Abstract
This study aims to develop an automated method for de-noising cardiac signals using independent component analysis (ICA) on a 37-channel magnetocardiography (MCG) system. The traditional approach of applying ICA involves manual visual inspection to determine the retention or removal of independent component (IC) related to signal or noise, which is time-consuming and lacks assurance in preserving essential attributes of signal components during the de-noising process. To address these challenges, we propose a novel approach. A feature set comprising spectral, statistical, and nonlinear time series properties is computed from the ICs of thirty subjects. These features are then evaluated by a few machine learning (ML) models to optimally select ICs for de-noising cardiac time series. It is found that ICs evaluated by a gradient boosting decision tree (GBDT) classifier could accomplish the task of efficiently selecting components to de-noise MCG with an accuracy of 93%. The performance of the proposed method is qualitatively and quantitatively compared against conventional methods for noise elimination and preserving signal features. The proposed method has extensive application in de-noising multichannel MCG signals where the characteristics of the noise are not clearly known and for routine diagnostic assessments of subjects with cardiac anomalies in hospital settings.
Similar content being viewed by others
Data Availability
The data that support the findings of this study are available from the corresponding author on reasonable request.
References
U.R. Acharya, F. Molinari, S.V. Sree et al., Automated diagnosis of epileptic EEG using entropies. Biomed. Signal Process. Control 7(4), 401–408 (2012)
U.R. Acharya, K. Paul Joseph, N. Kannathal, et al., Heart rate variability: a review. Med. Biol. Eng. Comput. 44(12), 1031–51 (2006)
A.K. Barros, A. Mansour, N. Ohnishi, Removing artifacts from electrocardiographic signals using independent components analysis. Neurocomputing 22(1–3), 173–186 (1998)
A.J. Bell, T.J. Sejnowski, An information-maximization approach to blind separation and blind deconvolution. Neural Comput. 7(6), 1129–1159 (1995)
M. Blanco-Velasco, B. Weng, K.E. Barner, ECG signal denoising and baseline wander correction based on the empirical mode decomposition. Comput. Biol. Med. 38(1), 1–13 (2008)
L. Breiman, Classification and regression trees. Routledge. (2017)
D. Brisinda, P. Fenici, R. Fenici, Clinical magnetocardiography: the unshielded bet—past, present, and future. Front. Cardiovascular Med. 10, (2023)
D. Brisinda, A. Meloni, R. Fenici, Clinical multichannel MCG in unshielded hospital environment. Neurol. Clin. Neurophysiol. 8, (2004)
E. Colin Cherry, Some experiments on the recognition of speech, with one and with two ears. J. Acoust. Soc. Am. 25(5), 975–979 (1953)
S. Comani, V. Srinivasan, G. Alleva, et al., Entropy-based automated classification of independent components separated from fMCG. Phys. Med. Biol. 52(5), (2007)
M.E. Davies, C.J. James, Source separation using single channel ICA. Signal Process. 87(8), 1819–1832 (2007)
D. DiPietroPaolo, H.P. Mueller, G. Nolte et al., Noise reduction in magnetocardiography by singular value decomposition and independent component analysis. Med. Biol. Eng. Compu. 44(6), 489–499 (2006)
D. Djuwari, D.K. Kumar, M. Palaniswami, Limitations of ICA for Artefact Removal. In: 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference. pp. 4685–4688 (2006)
R.O. Duda, P.E. Hart, Pattern classification (Wiley, Sons, 2006)
A. Géron, Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow. O’Reilly Media, Inc (2022)
R.M. Gulrajani, Bioelectricity and biomagnetism (Wiley, New York, 1998)
T. Hastie, R. Tibshirani, J.H. Friedman, The elements of statistical learning: data mining, inference, and prediction, vol. 2 (Springer, Cham, 2009)
A. Hyvarinen, Fast and robust fixed-point algorithms for independent component analysis. IEEE Trans. Neural Netw. 10(3), 626–634 (1999)
A. Hyvärinen, E. Oja, Independent component analysis: algorithms and applications. Neural Netw. 13(4–5), 411–430 (2000)
S. Ikeda, K. Toyama, Independent component analysis for noisy data: MEG data analysis. Neural Netw. 13(10), 1063–1074 (2000)
M. Iwai, K. Kobayashi, M. Yoshizawa et al., Automatic component selection for noise reduction in magnetocardiograph based on independent component analysis. J. Magn. Soc. Japan. 41(2), 41–45 (2017)
S. Kawakami, H. Takaki, S. Hashimoto, et al., Fragmentation assessed by magnetocardiography but not electrocardiogram can predict future cardiac events in patients with non-ischemic dilated cardiomyopathy and narrow QRS. Circulation. 130(suppl_2), A17191-A17191 (2014)
K. Kobayashi, Y. Uchikawa, T. Simizu et al., The rejection of magnetic noise from the wire using independent component analysis for magnetocardiogram. IEEE Trans. Magn. 41(10), 4152–4154 (2005)
K. Kobayashi, M. Iwai, Quantitative independent component selection using attractor analysis for noise reduction in magnetocardiogram signals. IEEE Trans. Magn. 54(11), 1–4 (2018)
K. Kobayashi, Y. Uchikawa, K. Nakai et al., Visualization of the current-density distribution for MCG with WPW syndrome patients using independent component analysis. IEEE Trans. Magn. 40(4), 2970–2972 (2004)
P. Korhonen, T. Husa, I. Tierala et al., Increased intra-QRS fragmentation in magnetocardiography as a predictor of arrhythmic events and mortality in patients with cardiac dysfunction after myocardial infarction. J. Cardiovasc. Electrophysiol. 17(4), 396–401 (2006)
S. Luo, P. Johnston, A review of electrocardiogram filtering. J. Electrocardiol. 43(6), 486–496 (2010)
J. Malmivuo, R. Plonsey, Bioelectromagnetism: principles and applications of bioelectric and biomagnetic fields (Oxford University Press, USA, 1995)
N. Mariyappa, S. Sengottuvel, P. Rajesh et al., Denoising of multichannel MCG data by the combination of EEMD and ICA and its effect on the pseudo current density maps. Biomed. Signal Process. Control 18, 204–213 (2015)
B. Mijović, M. De Vos, I. Gligorijević et al., Source separation from single-channel recordings by combining empirical-mode decomposition and independent component analysis. IEEE Trans. Biomed. Eng. 57(9), 2188–2196 (2010)
S.K. Mukhopadhyay, S. Krishnan, A singular spectrum analysis-based model-free electrocardiogram denoising technique. Comput. Methods Programs Biomed. 188, 105304 (2020)
H.P. Müller, G. Nolte, D.D.P. Paolo et al., Using independent component analysis for noise reduction of magnetocardiographic data in case of exercise with an ergometer. J. Med. Eng. Technol. 30(3), 158–165 (2006)
M. Ohkubo, The emergence of non-cryogenic quantum magnetic sensors: Synergistic advancement in magnetography together with SQUID. Rev. Sci. Instruments. 94(11), (2023)
C. Parasakthi, R. Patel, S. Sengottuvel, et al., Establishment of 37 channel SQUID system for magnetocardiography. In: AIP Conference Proceedings. 1447(1). American Institute of Physics, pp 871–2 (2012)
R. Patel, K. Gireesan, S. Sengottuvel et al., Suppression of baseline wander artifact in Magnetocardiogram using breathing sensor. J. Med. Biol. Eng. 37(4), 554–560 (2017)
F. Pedregosa, G. Varoquaux, A. Gramfort et al., Scikit-learn: Machine learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011)
C. Peng, S. Havlin, H.E. Stanley, et al., Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time series. Chaos Interdisciplinary J. Nonlinear Sci. 5(1), 82–7 (1995)
M. Potter, W. Kinsner, Competing ICA techniques in biomedical signal analysis. In: Canadian Conference on Electrical and Computer Engineering 2001 Conference Proceedings (Cat No01TH8555) (2001)
R.M. Rangayyan, Biomedical signal analysis (Wiley, New York, 2015)
M.A.D. Raya, L.G. Sison, Adaptive noise cancelling of motion artifact in stress ECG signals using accelerometer. In: Proceedings of the Second Joint 24th Annual Conference and the Annual Fall Meeting of the Biomedical Engineering Society In IEEE, 2, 1756–1757 (2002)
J.S. Richman, J.R. Moorman, Physiological time-series analysis using approximate entropy and sample entropy. Am. J. Physiol.-Heart Circulatory Physiol. 278(6), H2039–H2049 (2000)
S. Somarajan, N.D. Muszynski, D. Hawrami et al., Noninvasive magnetogastrography detects erythromycin-induced effects on the gastric slow wave. IEEE Trans. Biomed. Eng. 66(2), 327–334 (2019)
P. Parimita Swain, S. Sengottuvel, R. Patel, et al., A feasibility study to measure magnetocardiography (MCG) in unshielded environment using first order gradiometer. Biomed. Signal Process. Control. 55, 101664 (2020)
P. Takala, H. Hänninen, J. Montonen et al., Beat-to-beat analysis method for magnetocardiographic recordings during interventions. Phys. Med. Biol. 46(4), 975 (2001)
J.M. Tanskanen, J.J. Viik, Independent component analysis in ECG signal processing. In: Advances in Electrocardiograms-Methods and Analysis. IntechOpen (2012)
F.J. Theis, A. Meyer-Bäse, Biomedical signal analysis: Contemporary methods and applications (MIT Press, Cambridge, 2010)
S. Wallot, D. Mønster, Calculation of average mutual information (AMI) and false-nearest neighbors (FNN) for the estimation of embedding parameters of multidimensional time series in matlab. Front. Psychol. 9, 1679 (2018)
Acknowledgements
The authors thank Dr. K Gireesan, Dr. R. Nagendran, and Dr. N. V. Chandrasekar for their encouragement and support. Author C Kesavaraja expresses his gratitude to DAE for providing a research fellowship to carry out this study.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical Approval
This study was approved by the institutional ethics committee of Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Pondicherry.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A
Appendix A
Signal-to-error ratio and SNR were used to assess the performance of the proposed method in de-noising MCG signals.
1.1 Signal-to-Error Ratio (SER)
where x(t) is the raw measured signals, xde-noised(t) is the de-noised MCG signals after the ICA algorithm, and N is the total number of samples in the signal. The SER is expressed in units of dB [5, 35].
1.2 Signal-to-Noise Ratio (SNR)
Signal-to-noise ratio is calculated by considering the amplitudes of the R wave as x(t) and the peak-to-peak noise amplitude present in the time segment before the onset of the P wave as n(t), and it is expressed in dB [5].
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kesavaraja, C., Sengottuvel, S., Patel, R. et al. Machine Learning-Based Automated Method for Effective De-noising of Magnetocardiography Signals Using Independent Component Analysis. Circuits Syst Signal Process (2024). https://doi.org/10.1007/s00034-024-02655-9
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00034-024-02655-9