Automatic detection of epileptic EEG using THFB and auroregressive modeling

Abstract

In this work, the EEG signal is decomposed into its five subbands viz. delta (0.8-4Hz), theta (4-8Hz), alpha (8-15Hz), beta (15-30Hz), gamma (above 30Hz) using Triplet Half-band Filter Bank (THFB). Then, the autoregressive (AR) model is computed for each subband. Next, power spectral density (PSD) of the AR coefficients of each subbands is estimated for classfication of normal and epileptic EEG. It is observed that classification performed using THFB-AR modeling method gives better classification accuracy than existing method (approximate entropy).