Nonlinear Mapping From EMG to Prosthesis Closing Velocity Improves Force Control With EMG Biofeedback

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Human-Machine Interface

Background:

• Controlling myoelectric prosthesis grasping force using EMG biofeedback is challenging due to increased signal variability with higher muscle contractions.
• Existing linear mapping methods struggle to maintain performance as desired force increases.

Purpose of the Study:

• To implement and validate a novel EMG biofeedback system using nonlinear mapping to enhance myoelectric prosthesis force control.
• To investigate the effectiveness of nonlinear mapping compared to linear mapping and no feedback.

Main Methods:

• Developed an EMG biofeedback system employing nonlinear mapping, where EMG signal intervals are mapped to prosthesis velocity intervals.
• Evaluated the system with 20 non-disabled subjects performing force-matching tasks and four transradial amputees performing a functional task.
• Compared performance across conditions: no feedback, linear mapping biofeedback, and nonlinear mapping biofeedback.

Main Results:

• EMG biofeedback significantly improved the success rate of producing desired force compared to no feedback (65.4% vs. 46.2%).
• Nonlinear mapping yielded a higher success rate than linear mapping (62.4% vs. 49.2%).
• The highest success rate (72%) was achieved with EMG biofeedback combined with nonlinear mapping in non-disabled subjects, a trend also observed in amputee subjects.

Conclusions:

• EMG biofeedback is effective in improving myoelectric prosthesis force control.
• Nonlinear mapping is a superior approach to linear mapping, effectively mitigating the challenges posed by increased myoelectric signal variability during strong muscle contractions.
• The proposed nonlinear mapping strategy offers a promising advancement for intuitive and accurate prosthetic control.