The study aimed to analyze and discuss the changing trends in the academic production on the topic of the use of deep learning and machine learning methods to learn about mental illnesses. The scopus database was searched to identify relevant publication that were gathered concerning the PRISMA guidelines of citation screening and selection. Among 625 articles published to date (2020-2025), 279 articles were used in the analysis. The co-authorship network in countries showed that 33 countries were active in a total of 64 countries. United States, China, and India were described to be centers of power, with the most amounts of total link strength, which indicates their central influences of international research collaboration. The co-occurrence analysis of the terms revealed that Deep Learning and Machine Learning were in the middle of the conceptual network, which highlights their importance in the given field of research. Very related words, such as Mental Health and Depression, were placed close to each other which are indications of thematic interconnections. In addition, the well-known scholarly journals like the American Journal of Psychiatry, the Schizophrenia Bulletin and the Journal of Affective Disorders have been discovered as the sources of core information to be included into the intellectual background of the sphere.
Al-Tarawneh et al. (2026) studied this question.