A Design and Implementation for Heart Sound Detection Instrument based on FPGA

Abstract

According to the problem that traditional heart detection equipment is bulky and expensive, approach to designing and implementing portable heart sound detection instrument is proposed in this paper. After the heart sound signal is filtered and amplified by hardware circuits and FPGA, the SCM receives the pre-denoised signal and calculates the heart rate, broadcasting the result in time. Experiments show that the instrument is convenient for patients to diagnose by themselves and make patients receive treatment without delay, which improves the timeliness and accuracy of cardiac auscultation. Introduction The traditional phonocardiograph has limitations of heart sound storage and processing for lack of quantitative analysis function [1]. With the development of digital technology, research on heart sound, including heart sound detection, analysis, and identification as well as clinical application has been a hot spot across the world [2, 3]. A large number of achievements involved with heart sound have emerged so far. Many results indicate that the filtering and envelope extraction is the key to system performance [4]. In this paper the method to design and implement the heart sound detection instrument is introduced. The system can not only analyze heart sound signal in time and frequency domain, but also broadcast the testing results in time. The flexibility, portability and reliability make the heart sound detection instrument practicable. Design of the Integral System The heart sound is induced by the vibration of myocardial contraction, heart valves closure and blood impacting ventricular wall and artery wall. There are four kinds of heart sound during a period of heartbeat. The first and the second heart sound can be heard in principle. The heart sound signal is weak and unstable with low frequency. During the heart sound acquisition process, noise cannot be avoided [5]. Such noise has effect on the heart sound analysis. Main noise includes disturbance of 50Hz power frequency and harmonic component, electrode contact noise resulted from breath and motion, muscle contraction noise and electronic device noise. These noises can be reduced to a great extent by taking action. Measures merely relying on hardware circuits are unable to eliminate noise disturbance. Digital filtering technology is indispensable for decreasing noises. Wavelet denoising and self-adaption filter has been used in this paper in order to improve the signal to noise ratio [6]. Low frequency approximation signals and detail signals except for high frequency signals on the scale of 5 2 as the input signal   d t , including heart sound signal overwhelmed by noises   s n and other interference elements   v n . That is to say       d t s t v t   . The reference input signal   x t , 3rd International Conference on Machinery, Materials and Information Technology Applications (ICMMITA 2015) © 2015. The authors Published by Atlantis Press 278 which is independent of   v t but related to   s t ,can be selected from detail signals on the scale of 5 2 . The reference input signal is 1 X T n n n p x x x         , , . The weight vector is 1 2 ω T p          , , . The output signal is       1