The reseach for the development of a automatical analytic system by digital analysis of lung sounds

肺音数字分析自动分析系统的开发研究

基本信息

批准号：
08670684
负责人：
KUDOH Shoji
金额：
$ 1.15万
依托单位：
NIPPON MEDICAL SCHOOL
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
1996
资助国家：
日本
起止时间：
1996 至 1997
项目状态：
已结题

项目摘要

The informations collected through auscultations, which we usually diagnose pulmonary diseases, are less objective. The results and the transmitted volume of information are depend on the auditors. When we judge the changing condition of a disease, the auscultation is not hundred percent reliable measure because of its vague and lack of reappearance. So we developed a phonopeumograph using soundspectrograph, time-expanded waveform and fast fourier transformation to evaluate lung sounds objectively. Firstly, we analyzed cough sounds that are part of lung sounds.Coughs are divided into non-productive and productive coughs, but a standard method to measure and analyze them has not been established. The analysis of a cough may provide important clues not only to aid diagnosis, but also for selecting drugs for treatment. In this study, cough sounds, productive and non-productive coughs, were recorded from patients with chronic airway diseases. Then we compared their coughs with those of voluntary cough of healthy subjects and analyzed auditory by phononeumograph. All voluntary cough sounds from patients and healthy subjects could be separated into two or three phases. In addition, some horizontal long stripes were observed in the low-frequency area of phases 2 and 3. It is believed that cough sounds are produced by airway wall flattering or resonance of upper airway. The duration of phase 2 which shows productive cough of the patients is significantly longer than that of voluntary coughs of healthy subjects and non-productive coughs of patients. On the soundspectrograph, vertical lines appear when continuous sounds break off. It shows that productive cough sounds last discontinuously.We confirmed the characteristics of cough sounds by this study. It is inevitable to have further investigation on other lung and the automatic analyzing system of lung sounds by our developing phonopneumograph.

我们通常诊断为肺部疾病的听诊收集的信息是不太客观的。结果和传输信息量取决于审核员。当我们判断疾病的变化状况时，由于疾病的含糊和缺乏重新出现，听诊并不是一百％可靠的措施。因此，我们使用Soundsprograph，时间扩展的波形和快速的傅立叶变换开发了声音函数仪，以客观地评估肺部声音。首先，我们分析了肺部声音一部分的咳嗽声音。厨师分为非生产性和生产性咳嗽，但是尚未确定一种测量和分析的标准方法。对咳嗽的分析不仅可以提供重要的线索，不仅可以帮助诊断，而且还可以选择用于治疗的药物。在这项研究中，从慢性气道疾病的患者中记录了咳嗽声音，生产性和非生产性咳嗽。然后，我们将他们的咳嗽与健康受试者的自愿咳嗽进行了比较，并通过语音仪分析了听觉。患者和健康受试者的所有自愿咳嗽声音都可以分为两个或三个阶段。此外，在第2和3阶段的低频区域中观察到了一些水平长条纹。据信，咳嗽声是由上呼吸道的air道壁讨人喜欢或共振产生的。显示患者生产性咳嗽的2阶段持续时间明显长于健康受试者的自愿咳嗽和患者的非生产性咳嗽。在SoundspectRograph上，当连续声音断开时会出现垂直线。它表明，生产性咳嗽听起来不连续。我们通过这项研究证实了咳嗽声的特征。不可避免地要通过我们正在开发的PhonopeNumograph的其他肺部和自动分析肺部的自动分析系统进行进一步研究。