Virtual Integration Environment as an Advanced Prosthetic Limb Training Platform.

Area of Science:

• Rehabilitation Engineering
• Human-Computer Interaction
• Neuroprosthetics

Background:

• Individuals with upper extremity (UE) loss face functional and psychosocial challenges.
• Advanced myoelectric prostheses offer intuitive control but require complex training.
• Virtual reality (VR) simulation is explored as a training tool for prosthetic control.

Purpose of the Study:

• To investigate the efficacy of a VR simulator for teaching dexterous prosthetic control to individuals with UE loss.
• To assess the ability of participants to learn and execute complex prosthetic movements using VR.

Main Methods:

• Thirteen military personnel with UE loss underwent passive and active motor training using a VR simulator.
• Participants followed avatar motions and controlled virtual limbs via surface electromyography (sEMG).
• Training involved three sets of increasing complexity (Basic, Advanced, Digit) with performance scoring.

Main Results:

• Passive training showed mean classification accuracy of 43.8%.
• Active training achieved >95% classification accuracy (prosthetic acceptance threshold) for Basic sets in all participants, and for Advanced/Digit sets in 50%.
• Significant improvements in active motor scores were observed over time for Basic and Advanced sets.

Conclusions:

• VR training platforms can efficiently train individuals with UE loss in advanced prosthetic control paradigms.
• Participants learned to generate distinct muscle contraction patterns for accurate prosthetic commands.
• Results support the viability of VR for training users of advanced myoelectric prostheses with pattern recognition controls.