A Dual-Attention Graph Network for fMRI Data Classification

2025-08-18

AI Summary

To address the limitation of static functional connectivity modeling in capturing the dynamic nature of brain activity from fMRI data, this paper proposes a Dual-Attention Graph Network (DAGN) for autism spectrum disorder (ASD) classification. Methodologically, DAGN integrates attention-driven dynamic graph construction with a hierarchical spatiotemporal attention mechanism: it first infers time-varying functional connectivity using a Transformer, then adaptively learns dynamic adjacency relationships via node-level attention; subsequently, it jointly employs GCN and a spatiotemporal Transformer to co-model both regional topological structure and temporal evolution. Evaluated on the ABIDE dataset, DAGN achieves 63.2% accuracy and 60.0% AUC—significantly outperforming static GCN (51.8%). Its core contribution lies in being the first to unify dynamic graph generation and spatiotemporal attention within an end-to-end graph learning framework, thereby enhancing selective modeling of salient brain regions and critical temporal segments.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Understanding the complex neural activity dynamics is crucial for the development of the field of neuroscience. Although current functional MRI classification approaches tend to be based on static functional connectivity or cannot capture spatio-temporal relationships comprehensively, we present a new framework that leverages dynamic graph creation and spatiotemporal attention mechanisms for Autism Spectrum Disorder(ASD) diagnosis. The approach used in this research dynamically infers functional brain connectivity in each time interval using transformer-based attention mechanisms, enabling the model to selectively focus on crucial brain regions and time segments. By constructing time-varying graphs that are then processed with Graph Convolutional Networks (GCNs) and transformers, our method successfully captures both localized interactions and global temporal dependencies. Evaluated on the subset of ABIDE dataset, our model achieves 63.2 accuracy and 60.0 AUC, outperforming static graph-based approaches (e.g., GCN:51.8). This validates the efficacy of joint modeling of dynamic connectivity and spatio-temporal context for fMRI classification. The core novelty arises from (1) attention-driven dynamic graph creation that learns temporal brain region interactions and (2) hierarchical spatio-temporal feature fusion through GCNtransformer fusion.

Problem

• Classifying fMRI data for Autism Spectrum Disorder diagnosis
• Capturing dynamic spatio-temporal relationships in brain activity
• Overcoming limitations of static functional connectivity methods

Innovation

• Dynamic graph creation with attention mechanisms
• Spatiotemporal attention for brain region focus
• GCN-transformer fusion for hierarchical feature extraction