Automatic seizure detection using higher order moments & ANN

Since seizures in general occur infrequently and unpredictably, it´s automatic detection during long term electro encephalograph (EEG) recordings is highly recommended. Automatic Seizure Detection Using Higher Order Moments is based on the time domain analysis of EEG signal and extract the features for seizure detection. Each channel of both seizure and normal EEG data were divided into frames of 256 samples. Then corresponding to each EEG segment, higher order statistical features of variance, skewness, kurtosis and entropy were calculated. The significant non linear and non-Gaussian characteristics, shown by many medical signals prompt the selection of these parameters. After the feature extraction, classification was done using a linear classifier. The proposed method was able to detect epileptic seizures with an accuracy of 97.75%.