Performance Comparison of Adaptive Power Line Interference Cancellers for ECG Signal

Cardiovascular diseases are major cause of worldwide mortality and are expected to remain so. Electrocardiogram (ECG) is used to diagnose various heart diseases and is adopted most widely in clinics. Various artifacts get added in to the original ECG signal and their removal is crucial thus allowing physicians to extract useful information from the original ECG signal. The most common artifact that gets added in ECG signal is Power line interference (PLI). Various filtering techniques have been implemented in literature to eradicate PLI from noisy ECG signal. This paper presents performance comparison of the filtering capability of different adaptive filters for PLI suppression from ECG signal. The investigated adaptive algorithms are least mean square (LMS), recursive least squares (RLS), state space recursive least squares (SSRLS) and Kalman filter. The comparison is carried out for PLI with known amplitude and frequency. Mean square error (MSE), power spectral density (PSD) and noise reduction ratio (NR) are used as performance metrics for comparison.