多分辨分析提取心率变异性中的睡眠结构信息

本文从工程分析的角度探索通过分析心率变异性推知健康人夜间睡眠结构的可行性、针对心率变化在睡眠过程中的特点,对心动间期信号作基于小波变换的多分辨率分析,深入考察了不同睡眠时相中的RRV(R-R interval variability)多分辨率信号,尤其在睡眠时相转换时的变化特点.建立了关于健康人睡眠规律和睡眠RR规律的规则库.综合分析RRV在多分辨率下的信息,用模糊逻辑推理系统对某时间段不同睡眠时相的隶属度进行推理.经对26例健康人睡眠RRV分析,与脑电人工分析结果比较,睡眠基本结构的平均符合率达到85%,醒觉和睡眠状态的平均符合率达到93%,证明用本方法进行睡眠结构分析是可行的.

Extraction of Sleep Structure Information from Heart Rate Variability Using Multi-Resolution Analysis

This paper is on a study of the feasibility to infer sleep structure information of healthy people at night by mans of analysis of HRV(Heart Rate Variability) from an engineering point of view. Focusing on characteristics of HRV during sleep, multi resolution analysis of signals based on wavelet transform was made. RRV(R R Internal Variability)multi resolution signals during different sleep phases were observed very carefully. Especially attention was paid to features of chage at transition of sleep phases. A database of rules of sleep RRV and regular sleep patterns of healthy people was set up. Information of RRV in multi resolution mode was analyzed comprehensively. Inference of index of subordination of different sleep phases at different time periods was carried out using fuzzy inference systems. 26 cases of sleep RRV of healthy people were analyzed and compared with results of analysis of brain electronic data obtained manually. The average rate of compliance of basic sleep structure was up to 85%, and the average ones for people who are staying up and sleeping reached 93%. That proves it is feasible to analyze sleep structure by using the method presented.