An objective quantitative diagnosis of depression using a local-to-global multimodal fusion graph neural network

Summary

This study developed an artificial intelligence (AI) system using a local-global multimodal fusion graph neural network (LGMF-GNN) to address the challenge of diagnosing major depressive disorder (MDD), a complex disease influenced by social, psychological, and biological factors. Utilizing functional MRI, structural MRI, and electronic health records, the system offers an objective diagnostic method by integrating individual brain regions and population data. Tested across cohorts from China, Japan, and Russia with 1,182 healthy controls and 1,260 MDD patients from 24 institutions, it achieved a classification accuracy of 78.75%, an area under the receiver operating characteristic curve (AUROC) of 80.64%, and correctly identified MDD subtypes. The system further discovered distinct brain connectivity patterns in MDD, including reduced functional connectivity between the left gyrus rectus and right cerebellar lobule VIIB, and increased connectivity between the left Rolandic operculum and right hippocampus. Anatomically, MDD is associated with thickness changes of the gray and white matter interface, indicating potential neuropathological conditions or brain injuries.

Highlights

• •

  Local-global network architecture offers a comprehensive view on depression diagnosis

• •

  Integrating functional, structural, and demographic data improves depression diagnosis

• •

  Interpretable AI identifies potential digital biomarkers for depression diagnosis

• •

  Achieves superior performance on depression diagnosis in a multi-center scenario

The bigger picture

Major depressive disorder (MDD) presents a multifaceted challenge to global mental health, given its intricate etiology involving social, psychological, and biological determinants. There is a lack of definitive diagnostic markers, and doctors therefore rely heavily on subjective methods to diagnose individuals. Interest in artificial intelligence (AI)-powered diagnostic methods is increasing because of their potential to offer a more comprehensive and objective evaluation. Current AI methods, however, often neglect the complex interplay of functional, structural, and demographic factors that characterize MDD. They seldom integrate insights from both in-depth brain region analysis and broad population-level associations. Moreover, there is a lack of strategies that can interpret AI models and identify diagnostic markers. There is, therefore, an urgent need to improve network architectures, develop effective multimodal fusion strategies, and enhance model interpretability.

Abstract

This study introduces an artificial intelligence (AI) system using local-global multimodal fusion graph neural networks for depression diagnosis. The system integrates functional and structural neuroimaging with health records, demonstrating state-of-the-art performance in diagnosing depression and its subtypes. Built and validated on large-scale, multi-center cohorts with 2,442 participants, the system reveals abnormal brain regions and connectivity patterns, along with digital structural features in depression patients. This advance contributes to the pursuit of objective diagnostic markers and better clinical diagnosis.