Background Depression has emerged as a significant public health concern in modern society, while conventional medical approaches exhibit notable limitations in its diagnosis and treatment. With the rapid advancement of technologies such as large-scale artificial intelligence (AI) models, the application of AI to drive research and innovation in mental health—and to further its translational implementation in clinical practice—has garnered increasing attention. Objective Focusing on the intersection of AI and depression, this bibliometric study aims to reveal key themes and developmental trajectories within this domain. Methods A bibliometric analysis was performed on publications retrieved from the Web of Science Core Collection. Utilizing the Bibliometrix package, CiteSpace, and VOSviewer, we quantified publication output, collaboration networks, and key research themes. The analysis encompassed English-language articles published between 2011 and 2024. Results A total of 1361 publications met the inclusion criteria. Since 2011, the annual number of publications on AI in depression has demonstrated a steady upward trajectory, entering a phase of accelerated growth after 2020. The peak annual output exceeded 300 articles. Analysis of geographical and institutional contributions identified China and the United States as the leading countries, with Lanzhou University being the most productive institution. The most prolific authors were Bin Hu, Xiaowei Li, and Raymond W. Lam, while R.C. Kessler emerged as the most cocited author. Keywords such as “machine learning,” “deep learning,” and “feature extraction” showed a marked increase in frequency, reflecting the field's evolving technical focus. Furthermore, the development of international collaborations underscores the increasingly globalized nature of research in this domain. Conclusion This study presents the latest comprehensive bibliometric analysis results of AI in the field of depression, clarifying current research hotspots and directions, and providing important resources for clinicians and researchers.
Tian et al. (Sun,) studied this question.