Huang Han-Ping, Hsu Chang Francis, Mao Yi-Chih, Hsu Long, Chi Sien
Department of Electrophysics, National Yang Ming Chiao Tung University, Hsinchu 30010, Taiwan.
Center for Industry-Academia Collaboration, National Yang Ming Chiao Tung University, Hsinchu 30010, Taiwan.
Entropy (Basel). 2021 Mar 31;23(4):412. doi: 10.3390/e23040412.
Gait stability has been measured by using many entropy-based methods. However, the relation between the entropy values and gait stability is worth further investigation. A research reported that average entropy (AE), a measure of disorder, could measure the static standing postural stability better than multiscale entropy and entropy of entropy (EoE), two measures of complexity. This study tested the validity of AE in gait stability measurement from the viewpoint of the disorder. For comparison, another five disorders, the EoE, and two traditional metrics methods were, respectively, used to measure the degrees of disorder and complexity of 10 step interval (SPI) and 79 stride interval (SI) time series, individually. As a result, every one of the 10 participants exhibited a relatively high AE value of the SPI when walking with eyes closed and a relatively low AE value when walking with eyes open. Most of the AE values of the SI of the 53 diseased subjects were greater than those of the 26 healthy subjects. A maximal overall accuracy of AE in differentiating the healthy from the diseased was 91.1%. Similar features also exists on those 5 disorder measurements but do not exist on the EoE values. Nevertheless, the EoE versus AE plot of the SI also exhibits an inverted U relation, consistent with the hypothesis for physiologic signals.
步态稳定性已通过多种基于熵的方法进行测量。然而，熵值与步态稳定性之间的关系值得进一步研究。一项研究报告称，平均熵（AE）作为一种无序度量，在测量静态站立姿势稳定性方面比多尺度熵和熵的熵（EoE）这两种复杂性度量方法表现更好。本研究从无序的角度测试了AE在步态稳定性测量中的有效性。为作比较，分别使用另外五种无序度量、EoE以及两种传统度量方法来单独测量10步间隔（SPI）和79步幅间隔（SI）时间序列的无序程度和复杂程度。结果显示，10名参与者中的每一个在闭眼行走时SPI的AE值相对较高，而在睁眼行走时AE值相对较低。53名患病受试者的SI的AE值大多高于26名健康受试者的AE值。AE区分健康人和患病者的最大总体准确率为91.1%。在这5种无序度量中也存在类似特征，但在EoE值中不存在。尽管如此，SI的EoE与AE的关系图也呈现出倒U形关系，这与生理信号的假设一致。
