The Study and Comparison between Various Digital Filters for ECG De-noising

摘要

Electrocardiogram (ECG) is a biosignal signal representing the electrical activities of the heart muscle. The common obstacle in analyzing and interpreting the ECG signal is the presence of unwanted artefacts and noises introduced during signal acquisition and transmission that contaminate or degrade the quality of ECG signal. A clean and noise-free ECG signal plays an imperative and vital part in the clinical analysis and diagnosis of cardiac abnormalities and diseases. Thus, the removal of these unwanted noises is an indispensable task to ensure a reliable ECG analysis performance. In this article, various types of noise contaminating the ECG signal are briefly discussed and several types of digital filter are modelled to de-noise the raw ECG signal. Their performance are evaluated and compared against each other with the signal-to-noise ratio (SNR) value using the ECG-ID Database obtained from PhysioNet. In this study, the structured qualitative and quantitative performance comparisons shows that the SNR of IIR digital filters outperform the FIR digital filters in de-noising the raw ECG signal, which the Chebyshev type II IIR filter shows the best ECG de-noising and filtering performance with average SNR value of 5.4995 based on 10 testing datasets.