A new study describes the testing of the device on a 47-year-old woman with quadriplegia who was unable to speak for 18 years after suffering a stroke. Doctors implanted the device in her brain during surgery as part of a clinical trial.
"It converts her intention to speak into fluent sentences," said Gopala Anumanchipalli, co-author of the study published Monday in the journal Nature Neuroscience.
Other brain-computer interfaces (BCIs) for speech often have a slight delay between thoughts of sentences and their computerized verbalization. These delays can disrupt the natural flow of conversation, leading to misunderstandings and frustration, according to researchers.
"It's a pretty significant advance in our field," noted Jonathan Brumberg from the Speech and Applied Neuroplasticity Lab at the University of Kansas, who was not involved in the study.
A team in California recorded the woman's brain activity using electrodes as she mentally articulated sentences. Scientists used a synthesizer they built with her pre-injury voice to generate speech sounds similar to what she would have produced. They then trained an artificial intelligence model to translate neuronal activity into sound units.
It works similarly to existing systems that transcribe meetings or phone calls in real time, explained Anumanchipalli from the University of California, Berkeley.
The implant is placed over the speech center in the brain to capture signals, which are then translated into speech fragments forming sentences. It is a "real-time" approach, noted Anumanchipalli, where each 80-millisecond speech fragment - approximately half a syllable - is sent to a recorder.
"It doesn't wait for a sentence to finish," Anumanchipalli explained. "It processes on the fly".
Decoding speech so quickly has the potential to maintain the fast pace of natural speech, said Brumberg. Additionally, using voice samples "would be a significant advancement in the naturalness of speech".
Although the work was partially funded by the National Institutes of Health (NIH), Anumanchipalli assured that it was not affected by recent NIH research cuts. More research is still needed before the technology is ready for widespread use, but with "sustained investments," it could be available for patients within a decade, he claimed.