Improving Performance of Pattern Recognition-Based Myoelectric Control Using a Desktop Robotic Arm Training Tool

James Austin1, Ahmed Shehata2, Michael Dawson2, Jason Carey2, Jacqueline Hebert2

  • 1University of Alberta BLINC Lab
  • 2University of Alberta

Details

Category

Poster Session

Sessions

10:00 - 17:00 | Tue 30 Oct | Foyer | B1P-E

Rehabilitation & Assistive Technologies

Full Text

Abstract

Performance using pattern recognition-based myoelectric prostheses is significantly impacted by user training with the selected control strategy. However, minimal research has been done into the effect of functional user training with different myoelectric control strategies, since doing so typically requires training and evaluating prosthesis users with differing device configurations and customized socket fittings. Intermediate platforms such as desktop-mounted robotic arms present an opportunity for consistent training of participants both able-bodied and with amputations. In this paper, a training environment and protocol for improving myoelectric prosthetic control with a desktop-mounted robotic arm was developed and assessed with pattern recognition as the control method. Pre-training and post-training performance for 10 able-bodied participants was evaluated using the Target Achievement Control test for 1, 2 and 3 degrees of freedom. Results showed significant differences in performance before and after 1 hour of desktop training. These results support the hypothesis that a desktop training protocol may improve performance with pattern recognition-based control.

Additional Information

No information added

Video

No videos found