Epileptic seizure detection using HHT and SVM

Abstract

The reliability and efficiency of classification strategies required to segregate between the categories of healthy patients and those suffering from epilepsy is of paramount importance. The erratic occurrence of epileptic seizures has stimulated the automatic seizure detection in EEG recordings. In this work, classification of EEG signals has been carried out using Hilbert Huang Transform (HHT) and Support Vector Machine (SVM). In this approach, the HHT based Time Frequency Representation (TFR) has been considered as Time Frequency Image (TFI). The time frequency image is segmented in accordance with the frequency bands of the rhythms. Also respective histograms of gray scale sub images are represented. Extraction of statistical features such as mean, variance, skewness and kurtosis of pixel intensity in the histogram is implemented. SVM with radial basis function (RBF) kernel has been employed for classification of seizure and non -seizure EEG signals.