Development of EMG-based mode and intent recognition algorithms for a computer-controlled above-knee prosthesis.

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Biomechanics

Background:

• Conventional above-knee prostheses lack adaptive control, limiting normal leg function.
• Computer-controlled prostheses can implement adaptive control using sensory inputs and gait data.
• Accurate determination of locomotion modes and transitions is crucial for prosthetic control systems.

Purpose of the Study:

• To investigate the use of electromyography (EMG) signals from hip muscles for adaptive prosthetic control.
• To determine if EMG signals can accurately differentiate between various locomotion modes.
• To assess the feasibility of recognizing transitions between different gait patterns.

Main Methods:

• EMG signals were recorded from the gluteus maximus, gluteus medius, and tensor fasciae latae muscles.
• Participants included two groups of 12 healthy individuals and one prosthetic patient.
• Locomotion modes tested were level walking, ramp ascent, and ramp descent (6 and 9 degrees).

Main Results:

• Distinct EMG activity patterns were observed for different locomotion modes.
• The study demonstrated the possibility of discriminating between level walking, ramp ascent, and ramp descent using EMG signals.
• EMG data showed potential for recognizing transitions between gait modes.

Conclusions:

• EMG signals from hip muscles show promise for enabling adaptive control in above-knee prostheses.
• Accurate mode recognition is feasible, supporting the development of finite state control approaches.
• Challenges remain in the practical implementation of EMG-based prosthetic control.