Studying ECG signals using nonlinear oscillators and genetic algorithm

• Published in: International journal of dynamics and control Vol. 13; no. 3
• Main Authors: Chowdhury, Sourav, Ghosal, Apratim, Roychowdhury, Suparna, Chaudhuri, Indranath
• Format: Journal Article
• Language: English
• Published: Heidelberg Springer Nature B.V 01.03.2025
• Subjects: Algorithms; Atria; Differential equations; Electrocardiography; Genetic algorithms; Oscillators; Pacemakers; Rhythm; Sinuses; Waveforms
• ISSN: 2195-268X; 2195-2698
• DOI: 10.1007/s40435-025-01610-z

Cardiovascular diseases are the leading cause of death and disability in the world, and thus, their detection is extremely important as early as possible so that it can be prognosed and managed appropriately. Hence, electrophysiological models dealing with cardiac conduction are critically important in the field of interdisciplinary sciences. The primary aim of this paper is to reproduce a normal sinus rhythm ECG (electrocardiogram) waveform which will act as the baseline for fitting and then fit any clinical ECG waveform that does not deviate much from normal sinus rhythm. To reproduce the ECG, we modeled the pacemaker complex using three coupled van der Pol oscillators with appropriate delays to generate the action potentials. These action potentials are responsible for the excitation of the non-pacemaker cells of the atria and ventricles whose electrical activity gets recorded as the ECG signal. The ECG signal is composed of a periodic set of individual waves corresponding to atrial and ventricular contraction and relaxation. These waves are modeled with the help of four FitzHugh–Nagumo (FHN) equations with impulses corresponding to the action potentials generated by the pacemaker cells. After the successful reproduction of a normal sinus rhythm ECG, we have developed a framework where we have used a genetic algorithm (GA) to fit a given clinical ECG data with parameters belonging to the above-mentioned system of delay differential equations. The GA framework has enabled us to fit ECG data representing different cardiac conditions reasonably well. We aim to use this work to get a better understanding of the cardiac conduction system and cardiovascular diseases which will help humanity in the future.