Graph averaging as a means to compare multichannel EEG coherence networks and its application to the study of mental fatigue and neurodegenerative disease

► The definition of a graph dissimilarity measure for EEG functional unit maps, which takes into account both node positions and node or edge attributes. ► A definition of the mean of two attributed graphs representing FUs, and its extension to an arbitrary number of such graphs. ► An algorithm for computing the mean of a set of FU-maps, with a quantitative measure of dissimilarity between this mean FU-map and each of the input FU-maps. ► Visualization of the mean FU-map employing a visual representation of the frequency of occurrence of nodes and the average coherence between nodes in the input FUs. ► The usefulness of the method is demonstrated in two case studies, one on mental fatigue and one on patients with corticobasal ganglionic degeneration (CBGD).