阻塞性睡眠呼吸暂停低通气综合征(OSAHS)患者鼾声中携带上气道阻塞部位和程度的信息。鼾声分析诊断OSAHS的Meta分析显示：阳性似然比为4.44, 阴性似然比为0.15，表明该技术仍有待完善(J Clin Sleep Med 发表中）。我们发现单纯打鼾患者的鼾声后也存在呼吸紊乱事件，而多数OSAHS的鼾声后无呼吸紊乱事件，把两者的鼾声整体进行比较是目前鼾声研究中的普遍缺陷。有学者证实呼吸紊乱事件后的鼾声有别于其它鼾声并应用于诊断OSAHS，但事件前鼾声对呼吸紊乱事件的发生理应更有预测价值。我们初步研究提示呼吸紊乱事件前、中、后鼾声的共振峰较其它鼾声存在显著差别，拟将进一步把鼾声和呼吸紊乱事件的实时结合作为研究重点，建立这些鼾声信号自动识别和分析的方法，分析呼吸紊乱事件发生过程中每一个鼾声的特点，试图寻找准确判断呼吸紊乱事件的鼾声特征信号，为鼾声分析准确诊断OSAHS奠定基础。

Signals of snore have been demonstrated to carry important information about the obstruction site and degree in the upper airway of obstructive sleep apnea-hypopnea syndrome (OSAHS) patients in recent years. The results of the accuracy of acoustic analysis of snoring in the diagnosis of OSAHS using a meta-analysis showed that positive likelihood ratio was 4.4 and negative likelihood ratio was 0.15. Those revealed that acoustic analysis of snoring is a relatively accurate but not a robust method for diagnosing OSAHS(J Clin Sleep Med, in press）. We found that there was a lot of snores with respiratory disturbance events in simple snorer, and a remarkable percentage of snores without respiratory disturbance event in OSAHS patients, the frequency of simple snores decreased progressively with increasing severity of OSAHS. We speculated that the major weakness of the previous studies was that the comparisons were only based on whole-night recordings rather than on single events. A few studies mentioned the post apneaic snores were significantly different to the other snores to diagnose OSAHS. However,it seem that snores pre- respiratory disturbance events should embody more informations of the occurence of apneas and hypopneas. .We found the formant of snore pre- ，during and post respiratory disturbance event were significantly different to the other snores, we would intend to extract the snoring sound segments of patients with and without OSAHS, who went through full-night sleep assessment by PSG. Using the PSG score sheet, we marked the snoring sound segments with simple sonring, pre- ，during and post -hypopneic and apneic classes. Furthermore, comparing and characterizing those classes (using several features), investigating the variability of snoring sound segments within each class (using total variation analysis), and investigating how the variability of snoring sound segments within each individual reflects on OSAHS severity (using regression analysis). Much more, the different gender, different age, different severity of OSAHS, different posture and different stages of sleep snoring were also observed. The characteristic snore signals were investigated by using analysis of coupled acoustic-structure systems with finite element method. The purpose is to establish a novel method of snore signal analysis and recognition detecting respiratory disturbance events to accurately diagnose OSAHS.