Epileptic seizure prediction using power analysis in beta band of EEG signals

A feature extraction method for epilepsy prediction and localization of the brain region responsible for the seizure is proposed in this paper. The epileptic seizure can be predicted in advance by using power analysis in beta band of the EEG signals. The brain waves are divided into five different frequency bands (alpha, beta, delta, gamma and theta). The changes in the pre-ictal period are essential for seizure to start. Increase in power of the beta band of EEG signals in pre-ictal period of epileptic patient can be the indication of the occurrence of the seizure. In this work, the topographical map of the brain region divided into five different zones and the power in beta band is calculated from these five different zones. So, the particular region of the brain in which the highest change in power or the region which is responsible for the epileptic seizure is also localized.