摘要阻塞性睡眠呼吸暂停低通气综合征(obstructive sleep apnea-hypopnea syndrome)是在现代人群中发病率比较高的睡眠紊乱疾病,严重影响着人们的正常生活。相关的研究发现,打鼾者的鼾声信号中携带了充分的诊断该疾病所需的声学特征信息。因此,研究打鼾者鼾声信号的声学特征用于OSAHS的诊断意义重大。61621
本文定量分析了一段实测的OSAHS病人的一小时的鼾声数据样本,利用Matlab软件提取了共振峰、800HZ功率比和频率特征(平均频率、中心频率、峰值频率)等特征矢量的数据,计算了每个特征矢量的一些基本参数;并从统计学的角度对这些特征进行了初步分析。得出了一些如鼾声信号的频率主要分布在1000HZ以下等结论,并且从各个特征的概率密度曲线得到了这些特征的大致分布。
毕业论文关键词 鼾声信号 共振峰 功率比 频率特征 概率密度
毕业设计说明书(论文)外文摘要
Title Analysis and Detection of Snore Characteristics
Abstract OSAHS (obstructive sleep apnea-hypopnea syndrome) is a sleep disorder disease of relatively high incidence in modern populations, which is seriously affecting people's normal life. The finding that snorers’ snore signals carry sufficient acoustic feature information to diagnose the disease is revealed in related research. Therefore, the study of the acoustic characteristics of snorers’ snore signal is significant for the diagnosis of OSAHS.
The recorded one hour OSAHS patients’ snoring data samples is quantitatively analyzed in this article, features vector data of formant, 800HZ power ratio and frequency characteristics (average frequency, center frequency, peak frequency) is extracted using Matlab software, the basic parameters of every feature vector are calculated; and a preliminary analysis of these characteristics is conducted from a statistical point of view. Some conclusions, for example the frequency of snoring signal distribute mainly beneath 1000HZ is obtained, and the approximate distribution of these characteristics is obtained from each feature’s probability density curve.
Keywords snore signal formant power ratio frequency characteristics probability density
1 引言 1
1.1 研究背景及意义 1
1.3 本文研究的内容 2
2 鼾声信号和OSAHS 3
2.1 鼾声信号 3
2.1.1 鼾声信号产生的模型 3
2.1.2 鼾声信号的发病机理 4
2.2 OSAHS的产生机理和鼾声信号的初步解读 4
2.2.1 OSAHS的产生机理 4
2.2.2 OSAHS病人鼾声信号的初步解读 5
3 鼾声信号特征的概述与估计 9
3.1 鼾声信号的特征综述 9
3.2 鼾声信号的预处理 9
3.2.1 鼾声信号的分帧 10
3.2.2 鼾声信号的加窗滤波 10
3.2.3 鼾声信号的端点检测 10
3.3 鼾声信号的线性预测(LPC)处理 11
3.3.1 LPC的基本原理