A stroked woman who lacked speech ability for many years has been helped to talk again.
Researchers have developed a technology that translates thoughts about speech into spoken words – in real time.
– We enable people with voice paralysis to be able to communicate again, says the researcher Gopala Anumanchipalli, at the University of California, in the United States.
It is researchers at the universities of San Francisco and Berkeley in the United States that have developed the technology that can translate thoughts about speech into spoken words in real time. They trained an AI model that translates thoughts, so-called neural activity, into sound units. It happens at the same time as the person thinks.
The new brain chip has been tested on a 47-year-old woman with paralysis, who lost her speech ability 18 years ago after a stroke.
“We enable people with voice paralysis to be able to communicate again,” says researcher Gopala Anumanchipalli, University of California, Berkeley, USA, which was part of the project.
Formulated their thoughts silently
Researchers and doctors at the University of California in San Francisco and Berkeley in the United States used electrodes to record brain activity while the woman silently formulated sentences in the tank.
The recorded nerve activity was then converted into real time into a synthetic speech, built with the help of the woman’s own voice from the time before the injury. This precisely this immediate reproduction is considered promising, as other types of brain-computer interface often cause a tangible delay, which can interfere with natural conversation flow.
– She tries to speak and without having to wait for a whole sentence or a word, and the computer can translate these signals into speech pieces which are then recorded to make the sound. So this is a very natural process in the sense that she does not have to wait to hear her voice, says Gopala Anumanchipalli.
Can help patients who lost speech ability
The ability to translate thoughts directly into meaning -bearing words can in the future help patients who lack speech ability, for example due to stroke, ALS or other neurological disease.
At the same time, researchers point out that the technology is still experimental and that more studies are needed to investigate security, stability and how well different patient groups can handle more advanced conversations.