Present methods of heart disease detection (auscultation of the heart with a phonendoscope, ECG recorder and phonograph) do not help to detect heart disease at early stages when the signs are mainly insignificant heart sounds and signals poorly distinguishable both by ear and by ECG recorder. It is shown that the human ear is far more sensitive to frequencies in the speech range (1000-2000 Hz) than to higher and lower frequencies. The heart sounds are primarily low-frequency vibrations that are to weak to be heard. This is the reason why digital recording and analysis of these vibrations may produce information that cannot be perceived by the physician’s ear.
The Scientific Research Biometric Technology Testing Centre of Bauman Moscow State Technical University and “ECHELON” scientific-production association have finished projecting hardware and software that are based on high-accuracy algorithms of wavelet analysis signals. New means of processing and obtaining multilevel time-and-frequency wavelets (acoustic cardiograms) of heart sounds and murmurs with the aim of conducting diagnostic cardiological testing are considered. Reception examples of obtaining acoustic cardiograms and the basic frequency-time characteristics of heart sounds and murmurs at various diseases.
At performance auscultation it is offered to carry out shift of frequencies of unheard low-frequency heart sounds in the audible sounds.