Skip to main content
. 2022 Aug 22;13:4933. doi: 10.1038/s41467-022-32595-4

Fig. 1. The processing flow to construct a population-based tractography connectome and derive its hierarchical relation.

Fig. 1

a The diffusion MRI data of 1065 subjects were used. b The data were reconstructed to calculate the diffusion distribution for fiber tracking. c For each subject, 52 white matter bundles were mapped using automated tractography. The track recognition was based on trajectory similarity with a tractography atlas without using the cortical parcellations. The tracking results were aggregated to construct a population-based probability atlas of 52 white matter pathways. d Cortical regions from cortical parcellations and the white matter trajectories of each subject were used to derive the connectome matrix. e The results from each subject were accumulated to construct the tract-region connectome based on population probability. f Hierarchical clustering was applied to the row vectors of the connectome to derive the hierarchical relation of cortical regions. g Hierarchical clustering was applied to the column vectors to derive the hierarchical relation of white matter bundles.