Abhinav Suri - Practical AI for Healthcare Professionals: Machine Learning with Numpy, Scikit-learn, and TensorFlow

This Apress imprint is published by the registered company APress Media, LLC part of Springer Nature.

The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Over the years ahead, artificial intelligence (AI) will play an ever-increasing and ultimately a transformative role for medicines future. Nearly every week, we are seeing peer-reviewed studies that demonstrate the potential of deep neural networks for improving the accuracy of interpretation of medical images, from scans to slides to skin abnormalities to real-time machine vision pickup of colon polyps during endoscopy. Beyond medical images, algorithms are getting validated for patients, capturing their own data, coupled with algorithmic assistance, to facilitate the diagnosis of heart rhythm abnormalities, urinary tract infections, ear infections in children, and many other common reasons that would require a visit to a doctor. This early phase of medical AI will inevitably progress with validation via prospective and randomized clinical trials that are sorely lacking at this juncture. As Antonio de Leva wrote in The Lancet, Machines will not replace physicians, but physicians using AI will soon replace those not using it.

But how will physicians get up to speed and learn about this field, which has undergone so much rapid change in the past decade owing to the subtype of AI known as deep learning (DL)? In this new book, Abhinav Suri, a medical student at UCLA, has provided an outstanding primer for uninitiated clinicians. Abhinav has the perfect background for this: a double degree in computer science and biology from Penn, an MPH degree from Columbia, and additional experience leading medical scan AI research at the Perelman School of Medicine. In just seven chapters, he succinctly lays out the basics and delineates the limits and potential flaws of AI, the different types of machine learning (ML) algorithms and deep neural networks, and snake oil AI. Weve needed such a book for the medical community to get grounded, not so that physicians can code, but rather to understand the power, nuances, and limitations as AI makes its way deeper into the practice of medicine.

Undoubtedly, we will see more educational tools to help promote understanding and optimal use of AI in healthcare over the years ahead. The main textbook of the overall field is Deep Learning by Ian Goodfellow and colleagues, but it is quite comprehensive and well suited for people who intend to code and get deep into neural networks. Suris new book sets a very good standard for the goal of getting a quick and pragmatic introduction into AI, catering to the specific needs of clinicians. It will get you to think like a computer which is a requisite step to get grounded. Awareness of the basics and nuances of AI will eventually become a standard part of every medical school curriculum, and this primer may be considered a very good start in that direction.

Eric J. Topol, MD

Professor and EVP of Scripps Research

Author, Deep Medicine

La Jolla, California