Effectiveness of gait cycle detection using unsupervised time series analysis
Effectiveness of gait cycle detection using unsupervised time series analysis
复制标题
使用无监督时间序列分析进行步态周期检测的有效性
DOI：
10.1109/iciibms55689.2022.9971502
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Iwami Takehiro
中科院分区：
文献类型：
--
作者：
Komatsu Akira;Iwami Takehiro
在本研究中，我们使用时间序列分析方法从附着在脚跟和大鱼际的足部压力传感器数据中检测步态周期，并验证该方法的有效性。在健康男性受试者的足跟和鱼际安装薄的足底压力传感器，进行直线路径上的步态测量实验，将测量的足底压力数据归一化为有无地面接触（站立相和摆动相），并生成步态周期波形。利用变化探测器从足底压力传感器的原始波形中检测出步态周期，并将检测结果与归一化后的步态周期波形进行比较。结果表明，通过时间序列分析检测到的波形与原始步态周期波形基本一致。然而，在站立阶段结束时观察到延迟。
In this study, we use a time series analysis method to detect the gait cycle from foot pressure sensor data attached to the heel and thenar and verify the usefulness of this method. After attaching thin foot pressure sensors to the heel and thenar of one healthy male subject, a gait measurement experiment was conducted on a straight path. The measured foot pressure data were normalized to the presence or absence of ground contact (stance phase and swing phase), and a gait cycle waveform was generated. The change finder was used to detect the gait cycle from the original waveform of the foot pressure sensor, and the results were compared with the normalized gait cycle waveform. As a result, the waveforms detected by time series analysis were generally in good agreement with the original gait cycle waveforms. However, a delay was observed at the end of the stance phase.