A Hybrid Morphology-Preserving ECG Denoising Approach Using Iterative Regeneration and Wavelet Method

Abstract

Electrocardiogram (ECG) signals play a crucial role in cardiac diagnosis. However, the presence of noise artifacts such as electromyographic (EMG) noise and powerline interference can severely degrade the quality of ECG recordings, leading to incorrect clinical interpretations. Among these, EMG noise poses a significant challenge due to its spectral overlap with the vital QRS complex. The Iterative Regeneration Method (IRM) operates by extracting multiple similar heartbeats using correlation-based selection and reconstructing the ECG signal while preserving its morphological integrity. Following IRM, Wavelet Shrinkage is applied to further suppress any remaining high-frequency noise without distorting critical ECG features such as the PQRST complex. Synthetic ECG signals contaminated with EMG noise and powerline interference are used to validate the proposed method. The denoising performance is evaluated using segment-wise Signal-to-Noise Ratio (SNR) across iterations. The results demonstrate that the hybrid IRM-Wavelet approach effectively reduces noise while maintaining the clinical morphology of the ECG signal, making it suitable for reliable cardiac diagnosis and automated ECG analysis.

Introduction

Electrocardiograms (ECGs) are essential for cardiac diagnosis, but their accuracy is often degraded by noise such as baseline wander, powerline interference, and especially electromyographic (EMG) noise, whose frequency content overlaps with critical ECG features like the QRS complex. Conventional denoising methods, including linear filters and wavelet techniques, struggle to remove such noise without distorting ECG morphology, while advanced data-driven approaches (e.g., EMD, ICA, deep learning) are computationally expensive or data-intensive.

To address these limitations, the text proposes a hybrid morphology-preserving denoising framework that combines the Iterative Regenerative Method (IRM) with wavelet shrinkage. The approach aims to suppress EMG noise while retaining essential ECG features such as the P-wave, QRS complex, and T-wave.

The method consists of three stages: preprocessing, IRM, and post-processing. Preprocessing applies zero-phase digital filtering to remove powerline and high-frequency noise and enhances beat similarity, followed by QRS detection using the Pan–Tompkins algorithm. In the IRM stage, ECG signals are segmented into aligned heartbeats and averaged to form an initial template that preserves morphology. EMG noise is estimated and iteratively removed, with the number of iterations adaptively controlled using Signal-to-Noise Ratio (SNR) thresholds to prevent overprocessing.

Wavelet denoising using discrete wavelet transform (Symlet-8) and soft thresholding is applied after IRM to further suppress residual high-frequency noise. An adaptive LMS filter is also discussed as a comparative approach. Post-processing refines the signal to improve clarity and clinical usability.

The proposed hybrid IRM–Wavelet method is validated on synthetic and benchmark ECG datasets (e.g., MIT-BIH Arrhythmia Database). Results show improved SNR and superior preservation of ECG morphology compared to standalone denoising techniques, making the method well suited for reliable cardiac diagnosis and automated ECG analysis.

Conclusion

A hybrid denoising approach combining the Iterative Regeneration Method (IRM) and Wavelet Transform was successfully developed and implemented for the effective removal of EMG noise, power line interference, and other artifacts from ECG signals. The proposed method preserves important morphological features of the ECG while enhancing signal quality, which is critical for accurate clinical interpretation. The iterative nature of IRM enables beat-wise adaptive denoising while the wavelet stage further suppresses residual noise without distorting the diagnostic features. The performance evaluation using SNR maps and visual comparisons confirms the robustness and efficiency of the hybrid approach over individual denoising techniques such as pure wavelet or LMS filtering.

References

[1] K. H. C. Li et al., “The current state of mobile phone apps for monitoring heart rate, heart rate variability, and atrial fibrillation: Narrative review,” JMIR mHealth uHealth, vol. 7, no. 2, Feb. 2019, Art. no. e11606. [2] L. Sornmo and P. Laguna, “Noise and artifacts,” in Bioelectrical Signal Processing in Cardiac and Neurological Applications,1sted.NewYork, NY, USA: Academic, 2005, pp. 416–441. [3] A. D. William et al., “Assessing the accuracy of an automated atrial fibrillation detection algorithm using smartphone technology: The iREAD Study,” Heart Rhythm, vol. 15, no. 10, pp. 1561–1565, Oct. 2018. [4] S.S.Lobodzinski, “ECGpatchmonitorsfor assessment of cardiac rhythm abnormalities,” Prog. Cardiovasc. Dis., vol. 56, no. 2, pp. 224–229, Sep. 2013. [5] A.Shvilkin et al., “Coronary artery occlusion detection using 3-lead ECG system suitable for credit card-size personal device integration,” JACC Adv., vol. 2, no. 6, Aug.2023,Art.no.100454,doi:10.1016/j.jacadv.2023. [6] L. Hadzievski et al., “A novel mobile transtelephonic system with syn thesized 12-lead ECG,” IEEE Trans. Inf. Technol. Biomed., vol. 8, no. 4, pp. 428–438, Dec. 2004. [7] K. Bonaventura et al., “Comparison of standard and derived 12-lead electrocardiograms registrated by a simplified 3-lead setting with four electrodes for diagnosis of coronary angioplasty-induced myocardial is chaemia,” Eur. Cardiol. Rev., vol. 8, no. 3, p. 179, Jun. 2012. [Online]. Available: https://doi.org/10.15420/ecr.2012.8.3.179 [8] P. Kligfield et al., “Recommendations for the standardization and inter pretation of the electrocardiogram. Part I: The electrocardiogram and its technology: A scientific statement from the American heart association electrocardiography and arrhythmias committee, council on clinical cardiology; the American college of cardiology foundation; and the heart rhythm society,” Heart Rhythm, vol. 4, no. 3, pp. 394–412, Feb. 2007. [9] S. Ardhapurkar et al., “ECG denoising by modeling wavelet sub-band coefficients using kernel density estimation,” J. Inf. Process. Syst.,vol.8, no. 4, pp. 669–684, Dec. 2012. [10] E. Ebrahimzadeh et al., “ECG signals noise removal: Selection and optimization of the best adaptive filtering algorithm based on various algorithms comparison,” Biomed. Eng.-Appl. Basis Commun., vol. 27, no. 4, 2015, Art. no. 1550038. [11] L. Smital, M. Vítek, J. Kozumplík, and I. Provazník, “Adaptive wavelet wiener filtering of ECGsignals,” IEEETrans.Biomed.Eng.,vol.60,no.2, pp. 437–445, Feb. 2013. [12] G.D.Cliffordetal.,“Model-basedfiltering,compressionandclassification of the ECG,” Int. J. Bioelectromagnetism, vol. 7, no. 1, pp. 158–161, Sep. 2013. [13] P. E. McSharry and G. D. Clifford, “A comparison of nonlinear noise reduction and independent component analysis using a realistic dynamical model of the electrocardiogram,” Proc. SPIE, vol. 5467, pp. 78–88, May 2004. [14] R.Sameni, M.B.Shamsollahi, C. Jutten, and G. D. Clifford, “A nonlinear Bayesian filtering framework for ECG denoising,” IEEE Trans. Biomed. Eng., vol. 54, no. 12, pp. 2172–2185, Dec. 2007. ATANASOSKI et al.: MORPHOLOGY-PRESERVING ALGORITHM FOR DENOISING OF EMG-CONTAMINATED ECG SIGNALS 305 [15] H.-T.Chiang,Y.-Y.Hsieh,S.-W.Fu,K.-H.Hung,Y.Tsao,andS.-Y.Chien, “Noise reduction in ECG signals using fully convolutional denoising autoencoders,” IEEE Access, vol. 7, pp. 60806–60813, 2019. [16] A. R.-Meymandi and A. Ghaffari, “A deep learning-based framework For ECG signal denoising based on stacked cardiac cycle tensor,” Biomed. Signal Process. Control, vol. 71, Jan. 2022, Art. no. 103275. [17] A. Mincholé and B. Rodriguez, “Artificial Intelligence for the electrocardiogram,\" Nature Med., vol. 25, no. 1, pp. 22–23, Jan. 2019. [18] K.C.Siontis et al., “Artificial Intelligence- enhanced electrocardiography in cardiovascular disease management,” Nature Rev. Cardiol., vol. 18, pp. 465–478, Jul. 2021. [19] M. Elgendi, M. Meo, and D. Abbott, “A proof-of-concept study: Simple and effective detection of P and T waves in arrhythmic ECG signals,” Bioengineering, vol. 3, no. 4, 2016, Art. no. 26, doi: 10.3390/bioengineer ing3040026. [20] J. Pan and W. J. Tompkins, “A real-time QRS detection algorithm,” IEEE Trans. Biomed. Eng., vol. BME-32, no. 3, pp. 230–236, Mar. 1985. [21] H. Sedghamiz, “Matlab implementation of pan tompkins ECG QRS de tector,” Mar. 2014, doi: 10.13140/RG.2.2.14202.59841. [22] D. Esposito, J. Centracchio, P. Bifulco, and E. Andreozzi, “A smart approach to EMG envelope extraction and powerful denoising for human Machine interfaces,” Sci. Rep., vol. 13, May 2023, Art. no. 7768, doi: 10.1038/s41598-023-33319-4. [23] V. Atanasoski et al., “A database of simultaneously recorded ECG signals with and without EMG noise,” IEEE Open J. Eng. Med. Biol., vol. 4, pp. 222–225, 2023. [24] A. L. Goldberger et al., “PhysioBank, PhysioToolkit, and PhysioNet: Components of a research resource for complex physiologic signals,” Circulation, vol. 101, no. 23, pp. 215–220, 2000. [25] G. B. Moody and R. G. Mark, “The impact of the MIT-BIH arrhythmia database,” IEEE Eng. Med. Biol. Mag., vol. 20, no. 3, pp. 45–50, May/Jun. 2001.

Copyright

Copyright © 2025 Kamani Vyshnavi, Dr. M. Sushama. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.