Classification of Depression, Anxiety, and Quality of Life in Diabetic Patients with Machine Learning: Systematic Review

Background: Diabetes is a chronic and costly disease that is emerging in low- and middle-income countries. Current research suggests that diabetes may be associated with depression, anxiety, and/or impaired quality of life. Therefore, assessment of these psychological, mental, physical, and social aspects can significantly improve overall diabetes care services. Recently, studies have been conducted on the classification of these components using new machine learning (ML) techniques. Objective: To summarize existing findings on machine learning models classifying diabetics with depression, or anxiety, or quality of life. Methods: Systematic review of original research between January 2010 and May 2023. A search covered three databases on (Scopus), (Web of Science), and (PubMed). Results: From 126 search results, 4 articles were selected after using the eligibility criteria: 3 machine learning models classifying subjects with versus without depression, and 1 model classifying subjects with versus without pain (Pain as a quality of life dimension). Conclusion: Almost all reported machine learning models have shown optimal performance results, although they need to be standardized and comparable. Furthermore, it is strongly recommended that they be improved and implemented in research and clinical practice.