New brain-computer tech lets paralyzed patient talk in real time

A new investigational brain-computer interface (BCI) developed by researchers at the University of California, Davis, has enabled a paralyzed patient with Amyotrophic Lateral Sclerosis (ALS) to communicate in real time using a synthesized version of his own voice. ALS causes loss of muscle control, including speech, making communication difficult or impossible. This BCI system uses surgically implanted microelectrode arrays in the brain’s speech region to capture neural activity, which is then decoded by advanced AI algorithms to produce near-instantaneous audible speech. The technology significantly reduces the delay seen in previous speech neuroprostheses, allowing for more natural, spontaneous conversations with a delay as low as one-fortieth of a second. The system was tested on a 45-year-old participant in the BrainGate2 clinical trial, who was asked to attempt speaking sentences displayed on a screen while his brain activity was recorded. The AI model mapped his neural firing patterns to intended speech sounds, enabling the participant to control