Comparison on LMS Adaptive Filter Performance in Denoising ECG Signal

Abstract

Since heart failure became a pressing issue, the development of the electrocardiogram (ECG) signal denoising technique has accelerated. Since heart failure issues might result in mortality, researchers have developed a good method for lessening the impact of noise in the ECG signal to avoid incorrect diagnoses and unnecessary operations. A precise denoising approach must be used to produce a free-noise ECG signal. The ECG signal needs to be cleaned up from major noises, such as baseline wander (BW), powerline interference (PLI), motion artefact (MA), and electromyogram (EMG). In the study, filters based on adaptive filters (AF) are recommended for the filtering stage. The design of the AF and the comparison performance displayed in the signal to noise ratio (SNR) test will be done using MATLAB simulation software. The results demonstrate that the proportionate normalise least mean square (PNLMS) adaptive filter performs better than other adaptive filters. The PNMLS adaptive filter reduces the BW, PLI, EMG, and MA noise effects by, respectively, 13.72 dB, 35.27 dB, 11.54 dB, and 75.41 dB.