Design of error normalized LMS adaptive filter for EEG signal with eye blink & PLI artefacts

Analysis of spectral behaviour of the electroencephalogram (EEG) signal is a major hurdle due to the presence of artefacts. These artefacts are the low amplitude signals generated from unconscious ocular activity and muscles activity of human body. In our research we mainly considered Eye blink artefacts and Power Line Interference (PLI) for denoising. Since the source and noise in received signals originate from different sources, Least Mean Square (LMS) Adaptive filtering has been comprehensively used for filtering. Even after filtering the results show that considerable artefact components still persist in clean EEG signals. In this paper, we propose Error Normalized LMS (ENLMS) algorithm as the overhead computation with LMS for further filtering the signals. Further we applied signum to the proposed algorithm and developed Error Normalized Sign Regressor LMS (ENSRLMS), Error Normalized Sign LMS (ENSLMS) and Error Normalized Sign Sign LMS (ENSSLMS). It is concluded that the proposed Adaptive filter reduces the Eye Blink and PLI artefacts present in EEG signals without removing significant information embedded in these records.