Applying the epidemic spreading model on the brain network to explain effective connectivity

Abstract

Network science studies a complex system as a network to capture the connectivity patterns and topological features. Different network topologies have been observed to shape dynamic spread- ing processes on the network in various ways, while the exact relationship is complicated and not yet fully understood. Epidemic models are often applied to describe the spreading process on a network and to facilitate studying the dynamic interactions. We apply a simple epidemic model, the Susceptible-Infected-Susceptible (SIS) model, on the structural brain network to explore the topological properties that drive the dynamic processes. A recent study examined the transfer entropy of empirical data and observed a dominant posterior-anterior spreading pattern in the brain. In both transfer entropy and delayed correlation measures, we show that hubs are more sending information to the network than lower degree nodes. With our continuous-time simu- lations, we also found the empirically-observed posterior-anterior global pattern. Based on our results, the brain topology of hubs mainly located at the back of the brain seems to be responsible for the emergence of the global pattern.