Comparative analysis of ECG signal processing methods in the time-frequency domain

The paper presents a comparative analysis of frequency-time methods for processing ECG signals. Initially for a comparative analysis of methods the structural feature of the morphology of ECG signals is analyzed. Based on the systematization of the revealed features of the ECG signal morphologies a comparative analysis of widely used Hilbert-Huang and Wavelet transformation methods was carried out. It is noted, their main advantages and disadvantages in comparison with the classical Fourier transform in the frequency-time domain. The analysis revealed that the Hilbert-Huang transform is superior to the known methods for processing ECG signals in the time-frequency domain. And along with this, it is noted in this paper that the Hilbert-Huang transformation opens up new possibilities for processing and analyzing long-term ECG signal records in the frequency-time domain. This approach will improve the accuracy of diagnosis of the cardiovascular system in the methods of electrocardiography. The possibility of using this method for long-term monitoring systems is determined by the formation of a posteriori unique basis, which is always justified for the processing and analysis of diagnostic signals and requires less computing resources.