ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 3. Vol. 30. 2024

DOI: 10.17587/it.30.159-167

M. R. Bogdanov, Ph. D., Associate Professor,
, Ufa University of Science and Technology, Ufa, 450000, Russian Federation,
G. R. Shakhmametova, Ph. D., Professor, Head of the Computational Mathematics and Cybernetics Department,
Ufa University of Science and Technology, 450000, Ufa. Russian Federation,
I. Sh. Shaibakov, Ph.D., Head of the Therapeutic Department,
GKUZ RB RCH No. 2, 450000, Ufa. Russian Federation,
N. N. Oskin, Director,
Siberian Telemetry Company, Penza, Russian Federation

Opportunities to Reduce the Risk of Cardiovascular Death by Improving Machine Learning Methods

Improving algorithms for automatic recognition of electrocardiograms requires increasing of training dataset, which is not always possible due to the rarity of certain cardiac diseases or ethical issues. It is possible to improve the algorithms for generating synthetic electrocardiograms using mechanistic models and generative-descriptive neural networks (GANs). At the same time, when evaluating the effectiveness of the proposed solutions, various authors offer different quality assessment metrics from subjective expert assessment to the squared mean error. We compare two approaches to generating synthetic electrocardiograms: pseudo-ECG generation using a one-dimensional cardiomyocyte model and GAN based on long-term short-term memory in terms of machine learning metrics: accuracy, recall, f1-score. We solve the problem of binary classification with bagging method, class 0 — normal sinus rhythm, class 1 — atrial fibrillation. We found that classifier trained on synthetic ECGs generated using GAN is slightly more effective compared to pseudo-ECGs generated using mechanistic models ones. At the same time, it turned out that GANs are not stable enough. The generation of synthetic electrocardiograms using both mechanistic models and GAN can be used to enrich the training set in case of recognition of rare cardiac diseases.
Keywords: synthetic data generation, electrocardiogram, generative-descriptive neural networks, one-dimensional cardiomyocyte model

Acknowledgements: RSF grant 22-19-00471 "Decision support system for the prevention and treatment of bronchopulmonary diseases, risk assessment of diseases and complications of their treatment in personalized medicine based on data analysis and artificial intelligence."

P. 159-167

References

1. Weinberg S., Iravanian S., Tung L. Biophysical journal, 2008, no. 95, pp. 1138—1150.
2. Laughner J. I., Ng F. S., Sulkin M. S., Arthur R. M., Efimov I. R. American Journal of Physiology-Heart and Circulatory Physiology, 2012, no. 303, pp. H753—H765.
3. Costabal F. S., Matsuno K., Yao J., Perdikaris P., Kuhl E. Computer Methods in Applied Mechanics and Engineering, 2019, no. 348, pp. 313—333.
4. Oliver J., Britton A., Bueno-Orovio K., Van A., Blanca R. Experimentally calibrated population of models predicts and explains intersubject variability in cardiac cellular electrophysiology, Biophysics and Computational Biology, vol. 110, no. 23, pp. E2098—E2105, available at: https://doi.org/10.1073/pnas.1304382110
5. Stephanie L. Hyland, Cristobal Esteban, Gunnar Ratsch. Real-valued (medical) time series genera tion with recurrent conditional gan, arXiv:1706.02633v2 [stat.ML] ,4 Dec 2017.
6. Radford A., Metz L., Chintala S. Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks, Conference paper at ICLR, 2016, arXiv:1511.06434v2 [cs.LG], 7 Jan 2016.
7. Costabal F. S., Matsuno K., Yao J., Perdikaris P., Kuhl E. Computer Methods in Applied Mechanics and Engineering, 2019, no. 348, pp. 313—333.
8. Plonsey R., Barr R. C. Bioelectricity: a quantitative approach, New York, Springer Science & Business Media, 2007.
9. Solosenko A., Petrenas A., Paliakaite B., Marozas V., Sornmo L. Model for Simulating ECG and PPG Signals with Arrhythmia Episodes, available at: https://physionet.org/content/ecg-ppg-simulator-arrhythmia/1.3.1/
10. Petrenas A., Marozas V., Solosenko A., Kubilius R., Skibarkiene J., Oster J., Sornmo L. Electrocardiogram modeling during paroxysmal atrial fibrillation: Application to the detection of brief episodes, Physiological Measurement, 2017, vol. 38, no. 11, pp. 2058—2080, http://doi.org/10.1088/1361-6579/aa9153
11. Goodfellow I., Pouget-Abadie J., Mirza M., Xu B., Warde-Farley D., Ozair S., Courville A., Bengio Y. Generative adversarial nets, Advances in Neural Information Processing Systems, 2014, pp. 2672—2680.
12. Wang Y. A mathematical introduction to generative adversarial nets (gan), arxiv:2009.00169v1 [cs.LG], 1 Sep 2020.
13. MIT-BIH Normal Sinus Rhythm Database, available at: https://physionet.org/content/nsrdb/1.0.0/.
14. Goldberger A., Amaral L., Glass L., Hausdorff J., Ivanov P. C., Mark R., Stanley H. E. PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals, Circulation [Online], 2000, vol. 101, no. 23, pp. e215—e220.
15. MIT-BIH Atrial Fibrillation Database, available at: https://physionet.org/content/afdb/1.0.0/
16. Yusupova N. I., Bogdanov M. R. Assessing the risk of sudden cardiac death using machine learning methods. Pattern Recognition and Image Analysis, Advances in Mathematical Theory and Applications, 2023, no. 3, vol. 33.
17. Baigildin S., Ushenin K., Fabarisova A., Bogdanov M., Solovyeva O. Linear transformations extract parameters of ionic currents and action potential features from pseudo-ECG in car-diomyocyte models, AAIP Conf Proc. 2313, 2020, p. 080007, available at: https://doi.org/10.1063/5.0035269
18. Ten Tusscher K. H., Bernus O., Panfilov A. V. Progress in biophysics and molecular biology, 2006, no. 90, pp. 326—345.
19. Muszkiewicz X. A., Liu X., Bueno-Orovio A., Brodie A. Lawson J., Burrage K., Casadei B., Rodriguez B. From ionic to cellular variability in human atrial myocytes: an integrative computational and experimental study, Integrative Cardiovascular Physiology and Pathophysiology, 2018 May 1, vol. 314, no. 5, pp. H895-H916, doi: 10.1152/ajpheart.00477.2017.
20. Britton O. J., Bueno-Orovio A., Van Ammel K., Rodri­guez B. Experimentally calibrated population of models predicts and explains intersubject variability in cardiac cellular electro-physiology, Biophysics and Computational Biology, 2013, vol. 110, no. 23, pp. E2098—E2105, available at: https://doi.org/10.1073/ pnas.1304382110
21. Dasi A., Roy A., Sachetto R., Camps J., Bueno-Orovio A., Rodriguez B. In-silico drug trials for precision medicine in atrial fibrillation: From ionic mechanisms to electrocardiogram-based predictions in structurally-healthy human atria, Front. Physiol., 15 September 2022. Sec. Cardiac Electrophysiology, 2022, vol.13, available at: https://doi.org/10.3389/fphys.2022.966046
22. Voigt N., Heijman J., Wang Q., Chiang D. Y., Li N., Karck M., Wehrens X.H.T., Nattel S., Dobrev D. MD. Cellular and Molecular Mechanisms of Atrial Arrhythmogenesis in Patients with Paroxysmal Atrial Fibrillation, Circulation, 2014 Jan 14, vol. 129, no. 2, pp. 145—156, doi: 10.1161/CIRCULA-TIONAHA.113.006641.
23. Zimik S., Majumder, R., Pandit, R. Ionic-heterogeneity-induced spiral- and scroll-wave turbulence in mathematical models of cardiac tissue, Biological Physics, 2018, vol. 1, pp. 1—8.

To the contents