Review of Artificial Intelligence (AI) in Healthcare

Resource Review Post by Marcio A. Diniz from Cedars Sinai Medical Center

The book Artificial Intelligence (AI) in Healthcare by Adam Bohr and Kaveh Memarzadeh is a collection of articles discussing the current applications of AI in the medical area such as development of new drugs, decisions tools for diagnostic and treatment, interpretation of medical imaging, remote patient monitoring and conduction of assisted surgeries. It is organized in 12 stand-alone chapters written by several different authors with expertise spanning the AI field.

The first two chapters introduce the healthcare system crisis due increasing costs and aging population, and the large amount of data that has been collected intentionally or not with the adoption of new technologies - electronic medical records and wearable devices among others - as a unique opportunity to take advantages of AI tools such as natural language processing and deep learning to improve patient care. In the chapters 3 to 9, authors discuss specific applications in medical research focused on patient care with an overview of the current literature and discussion of case studies. In the last three chapters – 10 to 12 – authors discuss specific applications focused on healthcare providers such as security, privacy, insurance, and the ethical and legal challenges that will need to be addressed to achieve a widespread use of AI tools in the healthcare system.

Although the authors state that the book is written for a broad audience including clinicians, health and life science professionals, policymakers, business leader, university students and patients with visible efforts made to intuitively introduce AI concepts and tools, readers would greatly benefit of some introductory level background in machine learning to fully appreciate the discussion of specific applications. Furthermore, the discussion about the implementation of these novel AI tools in clinical practice seems limited to technical issues, ignoring the most recent debate in the literature about the need to conduct clinical trials to rigorously test whether the implementation of specific AI tools in clinical practice bring benefit in clinical outcomes which goes beyond the standard performance measures used to evaluate machine learning algorithms.

The text could be a great tool as support material when organizing journal clubs of graduate programs in data science creating the opportunity for students to get familiar with the current stage of AI tools applied to healthcare and potential topics of research as every chapter presents concluding remarks indicating future directions of the field. Moreover, each chapter also presents a large list of references that can be further explored to dive in on specific topics. Overall, the book is well-written with illustrations to facilitate the understanding of AI concepts and could be considered a source material that provides a general perspective of AI and its contribution to the future of healthcare.

The link to the book is here.