TractGraphCNN: anatomically informed graph CNN for classification using diffusion MRI tractography
This work addresses the challenge of incorporating anatomical information into neural networks for brain imaging analysis, offering a domain-specific improvement for researchers in neuroimaging and computational neuroscience.
The authors tackled the problem of ignoring known neuroanatomical relationships in neural network design for brain connection analysis by proposing TractGraphCNN, an anatomically informed graph CNN framework, which demonstrated strong performance in sex prediction tasks on two large datasets (HCP and ABCD), with graphs based on white matter geometry outperforming those based on gray matter connectivity.
The structure and variability of the brain's connections can be investigated via prediction of non-imaging phenotypes using neural networks. However, known neuroanatomical relationships between input features are generally ignored in network design. We propose TractGraphCNN, a novel, anatomically informed graph CNN framework for machine learning tasks using diffusion MRI tractography. An EdgeConv module aggregates features from anatomically similar white matter connections indicated by graph edges, and an attention module enables interpretation of predictive white matter tracts. Results in a sex prediction testbed task demonstrate strong performance of TractGraphCNN in two large datasets (HCP and ABCD). Graphs informed by white matter geometry demonstrate higher performance than graphs informed by gray matter connectivity. Overall, the bilateral cingulum and left middle longitudinal fasciculus are consistently highly predictive of sex. This work shows the potential of incorporating anatomical information, especially known anatomical similarities between input features, to guide convolutions in neural networks.