Automatic Seizure Detection Using Higher Order Moments

Since seizures in general occur infrequently and unpredictably, it´s automatic detection during long term electro encephalograph (EEG) recordings is highly recommended. This paper presents a method of analysis of EEG signals, which is based on time domain analysis. Signal from each channel was divided into different frames of a predetermined length and higher order statistical features were calculated for each frame. Clinical data recorded from normal subject and epileptic patient were used to test the performance of the proposed method . It was demonstrated that the new scheme was able to classify the normal and epileptic EEG with an accuracy of 97.77% with less computation.