Progressive Rehabilitation Based on EMG Gesture Classification and an MPC-Driven Exoskeleton

Area of Science:

• Neuroscience and Biomedical Engineering
• Rehabilitation Robotics
• Assistive Technology

Background:

• Stroke is a major cause of long-term motor disability, affecting millions globally.
• Physical rehabilitation is crucial for restoring mobility after stroke.
• Existing rehabilitation methods may lack intensity, repetition, or adaptability.

Purpose of the Study:

• To design and develop a robotic hand exoskeleton for post-stroke motor rehabilitation.
• To create an adaptive control system that responds to patient effort and fatigue.
• To enhance the effectiveness and personalization of hand rehabilitation therapy.

Main Methods:

• Electromyographic (EMG) signal acquisition from forearm muscles.
• Automatic estimation of movement intention for five distinct hand gestures.
• Development of a predictive adaptive control algorithm to manage three levels of muscle fatigue.
• Integration of EMG-based intention detection with exoskeleton actuation.

Main Results:

• The robotic exoskeleton successfully interprets EMG signals to infer movement intentions.
• The adaptive control system effectively compensates for varying levels of muscle fatigue during exercises.
• The system provides repetitive, intense, and personalized assistance during rehabilitation therapy.

Conclusions:

• The designed robotic hand exoskeleton offers a promising approach to support stroke rehabilitation.
• EMG-based intention detection and adaptive control enhance therapeutic assistance.
• This technology has the potential to improve functional recovery and quality of life for stroke survivors.