Small-world network structure mimics spontaneous synchronization in epileptic seizures

The synchronous firing of neurons, though essential for normal physiological functions, can lead to seizures when it becomes excessive or otherwise abnormal. The cardinal symptom of epilepsy affects nearly 70 million people worldwide, but little is known about how synchronization in complex neural networks unfolds to produce unpredictable and varied seizures.

Gerster et al. employed a FitzHugh-Nagumo oscillator model to study the impact of network structure on the emergence and self-termination of seizures. The model used an empirical network structure obtained from diffusion-weighted magnetic resonance imaging data of human brains, with each of the network’s 90 nodes representing a certain part of the brain.

The researchers studied oscillation dynamics in different artificial network structures, including random networks, nonlocally coupled ring networks, ring networks with fractal connectivities, and small-world networks with various rewiring probabilities. They compared their simulations with electroencephalographic data of generalized epileptic seizures in patients.

The small-world network, with intermediate rewiring probability where each node was connected on average to six neighbors, was the most consistent in creating spontaneously occurring seizure-related synchronization. In the other structures, no seizure-related synchronization took place or synchronization remained too high throughout the simulation.

This suggests that self-initiation and self-termination of seizure-like synchronization rely on some balance of regularity and randomness, and that the value of the clustering coefficient should not be too high, such as in regular ring networks, or too low, such as in purely random networks.

Since seizures can be caused by macroscopic changes, such as brain lesions, the research could help identify specific neural network structures susceptible to epileptic seizures.

Source: “FitzHugh-Nagumo oscillators on complex networks mimic epileptic-seizure-related synchronization phenomena,” by Moritz Gerster, Rico Berner, Jakub Sawicki, Anna Zakharova, Antonin Škoch, Jaroslav Hlinka, Klaus Lehnertz, and Eckehard Schöll, Chaos (2020). The article can be accessed at https://doi.org/10.1063/5.0021420 .