Multiplex Graph Networks for Multimodal Brain Network Analysis

TL;DR

This paper introduces MGNet, a multiplex graph convolutional network that effectively analyzes multimodal brain networks by integrating tensor representations, achieving state-of-the-art classification results on challenging datasets.

Contribution

The paper presents a novel multiplex GCN model with tensor integration for multimodal brain network analysis, enhancing understanding of the human connectome across modalities.

Findings

Achieves state-of-the-art classification accuracy on HIV and Bipolar disorder datasets.

Effectively captures common and specific graph structures in multimodal brain data.

Provides a new tool for neuroscience research and brain disorder diagnosis.

Abstract

In this paper, we propose MGNet, a simple and effective multiplex graph convolutional network (GCN) model for multimodal brain network analysis. The proposed method integrates tensor representation into the multiplex GCN model to extract the latent structures of a set of multimodal brain networks, which allows an intuitive 'grasping' of the common space for multimodal data. Multimodal representations are then generated with multiplex GCNs to capture specific graph structures. We conduct classification task on two challenging real-world datasets (HIV and Bipolar disorder), and the proposed MGNet demonstrates state-of-the-art performance compared to competitive benchmark methods. Apart from objective evaluations, this study may bear special significance upon network theory to the understanding of human connectome in different modalities. The code is available at https://github.com/ZhaomingKong/MGNets.