Meta-modulation Network for Domain Generalization in Multi-site fMRI Classification

In general, it is expected that large amounts of functional magnetic resonance imaging (fMRI) would be helpful to deduce statistically meaningful biomarkers or to build generalized predictive models for brain disease diagnosis. However, the site-variation inherent in rs-fMRI hampers the researchers to use the entire samples collected from multiple sites because it involves the unfavorable heterogeneity in data distribution, thus negatively impact on identifying biomarkers and making a diagnostic decision. To alleviate this challenging multi-site problem, we propose a novel framework that adaptively calibrates the site-specific features into site-invariant features via a novel modulation mechanism. Specifically, we take a learning-to-learn strategy and devise a novel meta-learning model for domain generalization, i.e., applicable to samples from unseen sites without retraining or fine-tuning. In our experiments over the ABIDE dataset, we validated the generalization ability of the proposed network by showing improved diagnostic accuracy in both seen and unseen multi-site samples.