Machine Learning Applications in Endocrinology and Metabolism Research: An Overview

Abstract

Machine learning (ML) applications have received extensive attention in endocrinology research during the last decade. This review summarizes the basic concepts of ML and certain research topics in endocrinology and metabolism where ML principles have been actively deployed. Relevant studies are discussed to provide an overview of the methodology, main findings, and limitations of ML, with the goal of stimulating insights into future research directions. Clear, testable study hypotheses stem from unmet clinical needs, and the management of data quality (beyond a focus on quantity alone), open collaboration between clinical experts and ML engineers, the development of interpretable high-performance ML models beyond the black-box nature of some algorithms, and a creative environment are the core prerequisites for the foreseeable changes expected to be brought about by ML and artificial intelligence in the field of endocrinology and metabolism, with actual improvements in clinical practice beyond hype. Of note, endocrinologists will continue to play a central role in these developments as domain experts who can properly generate, refine, analyze, and interpret data with a combination of clinical expertise and scientific rigor.

Keywords: Adrenal; Artificial intelligence; Deep learning; Diabetes; Endocrinology; Machine learning; Metabolism; Osteoporosis; Pituitary; Thyroid.

Fig. 1. The increasing trend in the number of artificial intelligence or machine learning-related publications per year in the endocrinology and metabolism field. The included publications were confined to PubMed-indexed records until the search date (January 17th, 2020), with combinations of search terms including machine learning, artificial intelligence, deep learning, endocrinology, metabolism, diabetes, pituitary, thyroid, adrenal gland, and osteoporosis, using PubMed query as follows: search ((((((“Machine Learning”[Mesh]) OR “Artificial Intelligence”[Mesh]) OR “Deep Learning”[Mesh])) OR (((machine learning[Title/Abstract]) OR artificial intelligence[Title/Abstract]) OR deep learning[Title/Abstract]))) AND ((((((((endocrinology[Title/Abstract]) OR diabetes[Title/Abstract]) OR pituitary[Title/Abstract]) OR thyroid[Title/Abstract]) OR adrenal gland[Title/Abstract]) OR osteoporosis[Title/Abstract])) OR ((((((“Endocrinology”[Mesh]) OR “Diabetes Mellitus”[Mesh]) OR “Pituitary Gland”[Mesh]) OR “Thyroid Gland”[Mesh]) OR “Adrenal Glands”[Mesh]) OR “Osteoporosis”[Mesh])).

Fig. 2. A brief workflow of machine learning-based medical research.

Fig. 3. Top 30 frequently appeared words in the titles of machine learning (ML)-based endocrinology studies between 2015 and 2019. Among a total of 2028 literatures searched by PubMed querya on Jan 17th, 2020, text analysis was performed with nouns and adjectives parsed from the titles of 611 studies (English language, human study without review or meta-analysis) published within last 5 years. Cumulative counts of appearance of top 30 words were plotted as horizontal bar plot. Frequently appeared diseases and ML tasks were plotted as pie charts separately. aPubMed query: (Search ((((((“Machine Learning”[Mesh]) OR “Artificial Intelligence”[Mesh]) OR “Deep Learning”[Mesh])) OR (((machine learning[Title/Abstract]) OR artificial intelligence[Title/Abstract]) OR deep learning[Title/Abstract]))) AND ((((((((endocrinology[Title/Abstract]) OR diabetes[Title/Abstract]) OR pituitary[Title/Abstract]) OR thyroid[Title/Abstract]) OR adrenal gland[Title/Abstract]) OR osteoporosis[Title/Abstract])) OR ((((((“Endocrinology”[Mesh]) OR “Diabetes Mellitus”[Mesh]) OR “Pituitary Gland”[Mesh]) OR “Thyroid Gland”[Mesh]) OR “Adrenal Glands”[Mesh]) OR “Osteoporosis”[Mesh])).