Study On Location Of Heart Sounds And Software Development

Cardiovascular disease is one of the diseases which are threatening human being's life. Phonocardiogram (PCG), recording of heart sounds and murmurs, often reflects the malfunction of the heart early. Improve the early diagnosis and the ability to diagnosis of cardiovascular diseases is a long time pursuits for medical staff, and the development of a cheap, simple, reliable and practical instrument is an imperative task for biomedical engineering staff.The primary task of the paper is development digital heart sound analysis software based on LabVIEW. The software system is mainly composed of five functional modules. Heart sounds and electrocardiograph (ECG) signal synchronous acquisition and display module using VISA achieves synchronous acquisition and display, the use of sound play functions and programming skills to achieve real-time play of heart sounds. Patient database management module is fulfilled by using LabSQL which is free. This module can fulfill these performances: display record, find, delete and modify, as well as import of other patient database. Cardiac auscultation teaching module makes user easy to master the cardiac auscultation knowledge by cartoon, and allows users on top of the characteristics of typical murmurs by letting them practice listening the clinical heart sounds. Heart sounds signal analysis module which programmed by LabVIEW and MATLAB mix-language, locates S₁ and S₂ firstly, and then applies the Normalized Averaging Shannon Energy, Welch power spectrum density and the Short Time Fourier Transform to analysis heart sounds. The analysis results will show on the screen graphically. Patient report module shows that the basic information, typical ECG and PCG and diagnosis, and supports to print in a predetermined format. Organic combination of the above five modules comprise the entire software system.Another important task of this paper is study on a heart sounds location algorithm which could apply in digital heart sounds analysis, and validating the algorithm with clinical data, and diagnosing some arrhythmia problems, and judging whether there is noise. Before location, PCG and ECG should be processed to filter noise. Firstly, the method destruct original signals into approach signals and detail signals which have different frequency range with lifting wavelet. Secondly, deal the detail with proper threshold by the characteristics of signals. Finally, reconstruct the signals with the reverse of lifting wavelet and erase the main noise of signals. After pretreatment, the paper uses ECG to locate the heart sound. According to the relationship of ECG and PCG, the method detects the R-wave to locate S₁ using the wavelet transforms and the envelope. And then according to the relationship of T-wave and R-wave, detects T-wave to locate S₂. The paper locates the 853 clinical heart sounds. The overall recognition rate is 98.24%. Thus, the location method is feasible. Some arrhythmia problems could be diagnosed and clinical PCG could be diagnosed whether it is normal.