Dauda, Samson YusufMaduakola, Chinomso FrancisUmar, IbrahimLoko, A.Z.Lumbi, Lucas Williams2023-12-142023-12-142020-06-191. Gautam A, Kaur M. ECG Analysis using Continuous Wavelet Transforms (CWT). IOSR Journal of Engineering. 2012;2(4): 632-635. 2. Subbiah S, Patro R, Rajendran K. Reduction of Noises in ECG Signal by Various Filters. International Journal of Engineering Research & Technology. 2014;3(1):656-660. 3. Han G, Xu Z. Electrocardiogram Signal De-noising based on a new improved wavelet thresholding. Review of Scientific Instruments. 2016;87(8):084303. 4. Ravandale YV, Jain SNA. Review on Methodological Analysis of Noise Reduction in ECG. IOSR Journal of Electronics and Communication Engineering. 2015;21-28, e-ISSN: 2278-2834, ISSN: 2278-8735. 5. Ravindra PN, Seema V, Singhal PK. Reduction of noise from ECG signal using FIR low pass filter with various window techniques. Current Research in Engineering, Science and Technology Journals. 2013;1(5):117-122. 6. American Heart Association. Heart and Stroke Statistical Update; 2019.https://keffi.nsuk.edu.ng/handle/20.500.14448/6022Introduction: Electrocardiogram (ECG) provides a wealth of information and remains an essential part of the assessment of cardiac patients. However, noise distortions associated with the signal could lead to wrong interpretation and diagnosis. Aim: To carry out an extensive comparative analysis of Savitzky-Golay (S-G) and Butterworth filters for ECG de-noising using Daubechies wavelets in a MATLAB version 2015a. Methodology: Noisy ECG signals downloaded from physionet.org under MIT-BIH arrhythmia database were de-noised using S-G and Butterworth filters displayed in both time and frequency domains. A quantitative evaluation was done to assess the performance of the filters for Signal to Noise Ratio (SNR), Mean Square Error (MSE) and Signal to Interference Ratio (SIR). The results of SNR for this work are compared with the results of other researches with other methods. Results: Experimental result for de-noising with Butterworth filter shows abnormal spiky waves in time domain quite unusual in morphology of the original waves and in the frequency domain creates image signals which are indications of noise and baseline drift. While S-G filter maintains the signal power constant and only tries to decrease the noise power with peak preservation. Performance analysis for SNR, MSE and SIR using Butterworth filter gives mean values of 1.63 dB, 0.2036 and 0.259 dB, while that of S-G filter gives 32.78 dB, 0.0001 and 1852.358 dB respectively. Discussion: Significant reduction of noise by S-G filter and retaining the ECG signal morphology effectively as compared to Butterworth filter is an evident that S-G filter delivers better performance results as compared to Butterworth filter in terms of noise separation, artifacts and baseline drifts. Conclusion: The importance of ECG de-noising filters and the criteria for their selection must be clearly understood by hospital managements and cardiac health centers for good quality ECG in diagnosis and therapy for cardiac diseases.enElectrocardiogram (ECG); Savitzky-Golay filter; Daubechies wavelets; butterworth filter; de-noising technique; signals.Article