine. Deep learning for medical imaging, machine learning for decision support, and cardiomics with combined imaging and biomedical data can all be valuable assets in AI in cardiology and cardiac surgery. Many daunting challenges lie ahead for cardiovascular disease, which remains the biggest disease burden on this planet. At the same time, myriad invigorating discoveries and innovative solutions are evolving to address the many problems we face in cardiology and cardiac surgery. This AI paradigm is a once-in-a-generation opportunity for all of us to work in cohesion to transform the cardiac care of our precious patients for better outcomes, ultimately becoming the North Star for our collective efforts in artificial intelligence for cardiology and cardiac surgery.

Anthony C. Chang

Spring, 2023

Acknowledgments

I would first like to sincerely thank the colleagues who were especially supportive of me in my many AI-in-medicine endeavors: Drs. Alfonso Limon, Sharief Taraman, William Feaster, Robert Hoyt, Timothy Chou, Howard Lei, Terry Sanger, Louise Y. Sun, Robert Brisk, Francisco Lopez-Jimenez, Spyro Mousses, and Louis Ehwerhemuepha. I would also like to express my thanks to our Medical Intelligence, Information, Investigation, and Innovation Institute team members for their unwavering support: Debra Beauregard, Tiffani Ghere, Ashley Perez, Amber Osorno, Jenae Vancura, Vivian Nguyen, and Monica Suesberry. In addition, I would like to express my deepest gratitude to the Sharon Disney Lund Foundation board members, who have supported my vision of having a dedicated institute for artificial intelligence since its inception, particularly Michelle Lund and Robert and Gloria Wilson. I would also like to thank Kimberly Cripe, Kerri Ruppert Schiller, Paul van Dolah, Sandip Godambe, Coleen Cunningham, John Henderson, Doug Corbin, Melanie Patterson, Tom Capizzi, Jay Gabriel, and the entire chief executive suite of Children’s Health of Orange County, all of whom are exemplary in their professionalism.

Several organizations have contributed greatly to my personal and institutional growth in AI in healthcare. I want to thank the core faculty and support group of the American Board of AI in Medicine—Mijanou Pham, Brett McVicker, Dr. Orest Boyko, Dr. May Wang, Dr. Ioannis Kakadiaris, Flora Wan, Dr. Scott Campbell, Dr. Eric Eskioglu, and Eric Smith—as well as the many attendees of these courses and the weekly office hour that have been so interesting and insightful. I would also like to thank the fellow officers and principals of the Medical Intelligence Society: Drs. Piyush Mathur, John Lee, and Hoang H. Nguyen. The Alliance of Centers of AI in Medicine, with the pediatric subgroup Pediatric Centers of AI in Medicine, has been wonderfully supportive of all my efforts in AI in medicine, especially Dr. Curtis Langlotz, Johanna Kim, Dr. Yindalon Aphinyanaphongs, Dr. Muhammad Mamdani, Dr. Johan W. Verjans, and many other leaders of their outstanding centers of AI in medicine. The AI advocates at the International Society for Pediatric Innovation were of great help to me: Drs. Alberto Eugenio Tozzi, Darren Gates, Iain Hennessy, Todd Ponsky, Neil Sebire, and Andrew Taylor.

I would like to thank fellow cardiologists who were especially supportive of me: Drs. Zaidon Al-Falahi, Rima Arnaout, G. Hamilton Baker, Ami Bhatt, Jeffrey P. Jacobs, Kathy Jenkins, Pei-Ni Jone, Wyman Lai, Kevin Maher, Anthony McCanta, Jai Nahar, Mitch Recto, Alessandra Toscano, Tu Hao Tran, Gil Wernovsky, and David Wessel. My professional colleagues, Drs. Afshin Aminian, Amir Ashrafi, Arta Bakshandeh, Leo Celi, Peter Chang, John Cleary, Enrico Coiera, Ken Grant, Peter Holbrook, Matthieu Komorowski, Daniel Kraft, Peter C. Laussen, Diane Nugent, Randall Wetzel, Tony Young, and Neda Zadeh, have all been invaluable to me. To those whose names I have inadvertently left out, please forgive me because your support was cherished as well.

I want to thank my Stanford School of Medicine Biomedical Data Science program mentors (especially Drs. Ted Shortliffe, Nigam Shah, Russ Altman, and Dennis Wall) and classmates and teachers' assistants for their utmost patience and gracious encouragement during my 4-year sojourn as a curious and passionate student exploring the fascinating world of biomedical data science and artificial intelligence. My computer science colleagues at the Chapman University Department of Computer Science, especially Drs. Andrew Lyon, Michael Fahy, Erik Linstead, and Cyril Rakovic, were always available for guidance and counsel.

I would like to express my gratitude to the tireless and dedicated staff of the Artificial Intelligence in Medicine (AIMed) enterprise—Freddy White, Peter Moug, Ruth Carter, Sindy Cain, Alexis May, Andrew Johnson, Suzy White, Ally Baker, Hazel Tang, Gemma Lovegrove, and Toni Jenner—and the many colleagues and friends who have participated as faculty members and attendees at AIMed-related meetings around the world. My colleagues at the education and consulting group Medical Intelligence 10—Steve Ardire, Dr. Arlen Meyers, Dr. David Schneider, Rebecca Wiedemer, Zhen Xu, and Qingxin Zhang—have been extraordinarily encouraging. My colleagues at the Medical Intelligence 1 startup—Brendan Dunphy, Dr. Timothy McLerran, Andrew Nguyen, and Nikesh Shah—have also contributed to my AI journey.

Finally, I would like to thank all those esteemed authors and friends who have kindly contributed insightful chapters to this book and who have so kindly enlightened me as we collectively grow AI within the cardiology community. Finally, I would like to thank Timothy Bennett and Rafael Teixeira for their utmost attention to detail and patience, as well as Justyna Vinci and Judith Escales of Elsevier, who have provided continual encouragement throughout the entire book and journal process.

Anthony C. Chang

Spring, 2023

Section I

Basic concepts of data science and artificial intelligence

Outline

Chapter 1. Introduction to artificial intelligence for cardiovascular clinicians

Chapter 1: Introduction to artificial intelligence for cardiovascular clinicians

Anthony C. Chang ¹ , ² , ³ , and Alfonso Limon ²       ¹ Heart Failure Program, Heart Institute, Children's Health of Orange County, Orange, CA, United States      ² Sharon Disney Lund Medical Intelligence, Information, Investigation, and Innovation Institute (Mi4), Children's Health of Orange County, Orange, CA, United States      ³ Chapman University, Orange, CA, United States

Abstract

The impressive gains in deep learning (DL) started in 2012 and its successful utilization in image interpretation have led to the current momentum for artificial intelligence (AI) awareness and adoption. In 2016, Google DeepMind's AlphaGo software soundly defeated the best human Go champion Lee Sedol to introduce the capability of DL outside of image interpretation. More recently, there have been impressive exponential advances in natural language processing with transformer tools such as GPT-3, GPT-4, and now ChatGPT. DeepMind and its AlphaFold AI tool has been able to predict the three-dimensional (3D) structure of proteins since 2021 and was Science magazine's Breakthrough of the Year. All of these AI accomplishments heralded the recent new era in AI. Major universities with AI departments (such as Stanford, MIT, and Carnegie Mellon) and technology giants (such as IBM, Apple, Facebook, and Microsoft in the United States as well as other large companies such as Baidu, Alibaba, and Tencent [BAT] in China) are all fervidly exploring real-life applications of AI. There is also a movement to democratize AI so that no-code platforms can accommodate people who do not know how to code [1].

Keywords

Artificial intelligence; Cardiovascular clinicians; Deep learning technology; Human-machine intelligence continuum; Machine learning; Neuroscience

If you want to build a ship, don't drum up the men to gather wood, divide the work, and give orders. Instead, teach them to yearn for the vast and endless sea.

Antoine de Saint-Exupery, French pilot and author of Le Petit Prince

The impressive gains in deep learning (DL) started in 2012 and its successful utilization in image interpretation have led to the current momentum for artificial intelligence (AI) awareness and adoption. In 2016, Google DeepMind's AlphaGo software soundly defeated the best human Go champion Lee Sedol to introduce the capability of DL outside of image interpretation. More recently, there have been impressive exponential advances in large language models with transformer tools such as GPT-3, GPT-4, and now ChatGPT. DeepMind and its AlphaFold AI tool has been able to predict the three-dimensional (3D) structure of proteins since 2021 and was Science magazine's Breakthrough of the Year. All of these AI accomplishments heralded the recent new era in AI. Major universities with AI departments (such as Stanford, MIT, and Carnegie Mellon) and technology giants (such as IBM, Apple, Facebook, and Microsoft in the United States as well as other large companies such as Baidu, Alibaba, and Tencent [BAT] in China) are all fervidly exploring real-life applications of AI. There is also a movement to democratize AI so that no-code platforms can accommodate people who do not know how to code [1].

Even though the advent of data science as well as machine learning (ML) and DL has advanced information and analyses in domains such as finance, marketing, and even sports and fostered innovations such as virtual assistants, autonomous cars, and works of digital art and music; healthcare and medicine including cardiology and cardiac surgery remain behind these other domains in leveraging this new AI paradigm. The recent major escalation of venture capital into healthcare and AI domains, however, promulgated over 100 companies in AI in healthcare with an expected $50 billion to be spent on AI in healthcare by 2025 with more than $100 billion in savings. Private investment in AI overall increased to close to $100 billion in 2021 even though the number of AI companies decreased, indicating intensification of the investment concentration [2].

Since the first article published in the domain of AI in biomedicine in 1958 [3], there has been a relative paucity of published reports focused on AI in medical journals and a concomitant lack of serious interest among most clinicians in applications of AI in medicine, including cardiology and cardiac surgery, until very recently. In 2023, there will be a projected more than 50,000 publications on AI applications in medicine (under a myriad of AI-related search terms such as artificial intelligence, machine learning, deep learning, cognitive computing, natural language processing, etc.), including only about 1000–2000 or so in cardiology and cardiac surgery (out of about 75,000–100,000 total articles in these areas in cardiovascular medicine, or about 1%). Finally there is publication activity only in the past few years in the more prestigious journals (including those in cardiology) [4–8].

We all face the imbroglio of healthcare with its complex ecosystem and data in disarray, and this has led to a significant rise in professional burnout among its caretakers. We have a once-in-a-generation opportunity to capture this robust AI resource for clinical medicine and healthcare, and potentially