Multisession, noninvasive closed-loop neuroprosthetic control of grasping by upper limb amputees

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Upper limb amputation severely impacts daily living and quality of life.
• Current prosthetic control methods lack naturalness and require invasive procedures.
• Brain-machine interfaces (BMIs) offer potential for intuitive prosthetic control by translating neural activity.

Purpose of the Study:

• To evaluate the feasibility of a noninvasive electroencephalography (EEG)-based closed-loop BMI for controlling prosthetic grasp preshaping in transradial amputees.
• To assess neuroprosthetic performance without daily decoder recalibration.
• To identify factors influencing BMI stability and accuracy.

Main Methods:

• Two transradial amputees (A1 and A2) used an EEG-based closed-loop BMI system to control two grasping synergies.
• Participants underwent 12 sessions over 5 weeks, including an initial decoder calibration phase.
• Grasping success rate was measured as the primary outcome.

Main Results:

• Participant A1 achieved a consistent grasp success rate (63±3%), significantly above chance for 11 sessions.
• Participant A2, influenced by multiple medications, showed a lower success rate (32±2%), significantly above chance in only two sessions.
• EEG signal quality remained stable, but decoder variations indicated potential instability in short-timescale EEG features.

Conclusions:

• An EEG-based BMI is a viable approach for prosthetic control in amputees, warranting further research.
• Medication and other factors affecting brain activity require careful consideration to enhance BMI accuracy and long-term stability.