Comparison on Wavelet Adaptive Filter Performance in Denoising ECG Signal

Abstract

Heart failure has sped up the development of electrocardiogram (ECG) signal denoising techniques. Researchers have developed a workable method to reduce the impact of noise in the ECG data, which can help avoid incorrect diagnoses and unnecessary medical procedures because heart failure can lead to fatal. Obtaining a clean ECG signal requires the use of an exact denoising process, which includes feature extraction, filtering, and pattern identification. To ensure the accuracy and reliability of the ECG signal, it is necessary to eliminate significant sources of noise, including baseline wander (BW), powerline interference (PLI), motion artifact (MA), and electromyogram (EMG). The study suggests that the filtering stage should utilise filters that are based on adaptive filters (AF). AF has the capacity to automatically update its filter coefficients in response to variations in the input signal and will keep updated the filter coefficient to minimise the error. The design of the AF and the performance comparison observed in the signal to noise ratio (SNR) test will be conducted utilising the MATLAB simulation software. The findings indicate that combining decompose wavelet transform (DWT) and adaptive filter (AF) may improve performance compared to DWT and AF alone. The best combinations to reduce noise signals from ECG are 30.01 dB, 29.67 dB, 23.72 dB, and 28.31 dB for BW, PLI, EMG, and MA, respectively.