Closed-Loop Continuous Hand Control via Chronic Recording of Regenerative Peripheral Nerve Interfaces

Area of Science:

• Biomedical Engineering
• Neuroprosthetics
• Rehabilitation Robotics

Background:

• Upper limb loss significantly impacts daily life, with current myoelectric prostheses lacking simultaneous and consistent control.
• Restoring natural arm control requires managing multiple prosthetic degrees of freedom over time.

Purpose of the Study:

• To implement a Kalman filter for continuous prosthetic hand control using intramuscular electromyography (EMG) signals.
• To evaluate control using regenerative peripheral nerve interfaces (RPNI) and intact muscles in non-human primates.

Main Methods:

• Implanted indwelling bipolar intramuscular electrodes in seven RPNIs and one intact muscle in two rhesus macaques.
• Recorded EMG signals and used a Kalman filter to predict finger position during a hand-movement task.
• Enabled real-time closed-loop neural control of a virtual hand.

Main Results:

• Successfully reconstructed continuous finger movement offline with high correlation and low RMSE.
• Achieved a 96.70% target hit success rate in real-time virtual hand control.
• Demonstrated consistent recalibration longevity with some variation in signal amplitude over time.

Conclusions:

• This study is the first to demonstrate chronic indwelling electrodes for continuous position control via a Kalman filter.
• The novel peripheral nerve interface combined with Kalman filtering represents a significant advancement for naturalistic prosthetic limb control.
• Further research can build upon these findings to improve prosthetic functionality for amputees.