Influence of Wiring Cost on the Large-Scale Architecture of Human Cortical Connectivity
Figure 5
(A) Illustration of the identified cortical modules on a horizontal projection. Modules, represented by large circles, are drawn at the mean position of the regions they contain, with a radius proportional to their size (number of their regions), and are coloured by the average ‘anatomical structure colour’ (see section labels in Figure 2A) of their regions. Regions (smallest circles) are connected to their modules, and drawn with the colour of their modules. The widths of the inter-module connections (white lines) are proportional to the sum of connection strengths between regions in each of the two modules to regions in the other. (B) high-resolution cortical connectivity matrix ordered according to recovered modules (blue sub-squares along main diagonal, from M1 to M13 from top left to bottom right). Left and top colour stripes show colour-code of the corresponding cortical region's greater anatomical structure (see section labels in Figure 2A). For colour-code of matrix elements see colour-bar in Figure 3. (C–E) Module correspondence matrices of the three surrogate groups. Each element of the matrices denotes the normalized frequency (from black through red to white, see colour-bar on right side) of the two regions to be placed into the same module in the module partitions found in the corresponding surrogate network group. (F–H) Consistency of cortical module partition in surrogate groups. Consistency of module classification of each region is measured by its mean scaled inclusivity in each surrogate group using the obtained cortical modules as reference partition (see Methods). Colour-coded scaled inclusivity values of regions are shown on coronal (top) and horizontal (bottom) projections (see colour-bar on right side).