Myoelectric Prosthesis
Myoelectric prostheses, controlled by electrical signals from muscles, aim to restore lost limb function by providing users with intuitive and dexterous control. Current research focuses on improving control algorithms, including exploring neural networks (like convolutional and recurrent networks) and transformer architectures to enhance gesture recognition and force estimation from electromyography (EMG) signals, often minimizing the need for cumbersome external sensors. This work also emphasizes improving user experience through the incorporation of haptic feedback, shared control systems, and more accessible prosthesis designs, ultimately aiming to increase the functionality and usability of these devices for amputees.
Papers
Haptic Shared Control Improves Neural Efficiency During Myoelectric Prosthesis Use
Neha Thomas, Alexandra J. Miller, Hasan Ayaz, Jeremy D. Brown
The Utility of Synthetic Reflexes and Haptic Feedback for Upper-Limb Prostheses in a Dexterous Task Without Direct Vision
Neha Thomas, Farimah Fazlollahi, Katherine J. Kuchenbecker, Jeremy D. Brown