It's time once more to check in on the pioneers of brain computer interface technology.
The last time the world heard from Andrew Schwartz and his colleagues at the Motorlab, he was teaching monkeys how to pluck marshmallow treats with a thought-driven robotic arm. This year, those lovable macaques will finally step aside and give the humans a try.
With funding from DARPA and the NIH, researchers at the University of Pittsburgh Medical Center, led by Dr. Schwartz, will soon begin clinical trials on two different techniques for plugging into human brain power—one which collects neural impulses with an array of electrodes placed on the surface of the brain, and another that implants the electrodes directly into the tissue.
The first approach (ECoG, or electrocorticography) records a signal from a large population of neurons, while the second (SUR, or single unit recording) listens in on single cells. Both data sets, however, will be filtered into usable information that can then drive complex tasks with computers and robotic limbs.
But this takes training on both the human and machine ends of the equation. The algorithms in the machine must first learn the patterns of brain activity that occur when the human is either imagining or physically carrying out a task. As the system becomes familiar with the neural commands, the patient's control over it becomes more seamless.
The researchers took a preliminary look at how this training works by temporarily implanting one of the arrays in a patient undergoing open brain surgery as treatment for severe epilepsy. A video of the results shows a woman trying to play a computer game by imagining the moves. When she loses control of the ball on the screen, the patient throws out a physical cue—raising her am—to help the computer refine its estimation of her intentions.
Michael Boninger, the lead physician on the UPMC trials, says that training happens very fast. "You can see some changes in a matter of a session or two," he says.
He hopes to recruit three people for each trail, all with spinal cord injuries—which shouldn't be terribly hard, given the enthusiasm Boninger has encountered from people with severe disabilities.
"We get a ton of excitement from people we talk to," he says. "It has the potential to be the single most enabling technology for people with disabilities."
You can look for the experiments to be up and running this summer.