心音訊號感測與頻譜分析之研究

Abstract

心音訊號代表著心臟部分功能運作之狀態，心臟之生理與病理的症狀可於心音訊號中表現出來，所以心音訊號之分析對於心血管疾病的診斷也具有重要的意義，一般在醫院中醫生最常用的診斷方式為使用聽診器進行聽診以評估病人症狀， 雖然此法簡單又方便但難免會有判斷上的誤差，對於非醫療領域的人員更是不容易準確的從聽診分辨心音以及確診，所以近年來心音訊號分析之研究已成為國內外生物醫學工程領域研究人員熱門的研究主題。 本論文的題目為「心音訊號感測與頻譜分析之研究」，目的是藉由自行開發之心音訊號感測系統量取一般真人受測者及收集SAMⅡ假人心肺音標準模擬系統之胸腔四個不同位置上之心音訊號並將得到的原始心音數據配合數位訊號處理軟體進行時域分析以及頻域上的頻譜分析且比較以整理出各別的生理特徵及特性並進一步用於正常與異常心音之初步診斷。 本研究中利用快速傅立葉轉換對心音訊號進行頻譜分析以得到整體訊號的頻率分佈及第一、第二心音和心雜音中各別組成成份的頻率訊息，並利用小波轉換進行時頻分析以確定第一心音、第二心音和心雜音組成成份之時序關係，藉由此研究得出量測心臟不同位置心音訊號之差異和心音訊號不同的特徵分類並且對於心音之診斷給出直接又精確的證據，另外在量測系統的設計上主要是以SAMⅡ系統的心音資料蒐集和以日本福田電子公司之心音量測機台為基準所設計之心音感測系統並配合福田電子開發之感測器進行量測實驗以得到真人及假人之心音原始數據加上數位訊號處理軟體對這些心音訊號進行學術研究分析與探討，並對於此套心音量測系統進行整合以實現心音感測醫療器材之雛型化開發。

Heart sound signals represent the operation state of the part of heart functions, the physiological and pathological symptoms of heart can be shown in the heart sound signals(PCG, phonocardiogram), so the analysis of heart sound signals also has important significance for the diagnosis of cardiovascular disease, doctors generally assess the patient's symptoms with auscultation in the hospital, although this method is simple and convenient, it may inevitably have error judgement and it is not easy to accurately distinguish heart sounds and make right diagnosis for non-medical field people, so the study of heart sound signals analysis has become a popular research topic in the field of biomedical engineering in recent years.

The topic of this thesis is “The Study of Heart Sound Signals Sensing and Spectrum Analysis”, the purpose is to measure the chest four different positions of real subjects and use SAM Ⅱ simulation system to get the heart sound signals raw data and do the time and frequency domain analysis with the digital signal processing software, by the analysis and comparison, the different physiological features and characteristics of heart sounds will be sort out and the normal and abnormal heart sounds will be further analyzed to achieve the initial diagnosis. In this study, the Fast Fourier Transform(FFT) is used to perform the spectral analysis of the heart sound signals to obtain the frequency distribution of the whole signal and the frequency information of the individual components in the first、second heart sounds and murmur ; the Continuous Wavelet Transform(CWT) be used to perform time-frequency analysis that it can determine the timing relationship of the first heart sound、 second heart sound and heart murmur composition, by this study, it is possible to measure the difference between the heart sounds、obtain the different characteristics of the heart sounds and provide a direct and accurate evidence for the diagnosis of the heart sounds. In addition, the design of the measurement system is mainly based on SAM Ⅱ system heart sounds data collection and the heart sound sensing system with Fukuda heart sounds sensor to obtain the heart sounds raw data of the real people subjects and SAM Ⅱ system for the academic research analysis and discussion with digital signal processing software and statistical software, and I also integrated this sensing system to achieve a prototype development of heart sound sensing medical device.