Juliano Pinto is a former athlete. The 29-year-old Brazilian was paralyzed from the waist down following a 2006 car crash. His prognosis: life in a wheelchair.
But this year, something changed. An international team of 150 scientists, in a sprawling project that cost $14 million, built an exoskeleton that straps to Pinto’s body. He controls it with his mind. And on June 12, during the opening ceremony of the World Cup in São Paulo, Brazil, he thought-controlled that exoskeleton to stand on his own two feet and kick a soccer ball. It looked like an early prototype of Iron Man. The project’s ringleader, Brazilian neuroscientist Dr. Miguel Nicolelis, had been hyping this moment for years, once boasting, “We’re getting close to making wheelchairs obsolete.” Such is the promise of Brain Control Interface technology, or BCI. The mysteries of the brain have captured the imaginations of scientists for centuries, and we are now, finally, inching toward a reality in which we can use computers to tap into the brain, decode its signals and use that information to operate machines, robots or exoskeletons just by thinking.
There are 6 million people in the United States who are paralyzed. Wide-spread, thought-controlled medical solutions won’t be available tomorrow or next month or even next year. But what if, some day, all of those people could walk again?
How it Works
The human body is controlled by electrical signals. These signals are the key to cracking the brain’s code. “Whenever we have a thought, whenever we move a muscle, there’s an electrical signal that travels from the brain to the muscle,” says Dr. Sanjay Joshi, an associate professor in the Department of Mechanical and Aerospace Engineering at UC Davis. Dr. Joshi and a team from the university worked on the Walk Again Project, the venture that got Pinto up and kicking. The Davis labs are at the forefront of BCI research.
That sounds simple enough. But neurons in the human brain don’t operate in a linear, step-by-step fashion. They’re a connected system. Thousands of neurons talk to thousands of other neurons, creating a mind-boggling number of permutations that stump our most powerful computers. It’s something Dr. Nicolelis calls “the cerebral symphony.”
“Every act that we perform, every feeling that we experience, comes from these electrical brain storms that interact with each other,” Nicolelis told an enthralled Jon Stewart on The Daily Show. “So we’ve figured out a way to record these electrical signals and extract from them … and translate these commands to machines in digital language.”
The past several years have seen rapid advancements. In 2008, Nicolelis trained a monkey to use its brain to operate a robot. Work on humans soon followed. The military jumped into the action, realizing that BCI technology could potentially help the 1,300 soldiers who have lost limbs in Iraq and Afghanistan. That theory turned into reality in 2012 when Jan Scheuermann, who’s paralyzed from the neck down, used her mind to control a robotic arm, even shaking hands with reporter Scott Pelley on 60 Minutes. “I’ve always believed there’s a purpose to my illness,” she said at the time. “I didn’t think I would find out what it was in my lifetime, and here came this study where they needed me.”
“Many people think that we’re reading people’s minds. But what we’re actually doing is asking people to create a very specific signal.”
Dr. Sanjay Joshi, Department of Mechanical and Aerospace Engineering at UC Davis
Jan is a member of an exceptionally small club. She volunteered to have the invasive-electrode surgery, and according to Dr. Allison’s estimates, there are less than two dozen of these implants in the world. But there’s no shortage of volunteers. When a video was posted of Jan’s mind-controlling marvel, one commenter wrote, “I am 56 yrs. old (sic), in 2004 was in an auto accident, became a quadriplegic. I would love to be a study partner with this program. I live in Louisiana. My e-mail is [redacted]. The best thing that I can imagine is hugging my grandchildren.”
The idea of quadriplegics embracing their family members remains the golden fleece, but the technology still has a long, long way to go. “There’s still a lot of mystery as to how individual neurons work,” says Dr. Allison, and both invasive and external electrodes have their drawbacks.
By analogy, he asks us to consider a soccer stadium. “Imagine that you’re at the 2014 World Cup Finals. Inside that stadium there are tens of thousands of individuals, kind of like a large group of neurons.” Now, imagine a microphone attached to one particular seat. If the microphone listened to the individual conversations of the fans, their statements would inform you directly as to what was happening on the field. But there would also be lots of audio of all the surrounding chit-chat and the general cacophony of the crowd. This represents an invasive electroencephalography (EEG).
“Now imagine that you’re in a non-invasive stadium,” Dr. Allison continues. “Your microphone is outside the building. You can never tell what one fan is doing. But by listening to the roar of the crowd, you can probably guess that someone just scored.” So you won’t be able to tell if those individual neurons are going to the bathroom or buying popcorn, but you can still learn something about the game.
This is how Dr. Joshi’s team at UC Davis is able to accomplish quite a bit with their non-invasive technology, even without knowing precisely what the neurons are saying. “Many people think that we’re reading people’s minds. But what we’re actually doing is asking people to create a very specific signal,” he says. “If you were to think about moving your left eyebrow, that would create a specific electrical signal that can be measured.”
That “left eyebrow” concept is key. For quadriplegics, the trick is to get them to move a muscle that they can still control, such as a left eyebrow or the muscle that wiggles the ear, and then record the brain signal. And here’s where the magic happens. Once scientists understand the electrical signal that corresponds to raising the left eyebrow, they can map that signal to an external action, such as moving a wheelchair. Think of it like recording a keyboard shortcut.
“What we’d really like to do is to pick a specific set of neurons and then understand exactly what their intention is. We’re just not at that stage yet,” Dr. Joshi says. This explains why, at this point, the basic functions of BCI are so crude (forward, back, left, right, kick). “Every time you have a thought, your brain has a slightly different electrical signal. There’s all sorts of things your brain is controlling at the same time.” The brain is noisy. Even if we can think “left eyebrow raise” without getting distracted by the cable bill and the television, our brain is still doing subconscious chores like regulating breathing and heartbeats. These chores can garble the signal.
All of this is why the Walk Again kick itself, frankly, lacked a bit of the expected ooomph. Juliano Pinto did not triumphantly strut onto the field. He did not defiantly kick the ball into the net. His foot moved forward, ever so slightly, to tap the ball a few feet on a red pad the size of a yoga mat. In some ways, it was a scientific miracle. And in some ways, it was a marker of how far we still have to go.
But there might be some lower-hanging fruit. The team at UC Davis is conducting experiments with BCI that, realistically, could actually help paralyzed people in the shorter term. “We’ve been focusing on the more basic things, like helping someone move a cursor around on a screen or turn on a television,” Dr. Joshi says. “That might not sound like much, but it would be a huge help to someone who is paralyzed.”
Scientists at UC Davis have also built a tiny robot that can scurry from room to room. It’s sort of like the little gizmo from “Rocky III,” but this robot can be controlled by the mind. It gets even cooler. The robot has cameras and a video screen, which could allow a quadriplegic to communicate in real-time with other people (or service animals) in the house. These capabilities could soon extend far beyond the home. Dr. Joshi’s team just completed a test wherein a paralyzed person at UC Davis controlled a robotic arm — located in New York.
The typical Brain Control Interface has four components, as explained by Dr. Brendan Allison, an expert of BCI at UC San Diego: Signal acquisition: An electrode scans messages from the brain. Technically, it’s called electroencephalography, or EEG. The electrode can either be an external device (like a cap on your head) or “invasive,” which requires surgery and actually grafts the sensors beneath your scalp. Signal processing: This decodes the signal from the brain, translating all of the computer’s 1s and 0s into something meaningful.The output device: The physical tool, such as an exoskeleton, robot or mechanical arm.The operating system: This governs the interface.
This is not a short-term play. “As scientists, we have to be very careful not to over-hype a specific technology,” Dr. Joshi says. “I work with a lot of disabled people, and it’s important to say that we’re not there yet. We still have a lot of work to do.”
That said, while the focus of BCI has been on assisting the paralyzed, commercial opportunities loom around the corner. A house being developed in Japan, for example, aims to allow its inhabitants to turn on light switches via thought. (This technology will be commercially available in 2020, according to the Japanese newspaper Asahi Shimbun.) And just a few weeks before we spoke, Dr. Allison gave a talk at Google, discussing the potential for “low-obtrusive electrodes.” He can’t share what they discussed, so this is purely my speculation, but … “Google Brain,” anyone?
Some of this technology is already available. Do you want to control computers with your mind? You can. A company called Intendix sells an electrode cap that lets you use your brain — without any keyboard, mouse or audio — to spell on a computer screen. It’s an expensive party trick, as it costs several thousand dollars and only lets you spell 10 words per minute. (Note: One common misconception is that you simply think of the letter W and it magically appears out of thin air. In reality, the system shows a virtual keyboard of 26 letters and 10 numbers, and then you train yourself to virtually “type” one of the letters with your brain.)
For perspective, in the overall life-cycle of BCI, we are now like the earliest computer programmers, slapping high-fives when the computer performs 1 + 1 = 2. Back then, how crazy would it have been to suggest that, in our very lifetimes, we could slip supercomputers into our pockets?
Once we are truly able to understand this cerebral symphony, the stuff of science fiction becomes very, very real, and we move from Iron Man to X-Men. Dr. Nicolelis, as a child, was inspired by the Apollo missions, and he sees his crusade as something of a race to the moon. He’s unabashedly bullish.
“Imagine living in a world where people use their computers, drive their cars and communicate with one another simply by thinking,” he writes in Beyond Boundaries. “No need for cumbersome keyboards or hydraulic steering wheels. No point in relying on body movements or spoken language to express one’s intentions to act upon the world.”
A world where we barely need our arms or legs. Inspiring or chilling? He leaves us with one more glimpse of the future. “In that future, back at your beach house, sitting in your favorite chair facing your favorite ocean, you may one day effortlessly chat with any of a multitude of people anywhere in the world over the Internet without typing or uttering a single word. No muscle contraction involved — just by thinking.”
For some, that might be anathema. For America’s 6 million paralyzed people, though, like Juliano Pinto of Brazil, that future could restore their former life. And it’s likely not a question of if, but when.