Design of optimized adaptive filter for artifact removal from EEG signal

Abstract

Removal of artifacts from the acquired biomedical signal is a major issue. Various signal processing techniques are developed by the researchers for some decades to extract the clinical information from EEG signals after the removal of unwanted artifacts during the time of its recording. In this work, an optimized weighted RLS filter is designed to suppress cardiac interference from the EEG signal. The cardiac artifact are estimated with the help of an optimized weighted RLS algorithm and eliminated from the EEG signal. Based on validation measures, Mean Square Error(MSE), Gain in Signal to Artifact Ratio(GSAR), Signal to Noise Ratio(SNR), Information Quantity(IQ), and INPS are used. The performance of the proposed method compared with the earlier methods like the Least square notch filter, ICA method are given in the result section. For the proposed method the MSE,GSAR,SNR,IQ, and INPS are found to be 0.1898 μV2 , 10.4807dB, 68.1117dB, 0.330, 10.91dB respectively. In every case, the performance of the proposed method is better than the existing method which proves its robustness.