Exploring Position, Velocity, and Torque Control via Minimal sEMG Calibration for Robotic Ankle Prostheses

Area of Science:

• Biomedical Engineering
• Robotics
• Neuroscience

Background:

• Myoelectric control systems are increasingly used in lower-limb prosthetics to interpret user intent.
• Current AI-based systems face challenges with high latency and extensive training data requirements.

Purpose of the Study:

• To introduce a minimalist ridge regression approach for interpreting surface electromyography (sEMG) signals.
• To enable real-time control of ankle prosthetics with minimal calibration.

Main Methods:

• Developed a proof-of-concept system with three open-loop control modes: Position, Velocity, and Torque.
• Utilized a minimalist ridge regression algorithm for sEMG signal interpretation.
• Conducted Target Achievement Control tests with five able-bodied participants.

Main Results:

• Velocity control demonstrated the highest Completion Rate (83.3%), significantly outperforming Torque control (40.0%).
• Position control showed solid performance (71.4% CR).
• Differences in Overshoot and Movement Time were not statistically significant across modes.

Conclusions:

• Ridge regression shows promise for real-time myoelectric control, offering a potential solution for affordable, low-latency prosthetic operation.
• The developed system requires minimal calibration, making it more accessible.
• Further research can explore advanced control strategies based on this minimalist approach.