Scientists have developed a device that can translate thoughts into spoken words in real time. Although it is still experimental, they hope that the brain-computer interface can one day help give voice to those who cannot speak.
The implant was placed in a 47 -year -old woman with quadriplegia who could not speak for 18 years after a stroke. Doctors implanted it in their brain during surgery as part of a clinical trial.
“Turn your intention to speak into fluid sentences”said Gopala Anumanchipalli, co -author of the study published Monday in the journal Nature Neuroscience.
Other brain-computer interfaces typically have a slight delay between prayer thoughts and computerized verbalization. Such delays can interrupt the natural flow of conversation, which potentially leads to misunderstandings and frustration, the researchers declared.
“It is a fairly large advance in our field,” said Jonathan Bumberg of the speech laboratory and applied neuroscience of the University of Kansas, who was not part of the study.
A team in California recorded the brain activity of women using electrodes while she pronounced sentences in silence in her mind. The scientists used a synthesizer that they built using their voice before their injury to create a sound similar to their voice. They trained an AI model that translates neuronal activity into sound units.
It works similar to the existing systems used to transcribe meetings or telephone calls in real time, explained Anumanchipalli, of the University of California, Berkeley.
The implant itself is placed in the part of the brain that controls speech in such a way that you can listen to sounds, and those signs translate into speech fragments that make up sentences. It is a “transmission approach,” said Anumanchipalli, with each fragment of speech 80 milliseconds, approximately half of a syllable, sent to a recorder.
“It doesn’t wait for a prayer to finish,” said Anumanchipalli. “He is processing him on the fly.”
Decoding speech so quickly has the potential to follow the accelerated pace of natural speech, Bumberg said. The use of voice samples, he added, “would be a significant advance in the naturalness of speech.”
Although the work was partially funded by the National Health Institutes, Anumanchipalli said he was not affected by the recent cuts in that agency. More research is needed before technology is ready for generalized use, but with “sustained investments”, it could be available for patients within a decade, he said.
