ELECTROENCEPHALOGRAM ENHANCEMENT USING SIGN BASED NORMALIZED ADAPTIVE FILTERING TECHNIQUES

In this paper Adaptive filter is used as a primary method to filter the Electroencephalogram (EEG) or brain signal, as it does not require any priori information about the signal statistical charact eristics. Several simple and efficient sign based normalized adaptive algorithms are presented to cancel the noise in EEG signal. These are Normalized sign regressor Least Mean Square (NSRLMS), Normalized sign Least Mean Square (NSLMS) and Normalized sign sign Least Mean Square (NSSLMS). These algorithms enjoy less computational complexity because of the sign present in the algorithm and good filtering capability because of the normalized term . The filters developed using these algorithms are computational ly superior with multiplier free weight update loops. Based on these considerations NSRLMS, NSLMS and NSSLMS adaptive noise cancellers are developed for EEG signal enhancement. Power line interference (PLI) and Respiration artefacts are primarily considere d for denoising. Finally we have applied the algorithms on EEG signals obtained from CHB - MIT database for comparison of performance. The comparison of proposed schemes and conventional LMS indicates that NSRLMS outperforms existing realizations in noise re duction.