Research On The Characteristics Of The Pathology Of Heart Sounds

Heart sounds signal (PCG) is corresponding to the status of the mechanical activitiesof heart and cardiovascular system, which contains lots of information which reflected onthe heart and cardiovascular system of human body. In the clinical, PCG is also treated asone of the important basis to estimate patients’heathy condition. But it is also verydifficult to get the satisfactory diagnosis effects through the traditional auscultation due tothe complexity of these signals.The main objective of the subject is to solve these following problems with the helpof advanced technology of modern engineering and computer science: Realize theacquisition of heart sounds signal correctly, which contains the pathology andphysiological information about the heart and cardiovascular system. Analysis the variouspathology features of collected heart sounds through these information. Doing this canimprove the weakness of traditional auscultation such as highly depend on the doctors’experience and the auscultation range are limited; improve the clinical diagnosis of heartdisease. People can visibly acquire the heart sounds and analyze the pathologicalconditions.Signal acquisition, denoising and segmented, feature extraction, classification andrecognition will be researched and discussed on the paper:In signal acquisition, a potable PCG signal acquisition system is designed based onSTM32 in the paper, which has these followings useful functions: PCG signal acquisition,amplification, filter, A/D conversion, portable storage and displaying. The device willthen send the collected signals to the PC.Firstly, during the signal processing, after familiar with the traditional denoisingmethod, a pretreatment of the denoising algorithm based on the wavelet is choosed, inaddition to the residual impurities and interference signals, and get the more pure heartsound signals, this is very significative for the following extraction of the characteristicvalue of signals.Secondly, in the identification and feature extraction of heart sound signals, the improved HHT transform is used to extract heart sounds envelope in the research. Afterpretreatment and extraction the signal envelope curve, a relatively smooth envelope waveis got, without losing the basic elements of these signals. Finally, according to thecharacteristics of signals, choose a corresponding section strategy algorithm, in theenvelope signals. Identify and analyze the composition of the envelope signals, andanalyze the pathological characteristic of the heart sound signals, combined with thecharacteristic of the signal spectrum and energy distribution of the signals.This topic has designed a complete system of heart sounds collection and analysissystem, and the system play a help for the clinical diagnosis of heart sound signals.