Control Prosthetic Fingers Using Electromyography as a Sensing Device for Hand Gestures Recognition

This manuscript presents a study on enhancing the control of electrically powered prosthetic fingers using electromyography (EMG) signals. The proposed system incorporates the Myoarmband for gesture recognition. EMG signals are captured using sensors and translated into voltage values. These voltages are then utilized to control the motion of the prosthetic fingers, enabling intuitive and natural movement. Additionally, a deep learning approach was used to train continuous data collected from the sensors to accurately predict the desired sequence of voltage values. The neural network analysis allows for the assessment of the reasonability of the voltage sequence, providing insights into potential issues such as a corrupted myoarmband or underlying nervous system disorders. The experimental results demonstrate significant improvements in finger movement identification, showcasing the potential for advanced control systems that bridge the gap between human intent and artificial limb functionality.