Role of hierarchical heterogeneity in shaping seizure onset dynamics: Insights from structurally-based whole-brain dynamical network models

Abstract

The brain is a highly complex network system exhibiting nonlinear dynamics, and alterations in brain structure may give rise to pathological brain dynamics such as seizures. Recent efforts in whole-brain network modeling have provided a framework of integrating subject data with nonlinear dynamical models for reproducing empirically observed phenomenon. A question remaining largely unexplored is how the structure of brain network system support and affect seizure-like dynamics. In this paper, we simulate seizures using a structurally-based whole-brain dynamical model of coupled oscillators operating in the bistable regime, and seizure onset likelihood is quantified by the escape time for nodes to transit from background to oscillatory seizure state. Based on this model, the effect of spatial heterogeneity in brain regions on seizure onset dynamics is explored. Specifically, magnetic resonance imaging derived T1w/T2w myelin map is used to parameterize the model with biologically-relevant spatial heterogeneity that approximates structural hierarchy in the human cortex. The unique shaping effect of hierarchical heterogeneity on seizure onset dynamics is demonstrated through comparisons with surrogate models where the spatial order of heterogeneity is altered. We find that hierarchical heterogeneity could most effectively impede the propagation of excitation across brain networks. It may also improve the model’s capacity to better characterize latent seizure focus as observed clinically and to personalize for epilepsy patients. The robustness of results is demonstrated using an alternative way of representing spatial heterogeneity and a higher resolution of representing brain regions. Our findings highlight the significant role of hierarchical heterogeneity as a relevant mechanism of ictogenesis and its importance to be considered in the workflow of whole-brain modeling for epilepsy patients, which are fundamental to understanding the dynamical nature of the brain system and developing personalized treatments for epilepsy patients.