EARLY DIAGNOSTIC TECHNIQUE FOR CARDIOVASCULAR DISEASES BY MULTICHANNEL SPECTRAL ANALYSIS OF SLOW WAVES OF HEART SIGNALS

摘要

FIELD: medicine.;SUBSTANCE: invention refers to medicine, namely to non-invasive techniques for qualitative-quantitative analysis of the cardiovascular functional state. A pulse signal and an electric heart signal are recorded for 2-3 minutes. The slow waves are recovered from two heart signals; slow-wave spectra are detected in two channels. The windowed Fourier transform is used to calculate spectral ratio powers of the slow waves of the heart signal in the second-order slow-term within the range of 0.01 to 0.05 Hz, in the first-order slow-term within the range of 0.05 to 0.15 Hz, in the respiratory component within the range of 0.15 to 0.5 Hz. The derived data are used to form six informative criteria X…X6. As the heart signal recorded in one of the channels, a heart rate is calculated and used as the seventh informative criterion. The generated seven-element vector of the informative criteria is supplied to an input of a trained neuron network, outputs of which correspond to the allocated classes of the cardiovascular diseases.;EFFECT: technique enables early diagnosing aiming at preventing the disease progression, thereby preventing an increase of the primary hypertension incidence by analysing two heart signals.;3 cl, 9 dwg, 2 ex