Comparison of few wavelets to filter ocular artifacts in EEG using lifting wavelet transform

Abstract

We proposed a technique to remove ocular artifact (OA) from electroencephalogram (EEG) using Lifting Wavelet Transform (LWT). The objectives of OA filtering are threefold 1) to reduce the OA magnitude, 2) to retain the neural EEG potentials and 3) to make corrections at the right time localization. Three basic steps involved are to transform the EEG, hard thresholding the wavelet coefficients and the corrected EEG is obtained by inverse transform these threshold coefficients.It is of paramount important to select a suitable wavelet and threshold to accomplish this task. In this investigation, few wavelets were used and their effectiveness performances were compared visually. A level-dependent threshold used was suitable for both blinking and eyeball movements' corrections. From this study, we concluded that Haar and rbio3.1 outperformed db4 and cdf4.4 wavelets, and had achieved the objectives of OA removal.