EEG Based Epilepsy Diagnosis System Using Reconstruction Phase Space and Naïve Bayes Classifier

Electroencephalogram (EEG) is one the most used tools for the diagnoses and analysis of epilepsy. The diagnosis of epilepsy diseases are still made by physicians manually. This process is time consuming and subjective. In this study, EEG signal is analyzed by Discrete Time Wavelet Transform and Reconstruction Phase Space. Both techniques are used together to extract EEG features that allows Naive Bayes classifier to diagnose the epilepsy diseases and classify the corresponding EEG signals into “normal” or “abnormal” classes based on the extracted features. To assess the performance of the proposed system, we conducted a simulation experiment that involved 200 EEG signals from publicly available EEG dataset from University of Bonn. The proposed algorithm shows excellent accuracy compared with other techniques