Successful transfer of myoelectric skill from virtual interface to prosthesis control

Area of Science:

• Rehabilitation Engineering
• Motor Learning
• Prosthetics

Background:

• Prosthesis control is a learnable skill, enhanced by internal and augmented feedback.
• Augmented feedback, like biofeedback, can improve learning but its role in skill transfer to prosthetics is unclear.
• Optimizing training for skill retention may facilitate transfer to real-world prosthetic use.

Purpose of the Study:

• To determine if a training paradigm optimized for retention enhances the transfer of myoelectric skills to prosthesis control.
• To investigate the relationship between skill retention during training and prosthesis control performance.

Main Methods:

• Twelve limb-intact participants underwent five one-hour training sessions using a delayed feedback paradigm to promote retention.
• Participants learned a novel myoelectric skill via a computer interface.
• Prosthesis control was assessed pre- and post-training using tasks like grasp matching, the box and blocks test, and object manipulation.

Main Results:

• Prosthesis control performance significantly improved after five days of training.
• Retention metrics during training were significantly correlated with improved prosthesis control.
• No significant correlation was found between pre-test prosthesis performance and retention metrics.

Conclusions:

• Training protocols designed for skill retention can facilitate the transfer of abstract myoelectric control to prosthetic devices.
• This highlights the importance of skill acquisition principles in designing prosthetic training.
• Findings open new possibilities for developing effective prosthetic training strategies.