Brain implant turns paralysed man’s thoughts into words


New York: US researchers have efficiently developed a ‘speech neuroprosthesis’ that has enabled a person with extreme paralysis to speak in sentences, translating alerts from his mind to the vocal tract immediately into words that seem as textual content on a display screen.

The know-how, developed by researchers from University of California-San Francisco (UCSF), was in a position to decode words from mind exercise at a charge of as much as 18 words per minute with as much as 93 p.c accuracy.

The man, in his late 30s, suffered a devastating brainstem stroke greater than 15 years in the past that severely broken the connection between his mind and his vocal tract and limbs. Since his harm, he has had extraordinarily restricted head, neck, and limb actions, and communicates through the use of a pointer hooked up to a baseball cap to poke letters on a display screen.

UCSF researchers surgically implanted a high-density electrode array over the affected person’s speech motor cortex and recorded 22 hours of neural exercise on this mind area over 48 classes and a number of other months.

The electrodes recorded his thoughts as mind alerts, which had been then translated into particular supposed words utilizing synthetic intelligence.

The workforce thus created a 50-word vocabulary — which incorporates words corresponding to “water,” “family,” and “good” — which they may recognise from mind exercise utilizing superior laptop algorithms.

“To our knowledge, this is the first successful demonstration of direct decoding of full words from the brain activity of someone who is paralysed and cannot speak,” mentioned Edward Chang, Professor and neurosurgeon on the UCSF.

“It shows strong promise to restore communication by tapping into the brain’s natural speech machinery,” Chang added. The examine is detailed within the New England Journal of Medicine.

Further, to check their strategy, the workforce first offered the affected person with brief sentences constructed from the 50 vocabulary words and requested him to strive saying them a number of occasions. As he made his makes an attempt, the words had been decoded from his mind exercise, one after the other, on a display screen.

Then the workforce switched to prompting him with questions corresponding to “How are you today?” and “Would you like some water?” As earlier than, the affected person’s tried speech appeared on the display screen. “I am very good,” and “No, I am not thirsty.”

“We were thrilled to see the accurate decoding of a variety of meaningful sentences. We’ve shown that it is actually possible to facilitate communication in this way and that it has potential for use in conversational settings,” mentioned lead creator David Moses, a postdoctoral engineer in Chang’s lab.


Please enter your comment!
Please enter your name here