Fig. 3.

Identifying the components-of-interest in the resting-state. A. Component 8 (green) was identified as a good match spatially to the ATL-seeded network identified in the resting-state in Jackson et al. (2016) (blue). Overlap is shown in cyan. B. Component 21 (green) was identified as a good match spatially to both the mPFC-seeded default mode network (upper portion, blue) and an a priori DMN template (lower portion, blue). C. The two selected components are shown. The component hypothesised to be semantic is shown in green and the DMN component is shown in blue. Overlap is in cyan. Although both networks are more circumscribed than the seed-based analyses there is still some overlap, particularly within mPFC. All results have a voxel-level significance threshold of .001 and an FWE-corrected critical cluster level of .05.