Engineers at the University of California, Berkeley, have created a wearable biosensor device that recognizes hand gestures. Still in development, the UC team believes the device could eventually give patients control over prosthetic hands and other artificial body parts.

Health Tech Insider has covered a variety of technologies with prosthetic control applications. Recent innovations range from this complex brain-computer interface that required 10-hour electrode implantation brain surgery to this neural interface mouse introduced at CES 2021. The UC device — a non-invasive, lower-cost option compared to brain surgery — combines existing biosensors, signal processing, and AI algorithms in a flexible, lightweight arm cuff that detects electrical signals on the forearm. Although similar to the neural mouse, the UC device exhibits greater sensitivity, allowing for fine digital movements in a prosthetic hand.

To create the cuff, the UC team used conductive silver ink to print 64 electrodes on a flexible PET substrate. These electrodes pick up electrical signals in the forearm during gestures. A chip attached to the armband programmed with the AI algorithm reads and processes the signals, recognizing patterns associated with 21 different hand gestures.

The cuff requires initial calibration to train the algorithm to recognize individual signal patterns. Each user wears the cuff and makes each of the 21 hand gestures one time while a computer records the patterns and updates the chip on the armband. The algorithm uses hyperdimensional (HD) computing, an advanced form of AI learning. HD allows the algorithm to incorporate minor changes in the way a user performs a gesture into its pattern recognition system, so the user doesn’t have to worry about moving in the exact same way every time.

Implanted brain-computer interface technology, still in the early stages of development, may ultimately transform the lives of people with upper limb paralysis. External neural interfaces like the UC arm cuff could make it easier for patients with limb malformation or amputation to perform daily tasks, so that living with a prosthetic feels more natural.