Species-Preserved Structural Connections Revealed By Sparse Tensor Cca

Zhibin He,Ying Huang,Tianming Liu,Lei Guo,Lei Du,Tuo Zhang
DOI: https://doi.org/10.1007/978-3-030-33226-6_6
2019-01-01
Abstract:Comparative evolution studies can advance the understating of the brain's functional and structural mechanisms. Efforts have been denoted in the literature to identify structural common connectome preserved between a pair of species, such as macaques and humans. However, very few studies were reported to identify species-preserved structural connections systematically and simultaneously across more than two species at a connectome-scale. In this work, we used diffusion MRI (dMRI) and Brodmann areas as established tools to estimate the whole-brain connectome for three primates: macaque, chimpanzee and human. We designed a sparse tensor canonical correlation analysis (STCCA) algorithm to identify the connective components that are strongly correlated among the three species. Joint analysis of the components can help to identify the white matter pathways preserved among three species. These preserved connections are consistent with the existing neuroscience reports, demonstrating the effectiveness and promise of this framework.
What problem does this paper attempt to address?