Advancing Mental Health Research with Graph Neural Networks: A Comprehensive Survey

作者：Wei Ma, Zihan Su, Q. B. Chen, Hanshu Zhai, Juanyuan Jiang, Shi Han · 年份：2025 · DOI：10.1109/icdew67478.2025.00030 · 被引用次数：2 · 研究领域：Mental Health Research Topics、Functional Brain Connectivity Studies、Mental Health via Writing

Graph Neural Networks (GNNs) have shown substantial potential in advancing mental health research, particularly in understanding disease mechanisms, improving diagnostic accuracy, predicting treatment outcomes, and enhancing public health strategies. GNNs excel at integrating diverse data types, such as neuroimaging, genetic, and clinical data, to model complex brain networks and disease relationships. This survey reviews the application of GNNs across five critical areas: (1) elucidating disease mechanisms, (2) diagnosis and classification of mental disorders, (3) risk prediction and intervention, (4) personalized medicine and treatment outcome evaluation, and (5) community prevention strategies. Despite challenges such as data heterogeneity, interpretability, and model scalability, GNNs offer promising solutions for more accurate, personalized, and dynamic mental health care.