Dynamic epileptic seizure propagation based on multiscale synaptic plasticity

Epileptic seizures are thought to be the result of an imbalance between excitatory and inhibitory processes in the brain, involving dynamic changes at both nodal and network levels. However, existing epilepsy network models often ignore these dynamics during seizures. In this study, we develop a dynamical network based on the Wilson-Cowan model, incorporating different connectivity structures and utilizing synaptic plasticity to achieve the excitation-inhibition (E-I) balance within nodes. The hyperexcitability of the epileptogenic node is realized by modulating the inhibitory synaptic connections within the model, and the changes of the E-I balance within nodes during seizures are explored. It is found that seizures further exacerbate the E-I imbalance in the epileptogenic node, and other normal nodes in the network exhibit the decreased inhibition due to seizure propagation. Notably, changes in the inhibitory synapses of recruited nodes are correlated with their in-degree as this correlation diminishes with the distance increasing. Furthermore, when considering the dynamic changes in network structure, we find that epilepsy networks with small-world properties can partially reduce the increase in network connectivity strength and changes of node inhibition to achieve seizure propagation suppression. These phenomena gradually diminish as the small-world properties are lost in the network. Through synaptic plasticity, we investigate the dynamic changes of nodes and networks in seizures, emphasizing the impact of seizure propagation on node characteristics and network structure.