From a graph theoretical perspective, the brain is conceptualized as a networked system composed of regions (nodes) functionally connected with different brain regions by “paths,” which are weighted by measures of statistical dependencies between their electrical activity at the nodes they connect (De Vico Fallani et al. 2012). Some properties of interest are whether there are groups of nodes more strongly connected to other nodes in their group than to nodes in other groups, whether there are paths of high connectivity between most nodes, and whether some nodes (hubs) have many paths of high connectivity with other nodes. Graph theoretical analysis offers a powerful way to understand the topological principles of brain networks in normal and clinical populations and across development (He and Evans 2010). These methods have been applied to resting state as well as task-related data (Reijneveld et al. 2007). Magnetoencephalographic (MEG) studies also have used graph theoretical methods. Although EEG and MEG signals originate from the same neurophysiological processes, magnetic fields recorded using scalp sensors are less distorted than electrical potentials by the skull and scalp. However,
