Method of Atrial Endocardial Signals Processing and Recording

The article considers the problem of applying existing methods based on Fourier transformation for processing specific signals of biological objects. Authors developed method for the processing and recording signals from the endocardial area of the right atrium in patients with paroxysmal form of atrial fibrillation based on Fourier transform's application. The developed method is based on a comparative analysis of sampling rates and amplitude resolution of signal. The criteria's system for determination the minimum sampling rate of the signalminimize errors during the procedure of endocardial atrial signal's interpolation in the case of paroxysmal atrial fibrillation was proposed. The developed system of criteria allows to assess the artifacts obtained from samples with small number of reference points for the subsequent spectral analysis based on Fourier transform. The practical application of the developed method allows to create a database of atrial endocardial signals in the norm, in the case of persistent and paroxysmal forms of atrial fibrillation. This database can be a training samples for using neural network algorithms. The use of neural network algorithms will allow analysis of the current atrial signal to identify sites of pathological electrophysiological excitation with minimal computational cost, thereby increasing the efficiency of radiofrequency ablation for treatment atrial fibrillation.