Removal of ocular artifacts from EEG using adaptive thresholding of wavelet coefficients
V. Krishnaveni, Sundaresan Jayaraman, L Anitha +1 more
Electroencephalogram (EEG) gives researchers a non-invasive way to record cerebral activity. It is a valuable tool that helps clinicians to diagnose various neurological disorders and brain diseases. Blinking or moving the eyes produces large electrical potential around the eyes known as electrooculogram. It is a non-cortical activity which spreads across the scalp and contaminates the EEG recordings. These contaminating potentials are called ocular artifacts (OAs). Rejecting contaminated trials causes substantial data loss, and restricting eye movements/blinks limits the possible experimental designs and may affect the cognitive processes under investigation. In this paper, a nonlinear time-scale adaptive denoising system based on a wavelet shrinkage scheme has been used for removing OAs from EEG. The time-scale adaptive algorithm is based on Stein's unbiased risk estimate (SURE) and a…