A “voice nerve prosthesis,” which translates brain signals into words that appear on the screen, helped a severely paralyzed man who lost his ability to speak to communicate in writing.
It is believed that words and sentences were deciphered directly from cortical activity during an attempt to speak from a person who was paralyzed and unable to speak.
“When I started this project, there was no real guarantee that it would work. I was thrilled to see the various meaningful sentences decoded correctly. That was really amazing.” Dr. David Moses, lead author of the University of California, San Francisco, said.Francisco (UCSF) said Medscape Medical News..
“It underscores the potential of this approach to speak to people with severe paralysis and speech loss,” Moses said.
The· The study was announced In the July 15th issue of New England Journal of Medicine..
First 50 words
In recent years, several research groups have been working to use brain signals to avoid disability. Psychologically controlled prostheses allow paralyzed people to control artificial limbs and enter characters into text one by one through a brain-computer interface.
The UCSF team states that their approach differs in important respects from these efforts. It involves translating signals aimed at controlling the muscles of the vocal tract to speak words, rather than signals that move arms or hands to allow typing.
“It is our philosophy that speech is the ultimate goal of any communicative neural prosthesis,” Moses said.
A diagram depicting the UCSF team transplanting a high-density subdural multi-electrode array into the area of the sensorimotor cortex that controls speech.
The first patient to test a voice neuroprosthesis was a man in his late thirties who was suffering from the brainstem. stroke Fifteen years ago, it led to widespread paralysis and language attrition (anus).
The UCSF team implanted a subdural, high-density, multi-electrode array in the area of the sensorimotor cortex that controls speech.
During 48 sessions, they recorded 22 hours of brain activity while men tried to say individual words from a vocabulary set of 50 “valued” words that could generate over 1000 sentences.
When a man tried to say common words such as “water,” “family,” and “good,” advanced computer algorithms analyzed brain patterns. The algorithm was finally able to decode the word and the full sentence.
When the team asks a man a simple question on the screen, such as “How are you today?” And “How about water?” — The man’s attempted speech appeared on the screen: “I’m very good” and “No, I’m not thirsty.”
Detecting and computationally analyzing neural signals to decode each word takes about 4 seconds.
According to the UCSF team, the system can decode words from brain activity at speeds of up to 18 words / minute (median 15 words / minute) with up to 93% accuracy (median, 75%).
Pioneering work
In a news release, Dr. Edward Chan, MD, a senior author at the University of California, San Francisco, said these early results in a single patient demonstrated the principle, “using the natural speech mechanism of the brain. A strong promise to restore communication with. “
“There’s definitely a long way to go, and we need to do a lot of work before it’s a real market application and widely useful,” Moses said. Medscape Medical News..
In the future, they plan to test the voice neuroprosthesis in more patients affected by severe paralysis and communication problems. They are currently working to increase the number of words in the available vocabulary and improve speaking speed.
“For Anarthrians, decoding the intended speech from brain signals alone is an extraordinary achievement,” said Lee Hochberg, MD, and Sydney Cash, MD, MD, Massachusetts General Hospital’s Center for Neurotechnology and Nervous Recovery. Please note Ancillary editorial..
“In this pioneering demonstration of how a person with anal arteritis caused by a brainstem attack can generate text simply by trying to speak, Amyotrophic lateral sclerosis, Cerebral palsy, stroke, or other disorders are approaching clinical benefit, “they say.
“Ultimately, success is demonstrated by how easy it is for patients to share their thoughts with all of us,” they add.
The study was funded by a sponsored research agreement with the National Institutes of Health, philanthropy, and Facebook Reality Labs (FRL), which was completed in early 2021. Author d The closure is available on NEJM.org with the full text of this article.