AUTHOR=Chu Ying , Wang Guangyu , Qiao Lishan 

TITLE=Multi-Scale Graph Representation Learning for Autism Identification With Functional MRI

JOURNAL=Frontiers in Neuroinformatics

VOLUME=Volume 15 - 2021

YEAR=2022

URL=https://www.frontiersin.org/journals/neuroinformatics/articles/10.3389/fninf.2021.802305

DOI=10.3389/fninf.2021.802305

ISSN=1662-5196

ABSTRACT=Resting-state functional magnetic resonance imaging (rs-fMRI) has been widely used for early diagnosis of autism spectrum disorder (ASD). With rs-fMRI, the functional connectivity networks (FCNs) is usually constructed for representing each subject, with each element representing the pairwise relationship between brain region-of-interests (ROIs). 
Previous studies often first extract handcrafted network features (such as node degree and clustering coefficient) from FCNs and then construct a prediction model for ASD diagnosis, which largely requires expert knowledge. Graph convolutional networks (GCNs) have recently been employed to jointly perform FCN feature extraction and ASD identification in a data-driven manner. However, existing studies tend to focus on single-scale topology of FC networks by using one single atlas for ROI partition, thus ignoring potential complementary topology information of FC networks at different spatial scales. In this paper, we develop a multi-scale graph representation learning (MGRL) framework for rs-fMRI based ASD diagnosis. 
The MGRL consists of three major components: 
1) multi-scale FC network construction using multiple brain atlases for ROI partition, 
2) FC network representation learning via multi-scale GCNs, 
and 3) multi-scale feature fusion and classification for ASD diagnosis.
The proposed MGRL is evaluated on 184 subjects from the public ABIDEI database with rs-fMRI scans. 
Experimental results suggest the efficacy of our MGRL in FC network feature extraction and ASD identification, compared with several state-of-the-art methods.