基于声学分析的鼾症人群鼾声来源及上气道阻塞部位识别模式的研究

Snoring disease is a highly prevalent sleep disorder, which is not only a social nuisance that may impair other people's sleep quality and disturb a harmonic interpersonal relationship, but also has close relevent to many chronic diseases. The source of snoring and the obstructive site of upper airway are two important factors related to pathogenesis of snoring disease and are cracial for surgical planning . But current medical technics for identifying them have many limitations and demand to be improved eagerly. In this study, snore recording and acoustic analysis will be combined with polysomnography, sleep nasendoscope, continuous upper airway pressure monitoring,and pharygol 3D CT. The acoustic property of different type of snoring sounds, which have different sources or come from OSAHS patient with different obstructive sites, will be extrcacted and correlated with anatomical and physiological character of upper airway. A practical snoring analysis based method to identify source of snoring and obstructive site of upper airway will be established. As a result, this study has great value on revealing the pathological significance of snoring sound, establishing a diagnosis system based on snoring analysis, improving the effct of surgical treament, and reducing medical burden for snoring disease.

鼾症是一种发病率高、与多种慢性病的发病相关、对健康和社会造成较大危害的疾病。鼾声来源及上气道阻塞部位是与鼾症发病机制及手术治疗密切相关的重要问题。传统的医学检测模式对于鼾症来源及上气道阻塞部位的识别具有明显的局限性，而基于鼾声检测的识别方案具有经济、简便、无创、对睡眠干扰小又与睡眠同步的优势。本项目以鼾声为研究对象，将鼾声监测、声学分析与睡眠监测、睡眠内镜、整夜上气道测压及咽部三维CT等医学检测相结合,研究不同来源、不同上气道阻塞类型等不同病理、生理状态下鼾声的声学特征，及其与医学解剖、生理特征的内在对应关系，建立一种新型实用的、基于鼾声声学分析的鼾声来源及上气道阻塞部位的识别模式。本研究将声学技术与医学技术相结合，为鼾症的医学检测提供了新的研究思路,对于深刻揭示鼾声产生的机理和病理意义、建立鼾症的声学诊断模式、提高鼾症手术疗效，以及节约医疗资源具有重要价值。

鼾症患者鼾声来源及睡眠时上气道阻塞平面的确定是鼾症临床研究及诊治中的关键问题。现有的医学检测模式对于确定鼾声来源及睡眠时上气道的阻塞平面具有明显的局限性。本课题以鼾症患者的鼾声为研究对象，运用声学信号分析及处理技术与现代医学检测技术，研究了鼾声产生时上气道组织塌陷振动的模式，及不同来源、不同上气道阻塞平面鼾症患者鼾声的声学特征及其与各种医学解剖、生理特征的内在对应关系，为鼾声来源和上气道阻塞平面的诊断提供了新的思路。 .通过研究自然睡眠状态鼾声评估比较了鼾声信号的声压级参数、基频、共振峰、波峰因素、功率比等多种基本声学特征;通过药物诱导睡眠内镜观察,发现了鼾声产生时上气道不同结构的多种振动模式及其本质特点，提出了新的鼾声来源分类方式;通过对睡眠内镜同步采集鼾声的研究获得了不同来源鼾声信号的声学特征及其对于判断鼾声来源的敏感性和特异性，发现基频、功率谱等声学特征对软腭及非软腭来源鼾声具有较高的区分能力;从解剖和生理方面研究了鼾声来源及其声学特征的影响因素，发现腭咽区最小前后左右径比和咽侧壁厚度是鼾声来源的主要影响因素，腭咽区最小左右径是基频和共振峰高低的主要影响因素;通过对整夜上气道测压同步采集鼾声的研究,发现共振峰、功率比等声学特征的动态变化是对不同阻塞类型上气道鼾声的重要识别模式之一。同时自主研制了适用于整夜全方位采集鼾声信号且不影响医生和病患正常活动的传声器阵列系统，建立了一种基于广义旁瓣对消结构的鼾声自适应增强处理方法有效抑制室内的各种环境干扰，研究了整夜鼾声信号分析、检测、识别、截取与建库的全自动处理算法，初步建立了不同来源、不同上气道阻塞平面鼾声的信号识别模式与数据库模型，以及鼾声信号监测、处理和特征定量分析方法。