Estimating Brain Functional Networks Based on Spatiotemporal Higher-Order Correlations for Autism Identification

Brain functional network (BFN) has become an important tool for the analysis and diagnosis of brain diseases, how to build a high-quality BFN based on resting-state magnetic resonance imaging (rs-fMRI) growing concern in neuroscience community. Although some methods have been proposed construct BFN, they only encode spatial characteristics ROIs, ignoring temporal characteristics. As result, it becomes challenging accurately capture true state brain. To address this problem, we propose novel method higher-order considering both domain In particular, get by differentiating rs-fMRI signal itself, then integrate information high-order BFN. evaluate method, conduct our experiments ABIDE database identify subjects with Autism Spectrum Disorder (ASD) from normal controls. Experimental results show that can achieve higher performance than baseline methods.