Scientists turn brain signals into speech, may help people who cannot talk

A state-of-the-art brain-machine interface created by University of California San Francisco neuroscientists can generate natural-sounding synthetic speech by using brain activity to control a virtual vocal tract – an anatomically detailed computer simulation including the lips, jaw, tongue and larynx.

The study was conducted in research participants with intact speech, but the technology could one day restore the voices of people who have lost the ability to speak due to paralysis and other forms of neurological damage.

Stroke, traumatic brain injury, and neurodegenerative diseases such as Parkinson’s disease, multiple sclerosis and amyotrophic lateral sclerosis (ALS, or Lou Gehrig’s disease) often result in an irreversible loss of the ability to speak. Some people with severe speech disabilities learn to spell out their thoughts letter-by-letter using assistive devices that track very small eye or facial muscle movements.

However, producing text or synthesized speech with such devices is laborious, error-prone, and painfully slow, typically permitting a maximum of 10 words per minute, compared to the 100 to 150 words per minute of natural speech.

From that work, Anumanchipalli and Chartier realized that previous attempts to directly decode speech from brain activity might have met with limited success because these brain regions do not directly represent the acoustic properties of speech sounds, but rather the instructions needed to coordinate the movements of the mouth and throat during speech.

“The relationship between the movements of the vocal tract and the speech sounds that are produced is a complicated one,” Anumanchipalli said.

“We reasoned that if these speech centers in the brain are encoding movements rather than sounds, we should try to do the same in decoding those signals.”

In their new study, Anumancipali and Chartier asked five volunteers being treated at the UCSF Epilepsy Center – patients with intact speech who had electrodes temporarily implanted in their brains to map the source of their seizures in preparation for neurosurgery – to read several hundred sentences aloud while the researchers recorded activity from a brain region known to be involved in language production.

Based on the audio recordings of participants’ voices, the researchers used linguistic principles to reverse engineer the vocal tract movements needed to produce those sounds: pressing the lips together here, tightening vocal cords there, shifting the tip of the tongue to the roof of the mouth, then relaxing it, and so on.

This detailed mapping of sound to anatomy allowed the scientists to create a realistic virtual vocal tract for each participant that could be controlled by their brain activity. This comprised two “neural network” machine learning algorithms: a decoder that transforms brain activity patterns produced during speech into movements of the virtual vocal tract, and a synthesizer that converts these vocal tract movements into a synthetic approximation of the participant’s voice.

The synthetic speech produced by these algorithms was significantly better than synthetic speech directly decoded from participants’ brain activity without the inclusion of simulations of the speakers’ vocal tracts, the researchers found. The algorithms produced sentences that were understandable to hundreds of human listeners in crowdsourced transcription tests conducted on the Amazon Mechanical Turk platform.

As is the case with natural speech, the transcribers were more successful when they were given shorter lists of words to choose from, as would be the case with caregivers who are primed to the kinds of phrases or requests patients might utter.

The transcribers accurately identified 69 percent of synthesized words from lists of 25 alternatives and transcribed 43 percent of sentences with perfect accuracy. With a more challenging 50 words to choose from, transcribers’ overall accuracy dropped to 47 percent, though they were still able to understand 21 percent of synthesized sentences perfectly.

“We still have a ways to go to perfectly mimic spoken language,” Chartier acknowledged. “We’re quite good at synthesizing slower speech sounds like ‘sh’ and ‘z’ as well as maintaining the rhythms and intonations of speech and the speaker’s gender and identity, but some of the more abrupt sounds like ‘b’s and ‘p’s get a bit fuzzy. Still, the levels of accuracy we produced here would be an amazing improvement in real-time communication compared to what’s currently available.”