Contrastive Graph Learning for Population-based fMRI Classification

TL;DR

This paper introduces a novel contrastive graph learning approach for population-based fMRI classification, leveraging functional connectivity graphs to improve accuracy and reveal patient subtypes, outperforming existing methods.

Contribution

It proposes a contrastive functional connectivity graph learning method that considers interpatient relationships and dynamic population graphs for enhanced fMRI classification.

Findings

Outperforms existing contrastive methods on ADHD200 dataset

Visualizes population relationships and identifies potential subtypes

Demonstrates improved classification metrics

Abstract

Contrastive self-supervised learning has recently benefited fMRI classification with inductive biases. Its weak label reliance prevents overfitting on small medical datasets and tackles the high intraclass variances. Nonetheless, existing contrastive methods generate resemblant pairs only on pixel-level features of 3D medical images, while the functional connectivity that reveals critical cognitive information is under-explored. Additionally, existing methods predict labels on individual contrastive representation without recognizing neighbouring information in the patient group, whereas interpatient contrast can act as a similarity measure suitable for population-based classification. We hereby proposed contrastive functional connectivity graph learning for population-based fMRI classification. Representations on the functional connectivity graphs are "repelled" for heterogeneous patient pairs meanwhile homogeneous pairs "attract" each other. Then a dynamic population graph that strengthens the connections between similar patients is updated for classification. Experiments on a multi-site dataset ADHD200 validate the superiority of the proposed method on various metrics. We initially visualize the population relationships and exploit potential subtypes.