AI TRENDS | Brain-Computer Interface Study Decodes Speech and Gestures From One Implant

Researchers at the University of California, San Francisco developed a new brain-computer interface system that, for the first time, simultaneously decoded speech and related body movements from a single brain implant and drove a digital avatar to express them. According to Sina Finance, the study was published on September 14 in Nature Neuroscience. Samantha Brossler, the paper's first author, said the research aims to move brain-computer interfaces from restoring a single function at a time toward restoring multiple communication and movement modes through one brain interface. She said gestures can add meaning, emphasis, emotion, and context, and in some cases can fully replace speech, such as nodding to agree, shaking the head to refuse, waving, shrugging, or giving a thumbs-up. The proof-of-concept study used 10 and 4 communication gestures, respectively, in two participants. The system used machine learning to accurately decode words and gestures from both participants' thoughts. The research team next aims to expand the method to a larger vocabulary and a wider range of continuous body movements. Effective communication prosthetics could help millions of patients who retain intact or near-intact cognition but have lost speech because of stroke, neurodegenerative disease, amyotrophic lateral sclerosis, and other brain disorders. In the long term, scientists hope the technology could also help people with cerebral palsy and autism who have difficulty speaking. The field has drawn growing research attention and investment, with active companies including Elon Musk's Neuralink and Precision Neuroscience.