Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again

0
1790

Eric Topol, MD
Basic Books
Hachette Book Group
New York, 2019


“Deep Medicine” is the book on artificial intelligence (AI) that all neurosurgeons need to read. Eric Topol, MD, is a cardiologist who understands health care from the inside. This is his third, and best, book on the future of health care. He understands that the electronic health record (EHR) was a disaster and he appreciates that it was designed for billing and also that it was responsible for much of the burnout and attrition in our profession. He believes that AI will put us back on the track of caring for patients.

Dr. Topol appreciates all the hype about AI and he understands the outlandish expectations of some optimists. He begs for realism and knows that AI can catalyze advances in basic science and drug development. His conclusion is that deep medicine is both possible and highly desirable.

o

The way we are now practicing leaves little room for meaningful improvement. Evidence suggests that even when we are completely certain of the diagnosis, we are wrong 40% of the time. Unfortunately, this uncertainty has never been acknowledged in medicine. We do have room for improvement.

The book begins with a refreshing review of the history of AI and a nice definition of all the terms one needs to understand. Neurosurgeons will especially appreciate the emphasis on neural network methodology and deep neural networks (DNN), which are at the heart of deep learning. The relationship between neuroscience and AI has been one where growing knowledge and breakthroughs in a single field affects several others. For example, understanding the brain has helped unlock problems in computer science. A striking achievement in AI with respect to comprehending brain function has been to model special navigation. It has been predicted that the 21st century will be the century of the brain. The brain is the most complex piece of highly excitable matter in the known universe. Machine learning and human brain processing complement each other and should yield high levels of accuracy.

AI will make it possible to build the virtual medical assistant of the future. Neither patients nor physicians will be able to process all the data necessary for optimal care. The virtual medical assistant’s ultimate success will be predicated on changing human behavior, because so much of the burden of disease is related to poor lifestyle. The virtual medical coach can promote self-driving healthy humans and be more important than self-driving cars.

But the future of health care is not just managing data. We need to restore the primacy of the physical exam. The neurological exam can be the mechanism to help restore a proper physician-patient relationship. Technology must not lead to human disconnection. AI has the potential to bring back real medicine: Presence. Empathy. Trust. Caring. Being human.

[aans_authors]

Print Friendly, PDF & Email
o