Chapter

Universal Access in Human-Computer Interaction. Applications and Services

Volume 6768 of the series Lecture Notes in Computer Science pp 479-488

Intensity Analysis of Surface Myoelectric Signals from Lower Limbs during Key Gait Phases by Wavelets in Time-Frequency

  • Jiangang YangAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Xuan GaoAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Baikun WanAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Dong MingAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Xiaoman ChengAffiliated withSchool of Science, Tianjin University of Technology
  • , Hongzhi QiAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Xingwei AnAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Long ChenAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
  • , Shuang QiuAffiliated withDepartment of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University
    • , Weijie WangAffiliated withDepartment of Orthopaedics and Traumatology, Ninewells Hospital, University of Dundee

* Final gross prices may vary according to local VAT.

Get Access

Abstract

This paper presented a time-frequency intensity analysis feature extraction approach of lower limb sEMG (Surface Electromyogram) to identify the key gait phases during walking. The proposed feature extraction method used a filter bank of non-linearly scaled wavelets with specified time-resolution to extract time-frequency aspects of the signal.The intensity analysis algorithm was tested on sEMG data collected from ten healthy young volunteers during 30 walking circles for each. Each walking cycle was made up of four key gait phases:L-DS(Left Double Stance), L-SS(Left Single Stance),R-DS(Right Double Stance),R-SS(Right Single Stance).The identification accuracy of 7 subjects using intensity analysis reached 97%, even up to 99.42%.The others were about 95%. The algorithm obviously achieved a higher accuracy of sEMG recognition than the other algorithms such as root mean square and AR Coefficient. In the future, the feature of sEMG signal under different key gait phases may be used in the control of Functional Electrical Stimulation (FES) and other intelligent artificial limbs.