Using Graph Convolutional Networks to Address fMRI Small Data Problems
Thomas Screven, Andras Necz, Jason Smucny, Ian Davidson
TL;DR
The paper addresses prognosis prediction from task-based fMRI in a small-data regime by introducing a six-view spectral GNN framework. It builds a population graph from phenotypic similarity and uses spectrally embedded ROI connectivity as node features, with predictions aggregated via majority voting across views. The results show substantial improvements over deep neural network baselines and reveal that gains arise in part from smoothing the data and reducing triangle-inequality violations through graph propagation. Two key innovations—spectral embedding with $k=10$ eigenvectors and graph-based smoothing over subject similarities—enable reliable prognosis in an 82-subject dataset, with potential clinical impact for small-sample neuroimaging studies.
Abstract
Although great advances in the analysis of neuroimaging data have been made, a major challenge is a lack of training data. This is less problematic in tasks such as diagnosis, where much data exists, but particularly prevalent in harder problems such as predicting treatment responses (prognosis), where data is focused and hence limited. Here, we address the learning from small data problems for medical imaging using graph neural networks. This is particularly challenging as the information about the patients is themselves graphs (regions of interest connectivity graphs). We show how a spectral representation of the connectivity data allows for efficient propagation that can yield approximately 12\% improvement over traditional deep learning methods using the exact same data. We show that our method's superior performance is due to a data smoothing result that can be measured by closing the number of triangle inequalities and thereby satisfying transitivity.