Fall recognition based on surface electromyography phase synchronization analysis(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2018年第3期
- Page:
- 313-322
- Research Field:
- 医学信号处理与医学仪器
- Publishing date:
Info
- Title:
- Fall recognition based on surface electromyography phase synchronization analysis
- Author(s):
- DU Yucheng1; ZHANG Tingting2; WANG Xiaoyun2
- 1. Institute of Intelligent Control and Robotics, Hangzhou Dianzi University, Hangzhou 310018, China; 2. Guangdong Provincial Work Injury Rehabilitation Center, Guangzhou 510000, China
- Keywords:
- ; Hilbert phase synchronization; phase synchronization index; electromyography signals; muscles; fall
- PACS:
- TP29
- DOI:
- DOI:10.3969/j.issn.1005-202X.2018.03.013
- Abstract:
- The conscious activities of the human body are achieved by the internal and mutual synchronized oscillations of cerebral cortex and muscular tissue. Herein we aim to recognize unconscious fall from conscious daily activities by analyzing the phase synchronization of collected electromyography (EMG) signals. The EMG signals from tibialis anterior muscle, gastrocnemius muscle, rectus femoris and semitendinosus are all collected from 5 healthy participants when they do four different motions (walk, fall, sit down, sit down and stand up). Hilbert phase synchronization analysis is applied to calculate phase synchronization indexes, and the original surface EMG signals are used for comparing and analyzing the phase synchronization between muscles when participants do different motions. Subsequently, we apply wavelet packet decomposition to electromyography for studying the surface EMG signal phase synchronization in different frequency bands. The test results reveal that when participants fall, the phase synchronization between tibialis anterior muscle and rectus femoris and that between rectus femoris and semitendinosus are significantly different from the phase synchronization of other conscious motions, which indicates that we can use fisher linear classifier to recognize fall from other conscious motions by the phase synchronization of muscles. The fall recognition rates reach 85.5% and 91.0% when using full-band signals and the signals from chosen frequency band. In conclusion, the phase synchronization index reflects the situation of cooperative work between muscles, which can be used in fall recognition.
Last Update: 2018-03-21