Stroke Survivor Regains Voice with Revolutionary Brain-Computer Implant

Researchers have created a gadget capable of converting thoughts related to speaking into audible words instantaneously.

Even though it remains experimental, they are hopeful about the brain-computer interface might one day assist those who cannot speak to express themselves.

A recent study detailed the use of this device on a 47-year-old woman suffering from quadriplegia, unable to speak for 18 years following a stroke. The implantation was conducted through surgical procedures as part of an ongoing clinical trial.

"It transforms her intention to talk into complete sentences," stated Gopala Anumanchipalli, who is one of the study’s authors, as mentioned in the journal Nature Neuroscience.

Other brain-computer interfaces (BCIs) used for speech generally experience minor lags when converting thought-generated sentences into spoken output via computers. These delays may interfere with smooth dialogue exchange, possibly causing misunderstandings and irritation, according to researchers.

"This represents quite significant progress in our area," stated Jonathan Brumberg from the Speech and Applied Neuroscience Laboratory at the University of Kansas in the U.S., who wasn’t involved with the research.

How the implant works

A group in California captured the brain activity of the woman as she internally vocalized sentences, utilizing electrodes for recording.

Before her injury, the scientists utilized a synthesizer constructed with her voice recordings to generate a speech sound as she would have uttered it. They then employed an AI model designed to convert brain signals into phonetic components for this purpose.

It functions much like current systems utilized for transcribing discussions during meetings or telephone conversations in real-time, stated Anumanchipalli from the University of California, Berkeley, based in the United States.

The implant is positioned over the brain's speech center, allowing it to listen in, and these signals are then converted into segments of speech that form sentences.

This is referred to as a "streaming approach," according to Anumanchipalli, where every 80 milliseconds of audio—which equates to roughly half a syllable—is transmitted to a recording device.

Waiting for a sentence to end isn’t what happens here," Anumanchipalli stated. "The system processes it as it goes along.

Brumberg mentioned that understanding speech at such a rapid rate could help keep pace with the speed of normal conversation.

He mentioned that incorporating voice samples would significantly enhance the natural quality of speech.

Although the project received partial funding from the U.S. National Institutes of Health (NIH), Anumanchipalli mentioned that it remained unaffected by recent cuts to NIH research funding.

Additional studies must be conducted before the technology can be widely implemented, but with consistent funding, it might be accessible to patients in about ten years, according to him.