🦾 Carnegie Mellon Debuts Noninvasive Mind-Controlled Robotic Hand

Carnegie Mellon researchers have developed the first noninvasive system that enables users to control individual robotic fingers in real-time using brain signals, according to a study published in Nature Communications.

The system utilizes EEG to detect brain activity associated with finger movement and motor imagery, then translates these signals into robotic finger actions. A deep neural network decodes the user's intent in real-time, with a new fine-tuning method improving accuracy from noninvasive input.

Read More