Skip to main content

Paralyzed man sips beer using robot arm he controls with his mind

Paralyzed man sips beer using robot arm he controls with his mind

Share this story

Erik Sorto hasn't been able to move his arms or legs in a decade, after a gunshot wound 13 years ago rendered him quadriplegic. But after scientists implanted chips into his brain three years ago, he's been able to move a robotic arm — to shake hands; play rock, paper, scissors; and yes, drink beer, according to a study in the journal Science.

"I want to be able to drink my own beer — to be able to take a drink at my own pace, when I want to take a sip out of my beer, and to not have to ask somebody to give it to me," Sorto says in a statement from the California Institute of Technology. "I really miss that independence."

"I really miss that independence."Sorto is one of a handful of people who have been given brain implants to help move objects with their minds since 2006, when a paralyzed man named Matthew Nagle moved a cursor on a computer using only his thoughts. Since then, scientists have been trying to refine the process to benefit other paralyzed patients. Today's study differs from most previous research in the area of the brain researchers targeted for implants — and may lead to better control for patients.

Previously, scientists focused on the primary motor cortex, a part of the brain which coordinates the contractions muscles require to move — but that made for jerky movements. That may be because human limbs can move about 27 different ways, says researcher Richard Andersen, a neuroscience professor at CalTech. Instead, his group targeted a different area in the brain, one he'd studied in animals, called the posterior parietal cortex. While the primary motor cortex focuses on specific muscle movements, the posterior parietal cortex is about planning movements. Information from the implants in the posterior parietal cortex transmit the intent to pick up a pint of beer and lets the computer figure out how to make the movement.

"I wanted to run around and just high-five everybody.""If we can indicate the goal, we can have smooth, natural movements toward the goal," Andersen says. The first time Sorto tried to control a robotic arm, 16 days after surgery, he pantomimed a handshake with a researcher. "It was the first time he'd moved a limb in 10 years," Andersen says. "It was amazing for him."

Sorto agrees. "I was surprised at how easy it was," he says. "I remember just having this out-of-body experience, and I wanted to just run around and high-five everybody."

It's not yet clear which area of the brain is better for implanting the chips — four square millimeters that record the activity of about 100 neurons — since the two methods haven't been compared head-to-head. But it may be that the real best way to create new prosthetic would be to embed chips in both, says Andrew Pruszynski, an associate professor of physiology and pharmacology at Western University. (He wasn't involved in the research but co-wrote an editorial accompanying it.) That way, the computer gets information about intent and precise movements, more closely mimicking how we control our own limbs. "You’d have high-level signal of what the person wanted to achieve but get some info on the low-level signals for more precise control," he says.

It's a tough problem to solveThe implants aren't ready for prime time yet, Andersen says, and won't be for a while. That's because right now, Sorto has wire bundling connecting his implant to the computer that controls his robot limb. The wire bundling goes through a plug in his skin — which could be a prime site for infections. (Sorto hasn't had any infections there yet, Andersen says.) Ideally, the implants would need to be wireless, but the amount of information coming out of the chips is so large, it's a tough problem to solve, Andersen says.

Infections aren't the only worry. The chip is like a microscopic pincushion that's pushed into the brain, Pruszynski says. There's some damage associated with that, and the recovery process may mean that some of the points of contact might be lost over time. What's more, the body is a corrosive place — so the implant has to be stable for a long period of time, because it's not feasible to repeat serious surgery every 5 years, he says.

like a microscopic pincushion that's been pushed into the brain But what might really improve the device would be feedback. Right now, Sorto is controlling his limb by watching it. But that's not as fine-grained a movement as he might ideally like — it's like trying to move your hand when your arm's fallen asleep, and you can't properly feel where in space it is, or any kind of pressure against it, Pruszynski says. Andersen agrees that feedback would improve control, and his group is considering ways to pipe information into the brain, as well as out of it.

Andersen's group has implanted chips in two more patients, and will be exploring the technique further, in the hopes of creating touch feedback. Maybe moving objects or playing video games with your mind isn't going to be routine any time soon — but for patients like Sorto, the benefits go beyond just drinking his own beer. The study has given him a sense of purpose. "As much as the project needed me, I needed the project," Sorto says. "It gives me great pleasure to be part of the solution for improving paralyzed patients' lives."

Embedded video by the Caltech team