Resting-state networks can be identified with several analytical methods, including “seed” or region-of-interest based methods and independent component analysis (ICA).24 Region-of-interest based approaches measure the temporal correlation between an a priori selected brain region and all other brain voxels. The choice of the seed region is investigator driven and depends on the goals of the analysis. This approach identifies a network of brain areas (“nodes”; Panel 1) functionally connected with the seed region. ICA is a data-driven method that does not require a priori hypotheses about the regions of interest. This approach enables identification of multiple networks consisting of spatially independent and temporally correlated regions.25 Several networks have been consistently identified with either method (Figure 2):26 the default mode network (DMN), a posterior cingulate cortex-precuneus/medial temporal/lateral temporoparietal/medial frontal network that often deactivates during cognitively demanding tasks;27 bilateral executive-control networks made up of lateral frontal-parietal nodes;28 the salience network, an anterior cingulate/frontoinsular system with links to limbic and subcortical autonomic control centers,28 a dorsal attentional system embedded in high frontoparietal sensorimotor association regions,29 and networks related to primary visual, auditory, and
