ECG signal classification using machine learning techniques

Authors

  • M. Kovalchuk Taras Shevchenko National University of Kyiv
  • V. Kharchenko Taras Shevchenko National University of Kyiv
  • A. Yavorskyi Taras Shevchenko National University of Kyiv
  • I. Bieda Taras Shevchenko National University of Kyiv
  • Taras V. Panchenko Taras Shevchenko National University of Kyiv https://orcid.org/0000-0003-0412-1945

DOI:

https://doi.org/10.17721/1812-5409.2022/2.9

Keywords:

electrocardiogram, ECG, ECG classification, one-dimensional convolutional neural networks, 1D CNN

Abstract

The importance of electrocardiogram (ECG) analysis is difficult to overestimate. Rhythm of life, stress and other factors affect the frequency of diseases and their early appearance. At the same time, the technologization (digitalization) of life and hardware-software complexes, such as mobile electronic cardiographs and wearable devices in general, which are rapidly developing, open new opportunities for rapid analysis of human state by certain indicators, as well as allow to diagnose on the new higher level in almost real time.

There are many methods for analyzing cardiograms. In this paper, the authors propose a new approach based on an ensemble of individual classifiers, which effectively solves the problem of ECG analysis. The study is based on the PhysioNet Computing in Cardiology Challenge 2017 and the MIT-BIH Arrhythmia Database. The algorithm consists of the following stages: data filtering using moving average and Butterworth filters, R-peak localization via threshold and grouping method, ECG resampling for the better comparability, “Noisy” vs “NotNoisy” classification as the most hard-to-identify class, final classification as “Normal”, “Atrial Fibrillation”, “Other” using an ensemble of 1D CNN classifiers and a final classifier of selection using logistic regression, random forest or support vector machine (SVM).

The proposed method shows high accuracy by the metric F1, so it gives the background for further research, optimization and implementation. This way this algorithm could help to save human’s life by in-time detection of problems with cardiovascular system (CVS) at early stage.

Pages of the article in the issue: 70 - 77

Language of the article: Ukrainian

References

1. PANCHENKO, T.V., BUDICHENKO, V.O. (2016): Real-time Health Monitoring via ECG Analysis, “Artificial Intelligence”, No. 4 (74), pp. 98-100.
2. CLIFFORD, G.D., LIU, Ch., MOODY, B., LEHMAN, L.H., SILVA, I., LI, Q., JOHNSON, A.E., MARK, R.G. (2017): AF Classification from a Short Single Lead ECG Recording: The PhysioNet – Computing in Cardiology Challenge 2017, “Computing in Cardiology”, pp.1-4, doi: 10.22489/CinC.2017.065-469.
3. CHEN, D., LI, D., XU, X., YANG, R., NG, S.-K. (2021): Electrocardiogram Classification and Visual Diagnosis of Atrial Fibrillation with DenseECG, 10 p.,
https://arxiv.org/pdf/2101.07535.pdf
4. YAVORSKYI, A., TYSHCHENKO, B., PANCHENKO, T. (2021): Efficient ECG Analysis with High F1 Score and Low Computation Complexity, “Proc. 11th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS’2021)”, pp. 348-352.
5. PANCHENKO, T., YAVORSKYI, A., HU, ZH. (2022): Electrocardiogram Effective Analysis Based on the Random Forest Model with Preselected Parameters, “Lecture Notes on Data Engineering and Communications Technologies”, Vol. 135, pp. 137-145.
6. YAVORSKYI, A. (2021): Real-Time Analysis and Processing of Cardiogram Signals, “Bulletin of Taras Shevchenko National University of Kyiv, Series Physics & Mathematics”, No. 1, pp. 108–113.
7. YAVORSKYI, A., PANCHENKO, T., TYSHCHENKO, B. (2021): ECG Analysis with High Precision and Recall, “Proc. Problems of Decision Making under Uncertainties (PDMU-2021)”, pp. 78-79.
8. GOLDBERGER, A., AMARAL, L., GLASS, L., HAUSDORFF, J., IVANOV, P.C., MARK, R., MIETUS, J.E., MOODY, G.B., PENG, C.K., STANLEY, H.E. (2000): PhysioBank, Physio-Toolkit, and PhysioNet: Components of a New Research Resource for Complex Physiologic Signals, # 101 (23), pp. e215–e220, https://physionet.org/content/challenge-2017
9. MOODY, G.B., MARK, R.G. (1983): A new method for detecting atrial fibrillation using R-R intervals, “Computers in Cardiology”, No. 10, pp. 227-230.
10. MOODY, G.B., MARK, R.G. (2001): The impact of the MIT-BIH Arrhythmia Database, “IEEE Engineering in Medicine and Biology Magazine”, Vol. 20, No. 3, pp. 45-50.
11. BUTTERWORTH, S. (1930). On the Theory of Filter Amplifiers, “Experimental Wireless and the Wireless Engineer”, No. 7, pp. 536–541.
12. WU, L., XIE, X., WANG, Y. (2021): ECG Enhancement and R-Peak Detection Based on Window Variability, “Healthcare” (Basel), 9 (2), P. 227; doi: 10.3390/healthcare9020227.

Downloads

Published

2022-10-12

How to Cite

Ковальчук, М., Харченко, В., Яворський, А., Бєда, І., & Панченко, Т. В. (2022). ECG signal classification using machine learning techniques. Bulletin of Taras Shevchenko National University of Kyiv. Series: Physics and Mathematics, (2), 70–77. https://doi.org/10.17721/1812-5409.2022/2.9

Issue

Section

Computer Science and Informatics