Multi-Modal Graph Neural Network with Transformer-Guided Adaptive Diffusion for Preclinical Alzheimer Classification

2026-06-02

AI Summary

Existing graph neural networks struggle to simultaneously integrate multimodal features, preserve critical node information, and capture long-range dependencies in brain network modeling, hindering the early identification of preclinical Alzheimer’s disease. This work proposes a multimodal graph neural network incorporating a Transformer-guided adaptive diffusion mechanism that aggregates local neighborhood information via diffusion kernels and models long-range dependencies through multi-head self-attention. By dynamically regulating information propagation at the node level, the method effectively balances the preservation of salient brain region features with global topological learning. The proposed approach significantly outperforms current models in preclinical Alzheimer’s classification, achieving higher accuracy while reliably identifying key brain regions strongly associated with early-stage pathology, thereby offering a robust foundation for early diagnosis.

Abstract

The graphical representation of the brain offers critical insights into diagnosing and prognosing neurodegenerative disease via relationships between regions of interest (ROIs). Despite recent emergence of various Graph Neural Networks (GNNs) to effectively capture the relational information, there remain inherent limitations in interpreting the brain networks. Specifically, convolutional approaches ineffectively aggregate information from distant neighborhoods, while attention-based methods exhibit deficiencies in capturing node-centric information, particularly in retaining critical characteristics from pivotal nodes. These shortcomings reveal challenges for identifying disease-specific variation from diverse features from different modalities. In this regard, we propose an integrated framework guiding diffusion process at each node by a downstream transformer where both short- and long-range properties of graphs are aggregated via diffusion-kernel and multi-head attention respectively. We demonstrate the superiority of our model by improving performance of pre-clinical Alzheimer's disease (AD) classification with various modalities. Also, our model adeptly identifies key ROIs that are closely associated with the preclinical stages of AD, marking a significant potential for early diagnosis and prevision of the disease.

Problem

• Alzheimer's disease
• graph neural network
• multi-modal
• brain network
• preclinical classification

Innovation

• Multi-Modal Graph Neural Network
• Transformer-Guided Diffusion
• Adaptive Diffusion
• Preclinical Alzheimer's Classification
• Brain ROI Analysis