Figure 2

NAc modular connectivity includes more subcortical regions in CBP and more frontal regions in controls. a) The biggest difference in modular structure of the group mean connectivity matrices lies between modules 3 (the NAc module) and 4. Connectivity matrices are labeled as they best align to those previously defined [29]. Shown are the lateral and medial views of the brain. For CON, the NAc module integrates more frontal regions, while module 4 is primarily subcortical. In contrast, for CBP, the NAc module integrates subcortical structures, while moduIe 4 includes mostly frontal and parietal regions. b) Each brain region (see Table 1) is assigned its likelihood of membership with the NAc module based on the percentage of participants including that region with this module. The bar plot shows the difference between the groups. A greater percentage of CBPs integrated subcortical and medial temporal regions with the NAc, while more controls integrated midline cortical, temporal lobe, and prefrontal regions. c) The NAc module is divided into shared and exclusive regions. Mean connectivity across these regions are correlated to gain sensitivity scores. The inset plot indicates regions that both CON and CBP integrate into the NAc module (blue nodes), based on each groups’ mean connectivity matrix. Gray nodes indicate those regions that are exclusively integrated to the CON’s NAc module, while black nodes are exclusive to CBP. Scatter plots illustrate the Pearson correlation between gain sensitivity scores and the mean connectivity between all nodes within each group of regions. The only plot that is not significantly correlated is with the blue to black nodes, suggesting the fully integrated NAc module in CBP cannot account for gain sensitivity, whereas that of controls can. Black lines in all maps indicate functional links between regions thresholded at a whole-brain link density of 0.2.