Epileptic Seizure Detection Using Temporal Based Measures in EEG Signal

Neuronal activity analysis of the brain for epileptic seizure detection. This neuronal activity is recorded in electroencephalogram (EEG) data. Present work focused on two temporal based features of EEG signal, including statistical and morphological features. The discriminating potential of extracted features is examined by using logistic regression classifier and validated by 10 fold cross validation. The result of the proposed method is found as 99.38, 99.58, and 99.48 percent of sensitivity, specificity, and accuracy respectively by using EEG data of University of Bonn, Germany.