A team of scientists at Stanford University has developed a brain-computer interface (BCI) that can decode a person’s inner speech, allowing those with paralysis to communicate more easily and naturally. Unlike previous systems, this new BCI doesn’t require individuals to physically speak, making it a potential game-changer for people with severe speech and motor impairments.
The researchers tested the BCI on four people with paralysis caused by either a stroke or amyotrophic lateral sclerosis (ALS), a degenerative disease that affects nerve cells controlling muscles. The participants had electrodes implanted in their brains as part of a clinical trial for controlling assistive devices with thoughts.
Artificial intelligence models were trained to decode inner speech and attempted speech from electrical signals picked up by the electrodes, resulting in an accuracy rate of 74% for decoding sentences internally “spoken” in the mind. The team also found that inner speech produced weaker brain activity compared to attempted speech.
To address concerns about potential misuse, the researchers developed a password-protected BCI. Participants could use attempted speech to communicate at any time, but the interface started decoding inner speech only after they spoke a specific passphrase.
The study’s findings offer real hope for the future of BCIs and potentially restore communication in individuals with severe impairments. With advancements in technology, it may be possible to develop devices that can decode complete sentences when a person isn’t explicitly thinking in words.
Source: https://www.livescience.com/health/neuroscience/new-brain-implant-can-decode-a-persons-inner-monologue