A training strategy for learning pattern recognition control for myoelectric prostheses

People learn in different ways, the same is true for amputees learning to initiate the correct movement patterns in myoelectric prosthetics. First, the natural movements that are achievable are demonstrated. The amputee can then use a computer based training system driven by EMG and phantom limb movements to learn these movements. 

Consistent and distinguishable muscle patterns are the key to success. Consistency is achieved through repetition, while distinguishable is more challenging and must be monitored to define a need for modifications. Complex movements (e.g. the tendency to combine wrist supination with opening of the palm) can be confusing for prosthetics users. The use of a mirror allows patients to see the goal movements and exposes any limitations. 

Daily training with virtual control, first in clinic and then at home, is needed to achieve results. Outcome measures look at movement frequency, completion time and consistent muscle patterns. Problems remain in solving the clinical application, however pattern recognition control learning is tractable. > From: Powell et al., J Prosthet Orthot 25 (2013) 30-41. All rights reserved to the Authors.

Visit the Pubmed summary for more information or click here for the free full text version!

On top of that, read further into the basics of myoelectric prostheses here or watch an interesting video on mirror box therapy below: