Network Dynamics Help Devise Strategies for Epilepsy Surgery

A new study describes a mathematical model that can help distinguish ictogenic brain networks that act as the source of epileptic seizure activity.

Researchers at the University of Exeter (United Kingdom) have devised a predictive, network dynamics paradigm that allows the study of the relationship between brain network structure, node dynamics, and the generation of epileptic discharges. As part of the mathematical framework, the researchers quantified the ability of a brain network to generate emergent pathological dynamics, such as electrical discharges or seizures. They termed this ability as Brain Network Ictogenicity (BNI).

The specific model they developed for epilepsy surgery used spiking as a proxy for discharges. They claim that quantifying the ictogenicity of networks in this way allows them to determine the effect of perturbations to specific nodes of the network, thus uncovering the mechanistic contribution of each node to the emergent ictogenicity. Under this model, factors conventionally used to determine the region of tissue to resect, such as location of focal brain lesions or the presence of epileptiform rhythms, do not necessarily predict the best resection strategy.

The researchers then validated the paradigm framework by analyzing electrocorticogram (ECoG) recordings from 16 patients who had undergone epilepsy surgery. They found that when post-operative outcome was good, model predictions for optimal strategies aligned better with the actual surgery undertaken than when the post-operative outcome was poor. The model thus allows the prediction of optimal surgical strategies and the provision of quantitative prognoses for patients undergoing epilepsy surgery. The study was published on July 7, 2016, in Scientific Reports.

“Imagine someone was in a theatre and sending text messages to random audience members, making their phones ring. Current techniques are in essence akin to removing those people who receive the messages; they are contributing to the disruption and so removing them could make a difference,” said senior author Professor John Terry, PhD. “But clearly it would be better to identify and remove the individual sending out the messages - the original source. That is what our methods achieve - identifying the original source.”

“We were able to compare, for the first time, predictions made by a computer model applied to pre-surgical brain recordings with the post-surgical outcomes from a group of people with epilepsy,” said lead author Marc Goodfellow, PhD. “The potential for future treatment is clear - we are looking at a more accurate way of identifying exactly where to operate to give the best results for an individual, and so improve the lives of so many people who would otherwise have to live with the constant threat of seizures.”

Related Links:
University of Exeter