论文部分内容阅读
通过研究几种常用心率变异性(HRV)短时分析参数受非平稳性的影响,发现各短时参数在长时数据中均呈现出随时间变化的波动,但仅有HFnorm和样本熵(SE)受非平稳性影响大,其它如LFnorm和基本尺度熵(BE)的波动则与非平稳性无显著性相关。从而推论:HRV中的低频波动不只包含了非平稳干扰的影响,还蕴含了心脏动力系统的固有特性。重点针对BE方法,将不同人群HRV与三种典型噪声的熵值进行了比较,发现BE方法并不能准确的刻画HRV的长时相关性,而这正是BE方法良好抗非平稳能力的代价。
The short-term analysis of several short-term HRV parameters is affected by the non-stationarity. It is found that the short-term parameters show a time-dependent fluctuation in the long-term data, but only the HFnorm and the sample entropy (SE ) Are strongly affected by non-stationary, while others such as LFnorm and fundamental-scale entropy (BE) have no significant correlation with nonstationarity. The corollary is that low-frequency fluctuations in HRV not only include the effects of non-stationary interference, but also inherent characteristics of the cardiac dynamic system. Focusing on the BE method, we compare the entropy of HRV with three typical noises of different people. We find that the BE method can not accurately describe the long-term correlation of HRV, which is exactly the cost of the BE method to resist nonstationarity.