Originally Published by motherboard.com on June 06, 2013
For brain scientists Greg Siegle and Nicole Prause, understanding the brain is really tricky, and it's especially tricky when your subjects are in the middle of activities.
You can't, for instance, ask them to do their thing inside an fMRI machine, or with a wig of cables attached to their head. “All these brain-measuring systems have significant limitations,” says Siegle, a neuroscientist who has spent more than a decade poking at the brain using all sorts of sophisticated psychophysiology and neuroimaging tools.
With these, he says, “you can’t move your head very much, it costs a heck of a lot to buy, they take a long time to set up, and I could never really bring them out of the lab.”
Siegle isn't ashamed to say that his answer to this particular problem is a faddish-looking toy originally geared toward video gamers. The Emotiv, which is made by an Australian company and comes in $299 and $750 flavors, looks like a cyberpunk skull cap, with 14 electrode arms that branch out from the back of your head to surround your skull.
The device is a cheaper, wireless, and more user-friendly version of a standard EEG machine, the kind that scientists use to non-invasively detect the faint electrical signals in your brain, the brainwaves that fluctuate as you think and feel.
When it was released in 2009, the Emotiv promised to turn gamers’ brains into joysticks, and make the act of casting a spell, for instance, a matter of simply thinking of one. A parade of affordable brain-computer interfaces (BCI) made by companies like NeuroSky and NeuroVigil have shipped in the Emotiv's wake; a giant market, however, has not.
When they're working properly (they can be imprecise and sometimes annoyingly non-responsive), they do something that can border on magic; years after the technology hit the market, tech blogs still ooh and ahh over them. But the idea of mind-controlling your video game has mostly remained in the realm of sci-fi gimmick.
But the nerdy-looking toy's cult following has loftier goals
Engineers, hackers, scientists and artists have been mining the Emotiv's brain data, parsing waves and bringing them places far outside the walls of a video game. “The data’s very clean,” is Siegle's sober assessment of his toy headband. “We've been very impressed with it and we’ve incorporated it into some of our major research protocols.”
“The big goal is to try and really asses people where they are,” he says. “Depressed people are, for example, going over and over negative things for minutes.” People in the middle of sex, meanwhile, are going through something completely different. But in these and other cases, “behavioral measures aren’t always adequate. We want to actually see processes unfold and understand mechanisms in the time-course of minutes.”
The Emotiv headset, by contrast, has only 14 non-stick electrodes, situated on thin plastic arms that conform neatly around the skull. These electrodes allow the helmet to measure, its maker claims, four mental states (based on brainwaves) and thirteen conscious thoughts.
The signals it collects aren’t just electroencephalographic however: its gyroscopes allow the Emotiv to measure head movements, and a separate set of electrodes cover facial movements. (The electrodes for this rely on a cousin of EEG, electromyography, or EMG, which records the electrical activity of facial muscles).
All of its data is then transferred wirelessly over Bluetooth to a computer. Each new user must calibrate the device to their brain, but getting started is as simple as moisturizing the electrodes (a saline solution improves conductivity), slapping on a sci-fi headband, starting the software, and flipping a couple switches.
Like any brainwave-reading tool, the Emotiv is incapable of accessing the activities of neurons that sit below the surface of the brain, where most neural activity takes place. You need deep electrodes for that (these tend to be off limits for human subjects, but not for rats and everything else).
But what the Emotiv headset lacks in detail, it makes up for in its robust performance, collecting clean, uninterrupted data without a phalanx of cables, and despite movements of the head. During sex, Siegle has observed, there is a lot of head movement.
“We’re getting extremely strong results with the Emotiv,” he says. When he encounters skepticism towards BCI helmets in the neuro community, Siegle says he’ll somes challenge researchers to hand over their experiment protocols and run them on the Emotiv. On the sidelines of the Society for Psychophysiological Research conference in New Orleans last year, Siegle said he interrupted a lasagna dinner with a prominent EEG researcher in order to demonstrate his toy helmet.
“It took us four minutes to set it up, he ran the protocol for ten minutes and it took us six minutes to analyze it,” Siegle says. His colleague was impressed. “The data looked just like they do off his expensive systems at his home lab. It’s really very clean.” What's clean? It can't be as precise as a bigger system, could it? “Well, it’s cleaner than I would’ve expected. It’s clean enough that I will use it in research.”
The first consumer brain-computer-interface system arrived on shelves in 2003, produced by a Swedish electronics company called Interactive Product Line. Their system revolved around little more than a game called Mindball. The object of the game was simple: two opponents must wear headbands connected to biosensors reading their brain signals.
The one who relaxes the most while focusing on a small ball controlled by a BCI-formatted table can roll the ball into their opponent’s goal for a win. It was a cool idea in theory, but it came with a very un-cool $20,000 price tag, which is a bit steep for a toy if you're not Richard Garriott. You can now rent these systems for $1,500 per day.
It only took four more years until much more affordable BCI devices hit the markets, with prices dropping to as low as $50. While most worked decently enough to start selling, only one seemed to be worthy of science. Not for it's signal strength necessarily but for it's versatility, design, and low-cost—aspects seldom seen in most scientific equipment today—the Emotiv became an easy favorite.
That aspect has encouraged Siegle to use the Emotiv in other unconventional settings. Every year for the past five years he’s run an event called TREND, or Transdisciplinary Research in Emotion, Neuroscience and Development, an artist-in-residence series at the University of Pittsburgh.
His researchers spend one to four days testing different visual artists with an fMRI scanner, EEG, and eye-tracking devices (ETD), not simply to collect data, but to “exchange perspectives, methods for thinking about emotion and visual representation, and to apply the tools of neuroscience to understanding or advancing the artist’s vision,” says Siegle (he is himself a practicing glass-blower, poet and musician).
Earlier this year, cellist Katinka Kleijn donned an Emotiv in concert so that she could play a duet with her own brain. In 2011, Siegel put an Emotiv on sculptor Richard Claraval in order to observe his changing moods and states in the midst of his vigorous work. “We actually had it on a him as he was effectively using a chain saw,” says Siegle, “and still got data out of it.”
Most of the Emotiv’s developers, part of a small but growing open source “Idea Lab” community trade ideas and code through message boards dedicated to a full spectrum of popular Emotiv-controlled computer games. Developers can choose from nearly 24 different apps and various software developers kits that allow programmers to develop uses for the Emotiv, and, like neuroscientists, they are using the device in ways it was never intended to be used.
The Board of Imagination, says its inventor Will “Whurley” Hurley, was named “exactly how it works–you imagine where you want to go.” Using Bluetooth to link an Emotiv EPOC with a tablet affixed to a motorized skateboard, riders wearing the headset are able to effectively control the speed of the skateboard simply by thinking through neurofeedback. “Will” the skateboard to go faster and you’ll speed up; imagine slowing down and the motors relax.
While there were some different brain-computer interface headsets to choose from, the Emotiv seemed to be the best choice, Whurley says, not only for its simple design and reliable readings, but also for the development software and the growing community behind it. The Emotiv website community forum now boasts over 10,000 posts and 28 published articles by scientists and developers (only a handful of the papers have been peer-reviewed, however).
Biofeedback isn’t easy to master, especially when you’re flying down the pavement at about 20 miles per hour. During initial tests at Chaotic Moon Labs in Austin, Whurley, who works as a developer and general manager there, started to notice some curious challenges that went beyond the obvious ones of occasional electrical interferences, signal delays, and hardware malfunctions.
One challenge in particular was coping with the different ways people think about moving. To slow down or speed up, Hurley advised his testers to focus alternately on objects close by, or on the moving wheels on cars. While users adjust quickly, there’s no easy instruction manual for mind control.
“There’s a big difference between ‘thinking’ of moving and ‘the feeling’ of moving,” Whurley says. A similar project at the University of Minnesota, using a more specialized EEG cap, has recently been exploring the capability of steering a quadcopter drone through balloon hoops in a gymnasium using just thoughts.
The future of BCI technology, Whurley imagines (but really hopes) won’t sit on your head at all. “If you had something that read your brain across the room, without a headset, I’d use it all the time,” Whurley says.
You have probably seen enough dystopian sci-fi movies to realize that mass-producing consumer grade mind-reading machines might not be such a good idea. Developers have already created Emotiv smartphone apps that record your brain data on the go.
Others are dreaming of ways to control Google Glass not with voice commands but with the thoughts “measured” by an Emotiv, which takes us, one blogger pointed out, “one device convergence away from a completely computer-mediated existence!”
You probably won’t see the need these features anytime soon (or possibly ever). But if you end up at a retirement home, assisted living center, or physical rehabilitation clinic, brain-computer interfaces are looking increasingly like an appealing and realistic tool for restoring basic functions.
Indeed, the most dramatic example of BCI so far may be Braingate. The system first made waves in 2006 when researchers at Brown University successfully helped a paraplegic man move a computer mouse that was wired, “telepathically,” directly into his nerve-damaged brain.
By implanting tiny hair-thin electrodes inside the brain, Braingate enables paralyzed people like him to use their brains as remote controls with calibration good enough to, for instance, move a motorized wheelchair or even feed themselves.
Just three years later, the Emotiv was released and, with it, free open-source instructions on how to set-up your EPOC headset to move a motorized wheelchair. Just a few months ago, electrical engineers at the University of California at San Diego devised electronic tattoos as thin as a human hair and as small as a dime that stick right on to your skin. They essentially function as EEG sensors and are meant to eventually replace the current bulky sensors used on BCI devices like the Emotiv.
Not surprisingly at all, cheap EEG headsets are also fueling the growth of neuromarketing, the emerging “science” of better understanding the brain’s subconscious reactions in the hopes of making more effective advertisements. Jake Stauch, a 22 year-old whiz-kid entrepreneur from Charleston, South Carolina, is already ahead of the game.
After dropping out of Duke in 2012 Stauch founded NeuroSpire, a neuromarketing company that has developed a marketing test using the Emotiv, with the aim of providing to its clients (like Visa and Kia) reports and data visualizations about the brains of their target audiences.
Turning the brain into a type of interface, of course, also opens up the not-so-minor and not-so-friendly possibility of surveillance and hacking. Dr. Ivan Martinovic, a computer scientist from the University of Oxford who recently spoke at the Usenix Security symposium in Seattle on the feasibility of “side-channel attacks on brain-computer interfaces,” says that gleaning unspoken information from even a rudimentary brain measurement isn’t as difficult as one might think.
When someone is shown visual stimuli, there’s an involuntary or subconscious response in the brain, one that can be picked up by the Emotiv or even less sensitive BCI headsets. Show someone stimuli with personal significance—a photo of their house, for example, or a child—and a very specific part of the brain may “light-up,” showing what scientists call a “recognition response.” Show someone images that correspond to their ATM pin or a password, and the resulting data might produce precious clues for a hacker, like a surreptitious crime artist eking out the look of a suspect from their brainwaves.
“It’s better than a random guess,” says Dr. Martinovic, who puts the chances that hackers could brute-force their way into your brain, based on lab tests he conducted, at a 60 percent success rate. Today there are obviously much easier methods for gathering your personal data, but future criminals and surveillance experts would be remiss not to consider the potential for a brainwave scan, much as police rely (perhaps not so scientifically) on lie detector tests.
BCI helmets are also making scientists out of amateurs. Matt Otto, a musician and music professor at Kansas University who uses an EPOC, wants to know how to sing with his mind. “I wanted to see if the thing could detect the difference between when I'm actually singing through my instrument from my mind, or when I'm thinking ‘mechanically’ through the instrument,” he says.
As a teacher, Otto’s curious to understand how a student’s mind works when improvising so that he might better be able to instruct students on “proper” improvisation. “Part of the challenge of teaching is to make sure the person is actually doing it the right way. Like they're singing through their instrument as opposed to mathematically forming the ideas with their mind.” His students, he says, “can do a good impression of doing it right and not actually be doing it right.”
Otto—who traces his interest in the Emotiv and the workings of the brain to his meditative practice—has been following the trends in BCI technology since the helmet came out. “At some point it would be great if we had a mind-MIDI device that allowed you to play music from your mind,” he muses.
(On the Emotiv’s webforums, some developers have said they’ve hacked together a method for doing something similar, using a piece of Emotiv software called the 3D Brain Activity Map.)
“When you hear a melody in your head, the system could actually reproduce the melody into MIDI data that you could use to control synthesizers with,” imagines Otto. “If you were physically disabled, you could still be a great musician without having access to your arms.”
Still, mainstream adoptation of BCI may be years away. Siegle speculates that's because most gamers—the device's biggest potential market—still prefer using their fingers over relatively imprecise brainwaves to guide games. “Asking people to use an Emotiv involves adding hundreds-of-dollars in hardware, time to train the hardware, requiring special interface software, and getting your head wet,” he says. “Such requirements may deter some gamers from crying out for EEG as a key feature.”
So far, Siegle points out, the Emotiv's impressive list of partners includes not one major game company. But he highlighted two promising new partnerships: SMI, a German company, has designed a system that combines the Emotiv and eye-tracking software for use by scientists and marketers; and the Austrian developers of a new adventure game, Son of Nor, partnered with Emotiv and recently reached their Kickstarter goal of $150,000.
Like the brain itself, the possibilities of looking into the brain—and using it to guide robots, drive wheelchairs, and make art—are still nascent. But further neuroscience research will help improve the technology—which, in turn, could help advance that research.
“This was created for gamers who wanted to play World of Warcraft and cast spells with their mind,” says Siegle, the depression researcher, about his bizarre headband. “But the freedom it has given us to get out of a traditional lab is extraordinary.”