BrainGate, a consortium of researchers from universities including Stanford, Brown, and Case Western Reserve, has given a dozen patients the ability to control a cursor just by thinking about it. SOURCE: MARK HANLON/STANFORD UNIVERSITY
For the first 54 years of his life, Dennis DeGray was an active guy. In 2007 he was living in Pacific Grove, Calif., not far from the ocean and working at a beachside restaurant. He surfed most mornings. Then, while taking out the trash one rainy night, he slipped, fell, and hit his chin on the pavement, snapping his neck between the second and third vertebrae. DeGray was instantly rendered, as he puts it, “completely nonfunctional from the collarbone south.” He’s since depended on caregivers to feed, clothe, and clean him and meet most any other need. He had every expectation this would be the case for the rest of his life.
“My first six months were really something,” DeGray, now 64, says ruefully from his single room in a Menlo Park nursing facility, decorated with fairy lights, a National Lampoon poster, and a 6-foot-tall plastic alien. “And then the next two years were also something. And, frankly, this morning, it’s still something.” He operates his motorized wheelchair by blowing into a straw. Most of his days consist of TV and trips to the local park, the library, and neighbourhood restaurants, where familiar staff help him eat.
For the past year, though, the routine has been broken on Mondays and Wednesdays. Around noon, two or three scientists arrive at the nursing facility. They roll out a rack of computer equipment parked in a corner of DeGray’s room and plug a cable into a socket on the top of his head. Once he’s connected, a 1/6-inch-square silicon chip in his motor cortex allows him to move a cursor on a computer screen just by thinking about it.
This so-called brain-computer interface, or BCI, provides a way to directly measure neuron activity and translate it into information or action. To manipulate the cursor on his screen, DeGray imagines that his hand is resting on a ball on a table and that he’s trying to roll it in one of four directions: left, right, toward, away. When he first tried the system in September 2016, “it was like a bumblebee in the wind, bouncing around,” he says. Soon, though, he got the hang of it, and the researchers used his efforts to teach the computer to better interpret his brain activity. Today, with a keyboard laid out on the screen, DeGray can bang out nine and a half words per minute. If that doesn’t sound speedy in touch-typing terms, well, the Wright Flyer wasn’t a particularly fast aeroplane.
A cable leads from a socket installed on the top of a patient’s head to a rack of computer equipment that communicates with a nearby monitor.
DeGray has been working with BrainGate, a consortium of researchers from the likes of Stanford, Brown, and Case Western Reserve University that’s successfully treated a dozen patients. The BrainGate team is among a growing set of university scientists, government agencies, and startups trying to give humans the ability to sense, control, and communicate with the outside world through the power of thought.
So far these advances are limited to controlled settings, but there’s big money dedicated to getting them out into the world faster than most people imagined when DeGray broke his neck. Bryan Johnson, the founder of the payments service Braintree, has committed $100 million to a BCI startup called Kernel. Facebook Inc. is developing a skullcap it says will allow users to mentally type their thoughts at 100 words per minute. Tesla Inc. and SpaceX Chief Executive Officer Elon Musk are backing a similar technology from startup Neuralink that he says supports his vision of a “closer merger of biological intelligence and digital intelligence.” The Pentagon’s research and development arm, the Defense Advanced Research Projects Agency (Darpa), is funding nine BCI projects it aims to bring to the U.S. Food and Drug Administration for clinical trials in three to five years. Justin Sanchez, director of Darpa’s Biological Technologies Office, predicts that medical device makers will be able to apply BCI hardware to a wide range of projects. DeGray is focused on one in particular: bypassing damaged nerves to reconnect his brain and body. “Ten years from now,” he says, “a guy is going to fall down just like I did, and in short order, he’ll wake up in the morning, and someone will put his exoskeleton on, and he’ll get up and walk to Starbucks.”
The wiring together of brains and computers is a saddle-worn sci-fi trope. Think of William Gibson’s hacker heroes “jacking in” to cyberspace, or the captive humans plugged into the Matrix, or RoboCop. In practice, though, the brain is a lot tougher to hack. It contains 100 billion microscopic neurons, each connected to thousands of others. While some parts of the motor and sensory cortices correspond to parts of a person’s body, most elements of the brain, including the areas responsible for language and memory, aren’t as intuitively organized. In fact, we hardly understand them at all.
The least invasive tool for measuring brain activity is the electroencephalogram, or EEG, which works through an array of electrodes fastened to the scalp and measures the strength of the electric field in each spot. This kind of gear is safe, cheap, and imprecise, best suited to applications that ask researchers only to distinguish between the brain activity required for sharply contrasting thoughts: left vs. right, up vs. down. To restore function to quadriplegics, BCI devices need vastly better precision and speed. For now, the only way to achieve that is by affixing sensors directly to the cerebral cortex.
This brain-computer interface, or BCI, provides a way to directly measure neuron activity and translate it into information or action.
Cutting into the brain to insert electrodes is about as tricky and dangerous as you’d think. (Maybe a little more so.) But people have been doing it, albeit with some public outcry, since at least the 1950s, when controversial neurologist José Rodríguez Delgado experimented with the cortices of epileptics and schizophrenics. In the ’90s neuroscientist, Phil Kennedy implanted electrodes in the brains of subjects suffering from locked-in syndrome, a paralysis of almost all voluntary muscles besides those that control the eyes so they could type out messages. (The FDA halted Kennedy’s work because of safety concerns.) Around that time, Brown professor John Donoghue began developing neural interfaces to study how the brain turns thought into action, which ultimately led to the BrainGate project.
“The field is progressing very, very quickly,” says David Borton, head of the Brown Neuromotion Laboratory. The past year has been particularly impressive. Researchers at the University of Pittsburgh Medical Center connected touch sensors from a robot’s fingertips to a paralyzed man’s sensory cortex so he could feel what it was touching. At Case Western, scientists linked a paralyzed man’s motor cortex to a computer that electrically stimulated muscles in his arm, enabling him to bring a forkful of food from a dish to his mouth. At Brown, Borton’s team implanted electrodes and a wireless transmitter in a monkey’s motor cortex and connected it to a receiver wired to the animal’s leg, restoring its walking motion. Taken together, these procedures provide a roadmap for artificial workarounds of nervous system malfunctions caused by accident or disease. The minds of quadriplegic patients could be reconnected to their own muscles or patched into machines. Borton says it’s a question of when not if.
Just as Lasik surgery has gone from Kubrickian nightmare to the sort of thing you get done over a lunch hour, brain implants could come to be a reasonable intervention for conditions such as Parkinson’s, epilepsy, or chronic pain. They might even be used to improve healthy brains by adding memory storage or enabling communication by thought alone. Stare at the ceiling long enough, and it’s easy to worry about the darker possibilities of BCI. The kind of cybernetic fusion that gives us a doorway out of our bodies and minds could also give other people a way in. Once tiny robots can change people’s moods, what can’t they change? What does spam, social media addiction, or hacking look like inside your brain?
That’s a ways off, as DeGray notes. He’s been reading up on BCI research and is convinced that he’ll eventually be able to do a lot more than type nine and a half words a minute. “We’re building the foundation, learning how to directly control things from the cortex,” he says.
A keyboard laid out on the screen allows the patient to type.
Already several companies, including Raytheon Co. and Lockheed Martin Corp., have developed powered exoskeletons that augment the strength of healthy bodies. If scientists can develop sensors and actuators that allow quadriplegics to feel and manipulate objects, they can integrate human and exoskeleton into a fully functioning cyborg. That will be no small feat—neurologists don’t fully understand how our brains seamlessly coordinate sensation and action—but one-day paralysis will effectively be a solved problem.
The technology will also extend the distance between user and machine. Pressure-sensitive pads on a robot’s fingertips could feed into the sensory cortex of a user in the next room, the next state, or half a world away, and motor information travelling the other way could guide the robot hand to act. “It doesn’t really matter where the brain is located,” DeGray says. “I’ll be able to fly like a bird at some point. Literally, the sky’s the limit.”
DeGray is ready for a change. His breath-operated wheelchair can get him the 2½ miles from his bedroom to BrainGate’s Stanford lab and back, but the chair’s best feature, he says, lets him hike himself up an extra 13 inches by blowing into the straw to manipulate a digital menu on the chair’s screen. That 13-inch difference means “I can sit at a bar and watch the soccer game on TV and talk to the guy next to me, just like another guy.” Imagine, he says, stripping away the rest of the social barriers he feels. The potential to reconnect with people one-to-one, I suggest, could be enormous. “Enormous,” DeGray says. “All capital letters, double exclamation points at the back end. Enormous.”
– Jeff Wise