B.E. Alekseev1

1 St. Petersburg State Electrotechnical University "LETI" (St. Petersburg, Russia)
1 BorisBoris12@yandex.ru

Diseases of the cardiovascular system are the main cause of human death throughout the history. A significant part of such deaths is classified as premature. If serious cardiovascular disorders occur, most often cardiac arrhythmias, there is very little time left to provide medical care to the patient. In this regard, it becomes critically important to ensure a short reaction time to the development of a disorder, which could provide a reserve of time to prepare for medical intervention or perform actions aimed at preventing the development of the disease to a life-threatening one.

The goal of the research is to develop an algorithm for processing and analyzing short fragments of an electrocardiogram to classify them according to the degree of danger to human life.

The paper considers a classification algorithm based on the use of the AlexNet pre-trained convolutional neural network to classify the results of the Wavelet transform of short (2 seconds) electrocardiogram fragments. The training and testing were conducted on a database consisting of 1016 fragments, manually divided into 6 classes, ordered by the degree of danger to human life. The problems of data augmentation for further training of a neural network are considered, and an approach is proposed to increase the amount of information contained in an image obtained as a result of a Wavelet transform. The classification quality was assessed.

The results obtained make it possible to further improve monitoring systems with automatic heart rate classification in real time and will reduce the response time of medical personnel to the occurrence of dangerous disorders.

Alekseev B. E. Wavelet transform and convolutional neural networks for detecting short fragments of dangerous arrhythmias. Biomedicine Radioengineering. 2026. V. 29. № 3. P. 66–70. DOI: https:// doi.org/10.18127/ j15604136-202603-11 (In Russian)

1. World Heart Report 2023: Confronting the World’s Number One Killer. Geneva, Switzerland. World Heart Federation. 2023.
2. World Health Organization Fact Sheets. Cardiovascular Diseases. 2024. URL: https://www.who.int/europe/news-room/fact-sheets/item/cardiovascular-diseases (data obrashcheniya: 07.05.2025)
3. Nemirko A., Manilo L., Tatarinova A., Alekseev B., Evdakova E. ECG Fragment Database for the Exploration of Dangerous Arrhythmia (version 1.0.0). PhysioNet. RRID:SCR_007345. 2022. https://doi.org/10.13026/kpfg-xs25.
4. Goldberger A., Amaral L., Glass L., Hausdorff J., Ivanov P.C., Mark R., Stanley H.E. et al. PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulation [Online]. 2000. V. 101 (23). P. e215–e220. RRID:SCR_007345.
5. Katja Zeppenfeld, Jacob Tfelt-Hansen et al. ESC Scientific Document Group , 2022 ESC Guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: Developed by the task force for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death of the European Society of Cardiology (ESC) Endorsed by the Association for European Paediatric and Congenital Cardiology (AEPC), European Heart Journal, Volume 43, Issue 40, 21 October 2022, Pages 3997–4126, https://doi.org/10.1093/eurheartj/ehac262.
6. Krizhevsky A., Sutskever I., Hinton Geoffrey E. ImageNet classification with deep convolutional neural networks. Communications of the ACM. 2017. V. 60 (6). P. 84–90. DOI:10.1145/3065386.