Changes in connectivity patterns in the kainate model of epilepsy

Abstract

Epilepsy is a neurological disorder characterized by seizures, i.e. excessive and hyper synchronous activity of neurons in the brain. ElectroEncephaloGram (EEG) is the recording of brain activity in time through electrodes placed on the scalp and is one of the most used techniques to monitor brain activity. In order to identify pattern of propagation across brain areas that are specific to epilepsy, connectivity measures such as the Directed Transfer Function (DTF) and the Partial Directed Coherence (PDC) have been developed. These measures reveal connections between different areas by exploiting statistical dependencies within multichannel EEG recordings. This work proposes a framework to identify and compare interdependencies between EEG signals in different brain states. We considered an animal model of epilepsy characterized by spontaneous recurrent seizures. In three rats we identified a normal healthy baseline state and an epileptic state for which we estimated interdependencies between EEG channels using DTF and PDC and extracted significant differences between both states. We showed the feasibility of detection of connectivity patterns in a simple animal model of epilepsy. We found common patterns of propagation in the brain of the three rats during the baseline state. After the kainic acid injection, the connectivity pattern of interictal period is significantly altered compared to the baseline situation. Inter-rat variations are observed, but the intra-rat pattern alterations are consistent in time, revealing that the kainic acid permanently changes the brain connectivity.