To study the characteristics of electroencephalogram (EEG) and its associated artifacts

Abstract

Electroencephalogram (EEG) measures brain function by analyzing the scalp electrical activity generated by brain structures. Local current flows are produced when brain cells (neurons) are activated. However, only electrical activity generated by large populations of neurons concurrently active can be recorded on the head surface. The small electrical signals detected by the scalp electrodes are amplified thousands of times, then displayed on paper or stored to computer memory. EEG like all biomedical signals is very susceptible to a variety of large signal contamination which reduces its clinical usefulness. Many researches had discovered that EEG signals are noisy and non-stationary. These EEG signals are contaminated by artifacts due to blinking, eyeball movements and muscle movements. However the main contamination is due to ocular artifacts elicited by blinking and eyeball movements. This research proposed a novel approach of adopting lifting wavelet transform (LWT) to eliminate ocular artifacts. Three basic steps involved were to transform the EEG, hard thresholding the wavelet coefficients and the corrected EEG was obtained by inverse transform these threshold coefficients. It is of paramount important to select a suitable wavelet and threshold value to accomplish this task. Thus, to select a wavelet for artifact removal in electroencephalogram using this method, relative wavelet energies were determined before and after thresholding. Relative wavelet energy (RWE) gives information about the relative energy associated with different frequency bands and can be considered as a time-scale density. RWE can be used as a tool to detect and characterize a specific phenomenon in time and frequency planes. This study concluded that cdf4.4 outperformed db4 and haar wavelets by removing the artifacts at the correct times and frequency bands.