Amputees often feel disconnected from their bionic hands. AI could bridge the gap

"Researchers have built a prosthetic hand that, with the help of artificial intelligence, can act a lot more like a natural one. The key is to have the hand recognize when the user wants to do something, then share control of the motions needed to complete the task. The approach, which combined AI with special sensors, helped four people missing a hand simulate drinking from a cup, says Marshall Trout, a researcher at the University of Utah and the study's lead author."

"The success, described in the journal Nature Communications, is notable because "the ability to exert grasp force is one of the things we really struggle with in prosthetics right now," says John Downey, an assistant professor at the University of Chicago, who was not involved in the research. Problems like that cause many amputees to grow frustrated with their bionic hands and stop using them, he says."

"The latest bionic hands have motors that allow them to swivel, move individual fingers, and manipulate objects. They can also detect electrical signals coming from the muscles that are used to control those actions. But as bionic hands have become more capable, they have also become more difficult for users to control, Trout says. "The person has to sit there and really focus on what they're doing," he says, "which is really not how an intact hand behaves.""

A prosthetic hand integrates artificial intelligence with specialized sensors to detect user intent and share control of motions, enabling more natural actions. Four individuals missing a hand used the system to simulate drinking from a cup, achieving reliable grasping and sips when the AI and sensors assisted. Without shared control, users repeatedly crushed or dropped the cup. The system addresses difficulty in exerting appropriate grasp force, a common limitation that leads many amputees to abandon bionic hands. Modern bionic hands can move fingers and detect muscle electrical signals, but increased capability has made control more cognitively demanding compared with automatic neural circuits.

Read at www.npr.org