Machine learning in digital health, recent trends, and ongoing challenges

Abstract As a result of a growing and aging population, as well as an increase in associated costs, there is a continual stretching of health care services worldwide. This issue is motivating researchers all over the world to optimize medical resources by utilizing digital tools explicitly addressed to health care and well-being. One of the main fields of research in this regard is artificial intelligence (AI), the endowment of machines with human-like learning, reasoning, and decision-making abilities. Combined with high penetration of sensor-based technologies—such as smartphones and wearables—in modern society, advancements in AI mean we are entering a new age of health care. Soon, we will be able to monitor vital signs and lifestyle habits, in real-time, in such a way that will help clinicians to monitor patients’ evolution and progress in a nonintrusive and remote manner. This chapter intended to be an introductory, higher-level overview, of the core concepts relating to the branch of AI known as machine learning (ML). Readers are introduced to the ML train-test pipeline and given an overview of commonly used ML algorithms. The chapter finishes by discussing challenges that need to be overcome to help fully realize the potential of ML in everyday digital health settings.