Handle reaction vector analysis with fuzzy clustering and support vector machine during FES-assisted walking rehabilitation

  • Authors:

  • Weixi Zhu;Dong Ming;Baikun Wan;Xiaoman Cheng;Hongzhi Qi;Yuanyuan Chen;Rui Xu;Weijie Wang

  • Affiliations:

  • Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China;Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China;Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China;School of Science, Tianjin University of Technology, Tianjin, China;Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China;Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China;Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China;Department of Orthopaedics and Traumatology, Ninewells Hospital, University of Dundee, UK

  • Venue:
  • UAHCI'11 Proceedings of the 6th international conference on Universal access in human-computer interaction: applications and services - Volume Part IV
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper proposed Fuzzy clustering of C means and K means methods to extract the lateral features of lower limbs movement from handle reaction vector (HRV)data. With C-means clustering, the SVM recognition rate of lateral features was usually above 90% while, with K-means clustering, the recognition rate was close to 85%. The best recognition rate was even reaching up to 97% for some individual subject. Then the samples from all subjects were processed together with the cross-validation. Our experimental results showed that the HRV signal could be used with fuzzy clustering and support vector machine to effectively classify the lateral features of lower limbs movement. It may provide a new choice for FES control signal. The optimizing of the algorism parameters can be introduced to get better control in the future.