Extended home use of an advanced osseointegrated prosthetic arm improves function, performance, and control...

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Human-Machine Interface

Background:

• Restoring arm function with prosthetics is challenging.
• Advances in robotics, surgery, and machine learning are improving human-machine interfaces.
• Osseointegration (OI) and targeted muscle reinnervation (TMR) offer new possibilities for prosthetic control.

Purpose of the Study:

• To evaluate the long-term, at-home use of a dexterous prosthetic limb.
• To assess the effectiveness of machine learning-based electromyography (EMG) pattern recognition for prosthesis control.
• To investigate user integration, acceptance, and proficiency with advanced prosthetic technology.

Main Methods:

• One-year data logging of at-home use of the Modular Prosthetic Limb.
• Control via pattern recognition of EMG signals in an individual with transhumeral amputation, TMR, and OI.
• Monitoring of prosthesis usage, functional metrics, implant torque, and control performance.

Main Results:

• Continuous prosthesis usage increased by 1% weekly; functional metrics improved by up to 26%.
• Torque loading on the OI implant and prosthesis control performance increased monthly.
• EMG signal magnitude required for control decreased by up to 34.7% without performance loss, indicating improved control efficiency.

Conclusions:

• Extended, unconstrained use of an anthropomorphic prosthetic limb demonstrates significant functional benefits.
• Machine learning-based EMG pattern recognition enhances control efficiency and user proficiency over time.
• Customization of prosthesis movements for specific tasks contributes to continued usage and acceptance.