Graph Model Evolution During Epileptic Seizures: Linear Model Approach

Abstract

Epilepsy is a brain disorder characterized by sustained predisposition to generate epileptic seizures. According to the World Health Organization, it is one of the most common neurological disorders, affecting approximately 50 million people worldwide. A modern approach for brain study is to model it as a complex system composed of a network of oscillators in which the emergent property of synchronization arises. By this token, epileptic seizures can be understood as a process of hypersynchronization between brain areas. To assess such property, Partial Directed Coherence (PDC) method represents a suitable technique, once it allows a more precise investigation of interactions that may reveal direct influences from one brain area on another. During connectivity analysis, there may be a need to assess the statistical significance of the communication threshold and Surrogate Data, a method already applied for that purpose, can be used. Hence, the objective in this work was to carry out PDC connectivity analysis in combination with Surrogate Data to evaluate the communication threshold between brain areas and develop a graph model evolution during epileptic seizure, according to the classical EEG frequency bands. The main contribution is the threshold analysis adding statistical significance for connectivity investigation. A case study performed using EEG signals from rats showed that the applied methodology represents an appropriate alternative for functional analysis, providing insights on brain communication.