Arm Orthosis/Prosthesis Movement Control Based on Surface EMG Signal Extraction

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Control Systems

Background:

• Motorized orthoses require intuitive control systems.
• Electromyography (EMG) signals offer a promising biofeedback mechanism.
• Accurate EMG-based prediction is crucial for seamless prosthetic movement.

Purpose of the Study:

• To develop and evaluate an EMG-based control system for motorized arm orthoses.
• To predict intended lower arm movements using biceps and triceps EMG signals.
• To enable naturalistic movement reproduction via powered orthotic devices.

Main Methods:

• EMG signals captured from biceps and triceps using biometrical sensors.
• Signal filtering and processing via a dedicated acquisition system.
• Control algorithms incorporating autoregressive (AR) models and neural networks.

Main Results:

• Experimental validation of the EMG-based control system.
• Achieved a maximum prediction accuracy of 91% for desired movements.
• Optimal performance identified with a fourth-order AR-model and 100ms block length.

Conclusions:

• EMG-based control is effective for motorized orthoses.
• AR models provide accurate prediction of lower arm movements.
• This technology enhances the potential for advanced prosthetic limb control.