In a new study, three clinical trial participants with paralysis used the BrainGate brain-computer interface to type via direct brain control at the highest speeds and accuracy levels reported to date.

[Editor’s note: CAUTION: Investigational Device. Limited by Federal Law to Investigational Use.]

The study, which appears in the journal eLife, was led by researchers at Stanford University, a partner in the BrainGate research collaboration that includes Brown University, the Providence Veterans Affairs Medical Center, Massachusetts General Hospital and Case Western Reserve University.

Two of the participants were enrolled by Stanford, and the third was enrolled by the Providence V.A. With one or two tiny electrode arrays implanted in the motor cortex of their brains, each participant thought about moving a cursor with their hand and arm to direct it over an on-screen keyboard. One participant typed 39 correct characters per minute, or about eight words per minute.

“This incredible collaboration continues to break new ground in developing powerful, intuitive, flexible neural interfaces that we all hope will one day restore communication, mobility and independence for people with neurologic disease or injury,” said study co-author Dr. Leigh Hochberg, professor of engineering at Brown and a neurologist and neuroscientist at Massachusetts General Hospital and the V.A. Rehabilitation Research and Development Center for Neurorestoration and Neurotechnology in Providence.

Hochberg directs the pilot clinical trial of the BrainGate system. Development of BrainGate began at Brown more than 15 years ago in the lab of neuroscience and engineering professor John Donoghue.

The newly reported performance gains were not small, said Stanford electrical engineering professor Krishna Shenoy, whose lab developed algorithms for the system, which decodes brain signals and turns them into digital commands. Instead, they are approaching a point where they might become practical for restoring communication capabilities for people with paralysis of all four limbs, including people who can no longer speak due to diseases such as ALS or a brainstem stroke.

“This study reports the highest speed and accuracy, by a factor of three, over what’s been shown before,” said Shenoy, co-senior author along with Stanford neurosurgery professor Jaimie Henderson. “We’re approaching the speed at which you can type text on your cellphone.”

Hochberg said a strength of the results — and the collaboration — was that it produced such promising results at different sites and in different participants, two of whom have ALS and one of whom sustained a severe spinal cord injury.

“An idea — a new algorithm, a new interface — is developed and refined at one site, and then ‘shipped’ to a consortium partner where it’s independently validated by different scientists and (perhaps more importantly) a different trial participant who may have a different form of paralysis,” said Hochberg, a member of the Brown Institute for Brain Science. “This helps us to know that our results can generalize and that we’re en route to developing a powerful neurotechnology and assistive device. No doubt that there is still a lot of research left to be done, but I’m pleased with the progress we’re able to report today.”

The study was funded by the National Institutes of Health (grants R01DC014034, R011NS066311, R01DC009899, N01HD53404 and N01HD10018), the U.S. Department of Veterans Affairs Rehabilitation R&D Service, the Stanford Office of Postdoctoral Affairs, the Craig H. Neilsen Foundation, the Stanford Medical Scientist Training Program, Stanford BioX-NeuroVentures, the Stanford Institute for Neuro-Innovation and Translational Neuroscience, the Stanford Neuroscience Institute, Larry and Pamela Garlick, Samuel and Betsy Reeves, the Howard Hughes Medical Institute, the MGH-Dean Institute for Integrated Research on Atrial Fibrillation and Stroke, and Massachusetts General Hospital.