Discrete Action Control for Prosthetic Digits.

Area of Science:

• Biomedical Engineering
• Rehabilitation Robotics
• Neuroprosthetics

Background:

• Current upper-limb prostheses often lack intuitive control for multiple degrees of freedom (DOFs).
• Existing methods like position control face limitations in dexterity and clinical applicability.

Purpose of the Study:

• To introduce and evaluate "action control," a novel paradigm for simultaneous and independent control of multiple DOFs in upper-limb prostheses.
• To compare the performance of action control against the state-of-the-art position control method.

Main Methods:

• Developed a real-time myoelectric control system using surface electromyography (EMG) and multi-output, multi-class classification.
• Implemented action control where a decoder classifies movement intent (open, close, stall) for each DOF.
• Evaluated the system with unilateral and bilateral amputees controlling a six-DOF bar interface.

Main Results:

• Action control significantly outperformed position control in both task- and non-task-related measures.
• Median task performance improvements ranged from 20.14% to 62.32% for participants.
• All participants preferred action control over position control in qualitative assessments.

Conclusions:

• Action control offers a promising approach to enhance the dexterity of upper-limb prostheses.
• The method's requirement for discrete labels simplifies clinical implementation and suits bilateral amputees.
• This study marks the first demonstration of myoelectric digit control in bilateral upper-limb amputees.