Algorithmic Support for Spectral Processing of Cardiograms

The method of recording electrocardiograms as a non-invasive research method is widely used in modern functional diagnostics. Spectral diagnostic methods based on Fourier transform and wavelet transform are being developed. For the purposes of identification of cardiac rhythm disorders, the method of research selected is spectral (frequency) analysis of short-term ECG recordings, up to one period of heartbeats. Fourier series decomposition of the cardiac signal (ECG) in EDF-format for one period was carried out. It is determined that the maximum accuracy of cardiac signal description is achieved at the number of harmonics equal to half of the number of sampling points of the cardiac signal during the period. The correctness of the script developed for spectral analysis was checked by reconstructing the cardiac signal from its spectrum and comparing it with the original signal. The correlation between the spectrum and the shape of the cardiac signal has been established. The conclusion is made about the applicability of the spectral analysis method for the identification of heart rhythm disorders, as well as about the possibility of using the spectrum of electrical signals of heart contractions as a multidimensional function of the heart state. The direction of further identification of regularities by means of statistical analysis with interpretation of results by specialized specialists is indicated. The theoretical and practical value of this study lies both in determining the areas of application of spectral analysis of the cardiac signal for diagnosis and treatment, and in the practical results obtained, which can be used in the development of an expert system or a specific technical device.