An Effective Optimization Scheme for ECG Signal Denoising via Low-Rank Matrix Decomposition
- 112 Downloads
Electrocardiogram (ECG) signal denoising is an important preprocessing for ECG signal analysis. The contaminated ECG signal can be considered as the combination of the desired clean signal and the noise. Thus, ECG signal denoising can be considered as a problem of obtaining an optimal solution to the desired clean signal. In this paper, an effective optimization scheme for ECG signal denoising is presented based on low-rank matrix decomposition. First, the ECG denoising problem is formulated as low-rank matrix decomposition. So, an ECG beats matrix is assumed to be the combination of a sparse noise matrix and a low-rank matrix. Considering the repeatability of ECG signal, the rank of the ECG beats matrix is assumed to be one. Then, to fully exploit the low-rank property of the ECG signal, the matrix decomposition is modified by means of adding different weights to different singular values. Finally, the desired clean ECG signal is reconstructed by the low-rank component. The experimental results show that the proposed denoising method achieves the best performance of suppressing the electromyographic noise in the ECG signals compared with other optimization models.
KeywordsElectrocardiographic signal Low-rank decomposition Robust principle component analysis Weighted nuclear norm minimization Electromyographic noise
This work was supported in part by the National Natural Science Foundation of China (Grant Nos. 61001179 and 61571139), Guangdong Science and Technology Plan (Grant Nos. 2015B010104006, 2015B010124001 and 2015B090903017) and Guangzhou Science and Technology Plan (Grant Nos. 201604016022, 201803010065 and 201802020007).
- 3.M. Ayat, M.B. Shamsollahi, B. Mozaffari, S. Kharabian, ECG denoising using modulus maxima of wavelet transform, in Engineering in Medicine and Biology Society (2009), pp. 416–419Google Scholar
- 7.M. Dai, S.L. Lian, Removal of baseline wander from dynamic electrocardiogram signals, in Image and Signal Processing (2009), pp. 1–4Google Scholar
- 9.S. Gu, L. Zhang, W. Zuo, X. Feng, Weighted nuclear norm minimization with application to image denoising, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2014), pp. 2862–2869Google Scholar
- 19.Z. Lin, M. Chen, Y. Ma, The augmented Lagrange multiplier method for exact recovery of corrupted low-rank matrices. arXiv preprint arXiv:1009.5055 (2010)
- 20.A.J. Nimunkar, W.J. Tompkins, EMD-based 60-Hz noise filtering of the ECG, in Engineering in Medicine and Biology Society (2007), pp. 1904–1907Google Scholar
- 21.L. Sornmo, Time-varying filtering for removal of baseline wander in exercise ECGs, in Computers in Cardiology (1991), pp. 145–148Google Scholar
- 22.The MIT-BIH arrhythmias database. http://physionet.org/physiobank/database/mitdb/. Accessed 26 Oct 2016
- 23.The MIT-BIH arrhythmias database. http://physionet.org/physiobank/database/nstdb/. Accessed 26 Oct 2016
- 27.J. Wright, A. Ganesh, S. Rao, Y. Peng, Y. Ma, Robust principal component analysis: Exact recovery of corrupted low-rank matrices via convex optimization, in Advances in neural information processing systems, (Vancouver, B.C., Canada, 2009), pp. 2080–2088Google Scholar