Design and Implementation of a Prosthesis System Controlled by Electromyographic Signals Means, Characterized with...

Area of Science:

• Biomedical Engineering
• Robotics
• Artificial Intelligence

Background:

• Millions worldwide experience limb loss, necessitating advanced prosthetic solutions.
• Technological integration in prosthetics offers improved quality of life.
• Developing affordable, natural-movement prostheses is a significant scientific and social goal.

Purpose of the Study:

• To propose a low-cost electromyographic (EMG) electronic system for prosthetic control.
• To design and implement a functional hand-type prosthesis utilizing advanced signal processing.
• To enable natural and precise limb replication through a neural network-based controller.

Main Methods:

• Acquiring EMG signals from a user's healthy hand during various movements.
• Processing these EMG signals using a neural network controller trained on twenty signal characteristics.
• Developing a hand prosthesis that replicates controlled motions in real-time.

Main Results:

• The proposed neural network model achieved 95.2% accuracy in computer tests.
• Real-world experiments demonstrated 93% accuracy for the prosthetic hand control.
• Online execution yielded better response times compared to offline testing.

Conclusions:

• The developed low-cost EMG system and prosthetic hand offer a viable solution for limb amputees.
• The neural network approach with extensive feature extraction enhances prosthetic control accuracy.
• This technology presents a significant step towards affordable and functional advanced prosthetics.