An up-to-date systematic review on machine learning approaches for predicting treatment response in diabetes

Diabetes mellitus (DM) is defined as a group of metabolic disorders characterized by a long-term high blood sugar level caused by abnormal insulin secretion and/or action. Different medications have been developed but the treatment efficacy is patient-specific. The evidence-based prediction of DM treatment response can provide specific reference for self-management, clinical intervention and medication. Recently, some machine learning models have been proposed for the diagnosis of DM. Whereas, the applications in predicting treatment response are limited. The data-driven approach empowered by machine learning enables patient-tailored therapy based on multimodal big health data analysis. In this chapter, we overviewed the state-of-the-art machine learning techniques regarding the data, algorithm, and performance. We summarized the advantages, limitations, and future directions. This chapter provides an up-to-date reference for clinicians, data scientists, and biomedical engineers to improve the treatment for DM patients.