MM-GTUNets: Unified Multi-Modal Graph Deep Learning for Brain Disorders Prediction

TL;DR

This paper introduces MM-GTUNets, a novel multi-modal graph transformer framework that enhances brain disorder prediction by adaptively modeling complex inter-modal relationships and leveraging rich multi-modal data.

Contribution

The paper proposes MM-GTUNets, integrating Modality Reward Representation Learning and Adaptive Cross-Modal Graph Learning for improved multi-modal brain disorder prediction.

Findings

Outperforms existing methods on ABIDE and ADHD-200 datasets.

Effectively captures modality-specific and shared features.

Demonstrates robustness at large-scale graph modeling.

Abstract

Graph deep learning (GDL) has demonstrated impressive performance in predicting population-based brain disorders (BDs) through the integration of both imaging and non-imaging data. However, the effectiveness of GDL based methods heavily depends on the quality of modeling the multi-modal population graphs and tends to degrade as the graph scale increases. Furthermore, these methods often constrain interactions between imaging and non-imaging data to node-edge interactions within the graph, overlooking complex inter-modal correlations, leading to suboptimal outcomes. To overcome these challenges, we propose MM-GTUNets, an end-to-end graph transformer based multi-modal graph deep learning (MMGDL) framework designed for brain disorders prediction at large scale. Specifically, to effectively leverage rich multi-modal information related to diseases, we introduce Modality Reward Representation Learning (MRRL) which adaptively constructs population graphs using a reward system. Additionally, we employ variational autoencoder to reconstruct latent representations of non-imaging features aligned with imaging features. Based on this, we propose Adaptive Cross-Modal Graph Learning (ACMGL), which captures critical modality-specific and modality-shared features through a unified GTUNet encoder taking advantages of Graph UNet and Graph Transformer, and feature fusion module. We validated our method on two public multi-modal datasets ABIDE and ADHD-200, demonstrating its superior performance in diagnosing BDs. Our code is available at https://github.com/NZWANG/MM-GTUNets.