Performance Analysis of a PCG Acquisition System

Abstract

Heart Sounds Signal (HSS), a synonymous for Phonocardiogram (PCG) which is a record of the
audible sounds created by the vibrations caused due to mechanical activities of the heart and flow of blood
through the blood vessels during the cardiac cycle, determines the prima facie disorder in the cardiovascular
system of human body. Heart sound signals are characterized as non-stationery and non-deterministic signals.
They are generated due to impulsive closure and opening of the cardiac valves at different instant of time of
cardiac cycles. HSS are complex signals with numerous components and short-time intervals with noise
embedded in it. HSS are characterized by tone (frequency), intensity (decibels), duration (seconds) and pitch
(depending on the number of harmonics that accompany the background noise). For quantitative and qualitative
analysis of the PCG employing signal processing techniques, the heart sound needs to be captured in electrical
form using sophisticated electronic circuits. Noises are unwanted signals mixed with the actual signals and
restrict the measurements made using HSS. The acquired HSS needs to be denoised to make it free from noise
which is a part of preprocessing of the HSS.
In the present communication an electronic circuit has been designed and fabricated to acquire the HSS and
interfaced with the processing system for denoising. The output of such circuit is known as Phonocardiogram
(PCG). Butterworth filter of 3rd order and Chebyshev filter of type I and Chebyshev filter of type II have been
utilized for denoising. A comparison of the performances in terms of SNR of the filters employed has been
carried out to identify the effectiveness of the filters in performing the denoising operation of PCG. PCG has
been acquired using the designed circuit for persons under various physical state as well as for a comparison,
the PCG data available from open data sources have also been taken into consideration. It has been observed
that the Chebyshev filter of type I exhibited the best performance assuring its use to remove the problems posed
by artifacts, noise and interference of various forms in the analysis of HSS