The Application of Graph Neural Networks in Brain Disorders: A Bibliometric Analysis

Graph Neural Networks (GNNs) are deep learning models specifically developed to handle graph-structured data, and they have been extensively used in areas including social networks, physical systems, financial modeling, and molecular analysis. In recent years, GNNs have also demonstrated significant potential in medical applications, particularly in the diagnosis and monitoring of neurological disorders, by enabling the modeling of complex spatial and topological relationships in brain networks that conventional neural networks cannot effectively capture. Given the rapid growth of research on GNN-based neuroimaging applications, a systematic bibliometric analysis is essential to identify research trends, influential contributions, and emerging opportunities. Bibliometric analysis provides a quantitative approach to mapping the intellectual structure and evolution of a research field. This study presents a comprehensive bibliometric analysis of publications on the application of GNNs in brain disorder diagnosis. A total of 548 articles published between 2019 and 2026 were retrieved from the Web of Science database. Titles, abstracts, and author keywords were analyzed to identify research hotspots, key contributors, and collaboration patterns. The results indicate that GNN-based methods have been predominantly applied to Alzheimer's disease, followed by autism spectrum disorder and Parkinson's disease. Furthermore, China, the United States, and the United Kingdom were identified as the leading countries in terms of scientific output. These findings highlight the rapid expansion of GNN-based neuroimaging research and provide insights into future research directions in this emerging field.