A collaborative team of Duke neuroscientists, neurosurgeons, and engineers has developed a speech prosthetic that translates a person’s brain signals into spoken words, offering hope for individuals with neurological disorders that impede their ability to speak. The technology, featured in the journal Nature Communications, aims to improve communication through a brain-computer interface for conditions like ALS and locked-in syndrome. Current speech decoding technology is slow, operating at about 78 words per minute, while people typically speak at 150 words per minute. The team overcame this limitation by using high-density, flexible brain sensors with 256 microscopic brain sensors on a small, flexible implant. Although accuracy was 40% during initial testing, the researchers are optimistic about future cordless versions of the device. However, further development is needed to reach natural speech speeds.