Portable Neuroprosthetic Hand with Deep Learning-based Finger Control

This portable neuroprosthetic hand features a deep learning-based finger control neural decoder deployed on Jetson Nano. It is a self-contained unit with real-time control of individual finger movements. This system was evaluated on a transradial amputee using periveral nerve signals with implanted electrodes, with a finger control accuracy of 95-99% and latency of 50-120ms.