Identifying influential nodes in a network model of epilepsy (Q2022729)

The authors approach the problem of pathological brain connectivity as a possible cause of focal seizures propagation in certain forms of intractable epilepsy. Showing that there is poor understanding of the precise structures that support epileptic brain connectivity, they use a computational model for the macaque cortical connectome to simulate focal seizure spread. The results obtained have been used to propose the so called Ictogenic Centrality as a network centrality measure aiming at detecting the nodes that contribute most effectively in seizure propagation. The authors show that, based on the ictogenic centrality concept, the results regarding the identification of ictogenic nodes outperform other centrality measures in what concerns accuracy (0.947), specificity (0.939), and sensitivity (0.964). The proposed degree method is based on a single free parameter, thus being easily computed for large networks. The authors suggest that their results can be interpreted in the sense that baseline brain connectivity may predispose the temporal and frontal lobes toward ictogenicity even in the case of absence of any overtly pathological network reorganization.