Transfer of mode switching performance: from training to upper-limb prosthesis use

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Human-Computer Interaction

Background:

• Myoelectric prostheses offer advanced functionality through multiple articulation modes.
• Switching between these modes typically relies on pre-defined myosignals (triggers), yet training for this is understudied.
• Understanding effective training strategies is crucial for improving prosthesis usability.

Purpose of the Study:

• To evaluate if switching skills trained without a prosthesis transfer to actual prosthesis use.
• To determine the influence of feedback (visual vs. no visual myosignal) during training on skill transfer.
• To identify changes in clinically relevant performance measures and myosignal features during training.

Main Methods:

• A five-day pre-test-post-test study with two experimental groups and one control group.
• Experimental groups trained using a serious game with or without visual myosignal feedback.
• The control group used a laptop touchpad; all groups performed a prosthesis task in pre/post-tests.

Main Results:

• Both experimental groups demonstrated significant skill transfer from training to prosthesis use, unlike the control group.
• The degree of skill transfer was comparable between the two feedback conditions (visual vs. no visual myosignal).
• Improvements were observed in the clinical measure of 'accuracy' and the myosignal feature 'variation in phasing' during training.

Conclusions:

• Training switching skills using a serious game is effective and transfers to functional tasks.
• The type of feedback during training does not significantly impact the transfer of switching skills.
• While skills transfer, five training sessions were insufficient for achieving expert-level prosthesis performance.