基于呼吸音多元特征参数优化的睡眠呼吸状态分析研究

In recent years, the incidence of sleep-related breathing disorders has been trending younger. Although the requirement of individual health management increases, the sleep-related breathing disorders can not get timely intervention limited by the lacking cognition of long-term accumulation effect of chronic diseases, the high cost and complex operation of clinical testing (polysomnography, PSG) and so on. So the sleep-related breathing disorders will effect the spirit condition and life quality of human being seriously. It will lead to kinds of cardiac-cerebral vascular diseases and the traffic accident, threatening people’s life and endangering public safety. Therefore, the project will design a portable sleep breathing sound acquisition system with higher cost performance. The multivariate features are extracted based on collected breathing sound signal and statistical optimization is applied to process the features. Then a new comprehensive identification index is proposed to identify various sleep breathing states. The efficiency of new index for analyzing sleep-related breathing disorders is confirmed by taking the apnea-hypopnea index (AHI) from PSG as an reference. The correlation between the new index and changes of breathing ventilation states is researched as well. Finally, the ratio of normal and abnormal breathing states and other parameters will be used to evaluate the sleep quality and spirit states. The project aims to provide a reliable theoretical evaluation system for developing the follow-up family personalized sleep intervention, popularize popular science knowledge to increase the awareness of sleep and mental health, and to promote the development of healthy society.

近年来，睡眠呼吸障碍人群呈年轻化趋势。虽然人们对自身健康管理需求日益增长，但由于对慢性病长期积累效应严重性认识不足，且受到临床检测手段（PSG）费用高且需要专业医师指导等客观限制，睡眠呼吸障碍未能得到及时干预和管理。这越来越严重影响人们精神状态和生活质量，引发多种心脑血管疾病，威胁生命健康，导致交通事故频发，危害公众安全。因此，本项目拟设计高性价比的便携式睡眠呼吸音采集系统；提取呼吸音多元特征参数并进行统计优化；提出综合判别新指标以识别各种睡眠呼吸状态；以临床金指标PSG得到的低通气指数AHI为参照，确认新指标分析睡眠呼吸障碍的有效性及其与呼吸换气状态变化的关联性；最终应用正常、异常呼吸状态比例等参数，评价睡眠质量并研究其与精神状态的关系。本项目旨在为后续深入开展家庭个性化睡眠干预提供可靠的理论评价体系，为提升睡眠及精神健康认知度进行科普宣传，为实现社会大健康的发展目标助力。

本项目以睡眠健康为关注点，以睡眠相关疾病检测和日常睡眠健康管理的社会需求为研究背景，以呼吸音信号为主要研究对象，结合睡眠相关生理信号对鼾症、睡眠呼吸暂停综合症进行深入分析，在智能有效地评价睡眠质量的同时对睡眠呼吸相关疾病进行风险管理。.项目针对长程睡眠健康管理的需求，设计并完善了睡眠呼吸音信号采集系统，该系统对市售的蓝牙耳机进行性能改造，通过智能手机或平板电脑便携、无侵入地采集和储存呼吸音信号，并上传数据至后台进行分析。.项目采用数字信号处理技术，从时域、频域、时频域、能量、高维度等多角度提取睡眠呼吸音的多元特征参数。从音频信号这一简单但信息量丰富的生理信号入手，对睡眠呼吸中气体交换状态进行关联性讨论，优化并提出有效特征组对鼾音、低通气、呼吸暂停等不同呼吸状态进行有效识别。与血氧饱和度、心率参数、体动频率等相关临床判定指标进行对比、参考及联合分析，实现测试者睡眠健康质量评估，预警睡眠呼吸相关疾病，并对确诊病情发展情况进行监测。.本项目完成计划的研究内容，发表相关论文9篇，其中SCI收录4篇。获得国家发明专利授权1项，实用新型专利授权3项。培养研究生6人，本科生10余人。.本项目成果为基于呼吸音信号的睡眠健康分析研究提供了可靠的理论研究参考，也为健康中国发展进程中的家庭健康监护研究方向提供了开阔的实验思路。

内容获取失败，请点击重试