Sergey Stavisky
University of California-Davis, CA·Neuroscience·Early-career
People with ALS or stroke can lose the ability to speak while their thinking stays intact. A brain-computer interface could restore their voice by reading brain activity as they try to talk and turning it into words.
- Fields
Neuroscience
Biomedical Engineering
Computer Science
- Topics
brain-computer interface
iBCI
speech neuroprosthesis
- Grants
- 1 known grant, through Aug 2027
What they work on
- Record hundreds of individual neurons with implanted microelectrode arrays in the ventral precentral gyrus and superior temporal gyrus to learn which features of speech these neurons encode and when during speaking.
- Use dynamical systems and neural-population analyses to separate movement-related, sensory feedback and error-processing signals within speech-area neural activity in BrainGate2 participants.
- Build a 'Brain-to-Words' decoder that converts attempted speech into speech sounds, called phonemes, and then into text, using artificial intelligence and language statistics to correct errors.
- Develop a 'Brain-to-Voice' interface that turns neural activity into synthesized speech sounds in real time, testing whether control-theory design and user learning give intelligible voice output.
More about the research
To build one, researchers first need to understand how networks of brain cells produce speech. This project pursues both goals: to understand speech production at the level of single neurons and to build a device that lets people communicate fluently again.
Why it matters
People who lose speech to ALS, stroke or other neurological conditions have few options for communicating. This work could lead to an implanted brain-computer interface that turns attempted speech into text or a synthesized voice in real time. The goal is speed and accuracy approaching natural speech, which could restore everyday conversation for people with paralysis.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Lee Miller
Northwestern University, IL·Biomedical Engineering
Build decoders that predict intended muscle activity (EMG) from primary motor cortex spiking recorded in humans with high-level spinal cord injury, trained with EMG, hand posture and force data from able-bodied people doing the same actions.
Karunesh Ganguly
University of California-San Francisco, CA·Neuroscience
Compare two neural decoding methods for their ability to give paralyzed patients stable, plug-and-play control of a robotic arm and hand with ECoG for at least 8 weeks without recalibration.
Jonathan Kao
University of California-Los Angeles, CA·Biomedical Engineering·Early-career
Train new recurrent neural network architectures on the decision-making and movement tasks animals perform, constrained by anatomy and neural recordings, then analyze them to propose mechanisms of cortical computation.
The school
University of California-Davis: 49 labs on PI Pages, $1.0B a year on research, 647 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.