One possible explanation for the similarity between degree and betweenness centrality runs as follows. Studies in adults have demonstrated that voxels within putative functional areas possess similar rs-fcMRI timecourses (Cohen et al., 2008). Functional areas within functional systems (e.g. the various portions of the DMN) tend to have correlated rs-fcMRI activity, and thus voxels should generally possess a) strong correlations within their functional area, and b) moderate correlations to functionally-related functional areas. In voxelwise graphs, this scenario could give rise to a hierarchical modular structure of functional areas and functional systems, and a voxel’s degree would generally scale 1) with the size of its functional area, and 2) with the number (and sizes) of functional areas comprising its functional system. Stated differently, a voxel’s degree should scale with the size of its module. If the default module spanned the greatest number of voxels in adults, it would be unsurprising that the highest-degree voxels in adults would be identified within default-like regions. “Hubs” in this scenario are simply members of the largest (whether measured by nodes, voxels, or volume) graph module.
