陳少傑臺灣大學：電機工程學研究所林伯星Lin, Bor-ShingBor-ShingLin2007-11-262018-07-062007-11-262018-07-062005http://ntur.lib.ntu.edu.tw//handle/246246/53376哮鳴是一種連續且偶發的肺部聲音，它是氣流流經狹窄氣管發生振動而產生的聲音。哮鳴也常拿來當作某些肺部疾病的重要指標，例如像氣喘，因此我們有足夠的動機去設計一個自動偵測哮鳴的系統來偵測及分析這些哮鳴的聲音。藉由這套系統，可以將哮鳴的特徵，如發生的頻繁度、持續時間、主頻頻率、哮鳴聲的強度等萃取出來，以進一步幫助醫生做診斷。它也可以提供這類病患長時間的監控及分析。 本論文主要描述一快速且高效能的哮鳴辨識系統。首先，呼吸聲先經放大、濾除雜訊、數位化存入電腦後。接著由我們發展的程式讀入這些呼吸聲的檔案進行訊號處理及辨識。在本論文中我們提出三種訊號處理的方法，它們分別是時頻圖的雙邊濾波處理法(2D bilateral filtering of spectrogram)、逐次框取平均法(order truncate average method)、移動平均法(moving average method)。做完訊號處理的肺音資料接著被萃取哮鳴的特徵，送入倒傳遞類神經網路(back-propagation neural network)訓練。最後，我們再將另一組新的肺音資料送入已經訓練好的倒傳遞類神經網路進行辯識，類神經網路即可以分類哮鳴與正常的呼吸聲。 最後使用移動平均法的實驗結果，我們可以得到高達1.0的靈敏度(sensitivity)及0.895的鑑別度(specificity)。由於它的高效能及容易實現的特性，本論文所提出的哮鳴辨識系統可以有利的長期監控氣喘病患，更可以進一步協助醫師在呼吸疾病方面的研究。Wheezes are continuous adventitious lung sounds, which have been defined as lasting for at least 250 ms. They are probably produced when airflow makes narrow airways vibrate. Wheezes are common clinical signs in patients with obstructive pulmonary disease such as asthma. Automatic wheezing detection offers an objective and accurate way to analyze wheezing lung sounds. By the automatic wheezing detection system, the features about amount, time duration, frequency range, and power strength of wheezes can be extracted to help physicians diagnose. It also can provide long-term auscultation and analysis of a patient. This Dissertation describes the design of a fast and high performance wheeze recognition system. First, respiratory sounds are captured, amplified and filtered by an analog circuit; then digitized through a PC soundcard, and recorded in accordance with the Computerized Respiratory Sound Analysis (CORSA) standards. Since the proposed wheezing detection algorithm is based on three methods: 2D bilateral filtering of spectrogram, order truncate average (OTA) method, and moving average (MA) method. Some features are then extracted from the processed spectra to train a back-propagation neural network (BPNN). Eventually, the new testing samples go through the trained BPNN to recognize whether they are wheezing sounds. Experiment results of the MA method show a high sensitivity of 1.0 and a specificity of 0.895 in qualitative analysis of wheeze recognition. Due to its high efficiency, great performance and easy-to-implement features, this wheeze recognition system could be of interest in the clinical monitoring of asthma patients and the study of physiological mechanisms in the respiratory airways.中文部分 目　錄 第一章 簡介...伍 第二章 背景...陸 第三章 方法...柒 第四章 結果與討論...捌 第五章 結論...玖 英文部分 TABLE OF CONTENTS ABSTRACT......i LIST OF FIGURES......v LIST OF TABLES......ix CHAPTER 1 INTRODUCTION......1 1.1 Motivation......1 1.2 Objective......3 1.3 Dissertation Organization......5 CHAPTER 2 BACKGROUND......7 2.1 Characteristics of Normal and Adventitious Respiratory Sounds......7 2.1.1 Normal Breath Sound......9 2.1.2 Adventitious Breath Sound......10 2.1.3 Definition of Wheeze......11 2.2 Pulmonary Auscultation......12 2.2.1 Stethoscopes ......12 2.2.2 Digital Stethoscopes......13 2.3 Recording of Respiratory Sounds......15 2.3.1 Microphone Location and Attachment......16 2.3.2 Pick-Up Sensors......16 2.3.3 Amplifier......18 2.3.4 Filters......18 2.3.5 Analog-to-Digital Converter......19 2.3.6 Analog Recording......20 2.3.7 Digital Sound Recording......22 2.4 Basic Techniques for Respiratory Sounds Analysis......22 2.4.1 Expanded Time......21 2.4.2 Classical Spectral Analysis......23 2.4.3 Short-Time Fourier Transform......26 CHAPTER 3 METHODOLOGY......29 3.1 Overview of the System......29 3.1.1 Hardware Architecture ......30 3.1.2 Software Designs......31 3.2 Wheeze Recognition Based on 2D Bilateral Filtering of Spectrogram......32 3.2.1 Edge Reserving Filter ......34 3.2.2 Algorithm Based on 2D Bilateral Filtering of Spectrogram......37 3.3 Signal Processing by Order Truncate Average Method......43 3.3.1 OTA Method......43 3.3.2 Algorithm Implementation of OTA Method......45 3.4 Signal Processing by Moving Average Method......60 3.4.1 MA Method......60 3.4.2 Algorithm Implementation of MA Method......61 3.5 Back-Propagation Neural Network......75 3.5.1 Concept of Neural Network......75 3.5.2 Extracting of Wheezing Features......78 3.5.3 Training and Testing of Respiratory Sound Samples......79 CHAPTER 4 RESULTS AND DISCUSSION......83 4.1 Detection Based on 2D Bilateral Filtering of Spectrogram......83 4.2 Detection in MA Method......87 CHAPTER 5 CONCLUSION......97 REFERENCES......992353174 bytesapplication/pdfen-US哮鳴時頻圖的雙邊濾波處理法逐次框取平均法移動平均法倒傳遞類神經網路wheeze2D bilateral filtering of spectrogramorder truncate average methodmoving average methodback-propagation neural networkthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/53376/1/ntu-94-D88921027-1.pdf