Normalized Least Means Square (NLMS) Adaptive Algorithm as an Effective Method for Denoising Real Time ECG Signal

Abstract

Electrocardiogram (ECG) signals play a crucial role in the diagnosis of a range of cardiac disorders; however, they are frequently affected by various sources of noise, including baseline drift, muscle movements, electrode displacement, and electrical interference from power lines. This study introduces a robust adaptive filtering method founded on the Normalized Least Mean Squares (NLMS) algorithm, designed to effectively remove noise from ECG signals while retaining key diagnostic characteristics. The proposed methodology dynamically adjusts filter coefficients in real-time, successfully mitigating both high-frequency and low-frequency noise elements without altering essential waveform components, including the P-wave, QRS complex, and T-wave. To establish the superiority and robustness of the proposed technique, the performance of the developed prototype hardware it was evaluated using real ECG data acquired from the volunteers. The experimental result showed that the proposed stationary wavelet transform based ECG denoising technique outperformed the other ECG denoising techniques as more ECG signal components are preserved than other denoising algorithms. Comparing the outputs of both filtered and unfiltered ports, the results underscore the filter’s robustness and reliability for clinical ECG analysis, suggesting strong potential for deployment in real-time, patient monitoring systems and portable diagnostic devices. The practical application successfully demonstrated that the proposed filtering solution could function efficiently on hardware that is both relatively low-cost and portable, making it highly accessible for widespread utilization. Such advancements hold significant implications for the development of point-of-care monitoring systems and wearable health devices, ultimately enhancing the capability to deliver high-quality cardiac monitoring solutions in various healthcare settings.