Conor Russomanno hadn’t stopped wondering about the effects of the multiple concussions he’d suffered playing football and rugby at Columbia University. In 2011, less than a year after his last severe hit, he had passed a neurologist’s standardized test of cognition, but he still wasn’t himself, at least not all the time. “My mind [was] definitely different than it was before,” he says. “I was really, really amped and self-motivated and confident on certain days, and then I would hit these extreme lows on other days.”
The following year, as Russomanno was pursuing a master’s degree in design in New York City, a friend offered to sell him a MindFlex—a cutting-edge toy from Mattel that allowed users to make a ball hovering...
He deconstructed the toy and fitted it to a baseball cap with its single electroencephalogram (EEG) electrode resting on his forehead. He added a Bluetooth transmitter on the brim that allowed the device to communicate with his smartphone and built an app into which he could enter his activities and moods. Then Russomanno wore the hat one day while walking around New York City. “I wanted to start seeing if I [could] quantitatively track my habits and routines and stitch them to my emotions and internal states,” he explains.
The consumer market is creating these devices that didn’t exist . . . devices that open doors for researchers.—Olav Krigolson, University of Victoria
Russomanno’s device was just a prototype, and one day’s worth of data was not enough to answer the questions he had about his own mind. But his tinkering opened his eyes to the potential of BCI and the technology it’s based on. Around that time, his grandmother was diagnosed with a rare form of dementia, and a close friend suffered a temporarily paralyzing neck injury. Russomanno realized that portable EEG devices that feed information about brain activity into a computer, if done well, could have all sorts of applications, including monitoring brain health and assisting people with disabilities; the technology could “be applied in so many ways beyond my personal curiosity.”
There were already a handful of mobile EEG devices on the market and several more on the way, driven partly by neuroscientists’ desire to make the technology more practical for use in healthcare and other settings, and partly by the tech industry’s interest in developing EEG-based consumer BCI applications. Accordingly, companies selling the devices typically fell into one of two groups: those targeting researchers, with high-quality but extremely expensive equipment, and those targeting tech developers with low-cost hardware, usually with only a few electrodes and sometimes with fees to access the raw data. Russomanno saw an opportunity to democratize a technology that, at the time, was largely limited to big companies and well-funded labs.
In December 2013, he and Joel Murphy, his physical computing professor at Parsons School of Design, launched a Kickstarter campaign that promised to deliver an open-source, reasonably priced BCI device that could amplify and convert analog EEG signals into digital data to be streamed wirelessly to a computer. In less than two months, 947 backers pledged more than $215,000. In early 2014, Russomanno and Murphy founded a company, OpenBCI; they started shipping their first product—a small box that amplified the analog brain signals from up to eight EEG electrodes and sent the translated digital data to a computer—by the end of the year.
The timing couldn’t have been better. The consumer BCI industry was just beginning to blossom. In September 2014, Toronto-based InteraXon launched Muse, one of the first EEG-based devices truly targeted at consumers: a headband with four electrodes that communicated with smartphone or computer apps, designed to improve mindfulness and meditation by giving users auditory feedback on their cognitive state. A few years later, Paris and San Francisco–based Rythm (now called Dreem) launched the Dreem headband with six electrodes and apps to help consumers sleep. Since then, several new electrode headsets—and the hardware and software needed to process their recorded neural activity—have come on the market.
From health and wellness to gaming and virtual reality, the BCI market—which Brandessence Market Research valued in 2019 at $980 million and predicted would double in value in the next 15 years—is mainly driven by consumer demand. But the BCI industry also has an eye toward research, and with neuroscientists striving to make EEG mobile, consumer BCI wearables may be just what the field needs.
“The consumer market is creating these devices that didn’t exist . . . devices that open doors for researchers,” says Olav Krigolson, a University of Victoria neuroscientist. “It’s a new emerging technology and researchers are slowly figuring out the things you can do with it.”
Neuroscientists seek mobile EEG technology
Although EEG took a back seat in neuroscience with the advent of magnetic resonance imaging (MRI) in the 1970s and ’80s, the technology has been making a comeback in recent years, says Krigolson, thanks in no small part to the fact that it can be taken on the go. The first “mobile” EEG setups involved packing traditional equipment into backpacks, an approach that was cumbersome and produced noisy data. In the late 2000s and early 2010s, at least half a dozen companies sprang up to offer more-practical setups. These products were expensive, with price points in the thousands or tens of thousands of dollars. The equipment was designed with researchers, not consumers, in mind.
But soon, the first low-cost devices began to hit the market. In 2009, San Francisco–based EMOTIV launched its first headset, which had 14 electrodes—still far fewer than the 32 or 64 of a traditional EEG cap—and cost researchers just $750. That same year, NeuroSky released MindSet, a pair of consumer-targeted headphones with an arm that positioned a single electrode on the forehead—for $199. (MindFlex and another EEG-based toy—Uncle Milton’s Force Trainer, which similarly allowed users to control a ball by concentrating while hearing instructions from Yoda—were also released in 2009 and used chips sold by NeuroSky.) And in 2014, InteraXon launched Muse. With four electrodes and a price of $150, it became the first consumer product to make real inroads into research.
Researchers were initially skeptical about the quality of the EEG data that Muse and other low-cost devices could yield. Krigolson, for example, worried that the noise generated from muscle movements and other sources would wash out any signal that the device recorded from the brain. “Picking up muscle activity is very easy; it’s orders of magnitude bigger” than the neural signal, he notes. But when Krigolson and his colleagues, who have no financial ties to InteraXon, put it to the test a few years ago, they found that Muse did yield data of sufficient quality to detect event-related potentials (ERPs)—patterns of neural activity in response to a stimulus.
Enthused, Krigolson and his colleagues designed a protocol to detect cognitive fatigue and successfully tested it first on hospital workers and miners, and then on themselves during a week-long expedition to the Mars Habitat at the Hawaii Space Exploration Analog and Simulation. Other researchers also started using Muse; on its website, InteraXon lists nearly 200 publications on diverse topics including pain and post-traumatic stress disorder (PTSD) that have used data collected by the headband. “People have been using Muse for research for a long time now, as a recording device, as a therapeutic device, and everything in between,” says Subash Padmanaban, a research engineer at the company.
Still, the data from Muse and other mobile devices are not as pristine as data retrieved from research-grade equipment, notes Anthony Ries, a research psychologist at the US Army Research Laboratory who was part of a team that compared research-targeted mobile EEG technologies with conventional EEG systems. “Generally, there is a tradeoff between data quality and mobility,” he tells The Scientist by email. Thea Radüntz of the Federal Institute for Occupational Safety and Health in Berlin, Germany, agrees. She has compared devices from EMOTIV, NeuroSky, and others, and found that EEG setups with fewer electrodes often produce messier data and limit researchers’ ability to triangulate the source of the electrical activity being recorded.
But for many applications, just one or a few electrodes may be sufficient. Strong, brain-wide responses, for example, are “pretty easy to record with one of these mobile systems,” says Scott Burwell, a neuroscientist currently wrapping up a postdoc at the University of Minnesota who has used Muse and an EMOTIV headset for his research. “And in the clinical setting especially, people are just interested in registering that brain response, and you can do that with relatively few channels.”
New BCI products invite developers and researchers to tinker
The consumer BCI industry continues to grow, with established companies launching upgraded and novel products and a handful of new competitors entering the market with low-priced devices. While the BCI-enabling products primarily focus on consumer applications, these companies are also paying mind to research applications of the technology.
In 2019, Neurosity launched Notion, a headset with eight electrodes that sells for $899, with raw data freely available to users. Like Muse, Notion comes with neurofeedback software, but instead of encouraging a meditative state, it’s used to maximize focus and productivity, with the primary market so far being software coders, says Neurosity cofounder Alex Castillo. Notion has an on-board computer so that the data can be stored and later retrieved directly, and thus does not have to be streamed, something Castillo argues is critical for data fidelity and security. The device also comes preloaded with an app that can learn to interpret the user’s imagined motor commands, for applications in gaming and artificial reality/virtual reality (AR/VR).
Castillo says that Notion’s potential in health and wellness applications is never far from his mind. His brother has epilepsy, and at the end of 2019, researchers published a study that suggested it may be possible to predict an epileptic seizure up to one hour before it starts using EEG data and machine learning. “Maybe people who suffer from epilepsy can use this technology and predict this is going to happen and maybe not go for a drive,” he speculates. “Our device, it is for consumers, but it is clinical grade. We made it so you can use it for research, because we’re also looking to the future and this new generation of neuroscientists [who] are going to use it to develop their own research.”
For many applications, just one or a few electrodes may be sufficient.
This July, another new competitor, NextMind, began presales of its $399 headset. This device also uses eight electrodes but specifically targets the visual cortex, located at the back of the brain. The company has developed software it calls the NextMind engine, which ascertains visual attention and translates that into digital commands, allowing developers, particularly in the gaming and AR/VR industries, to get creative. “We are basically offering the platform to the industry such that the industry will be able to build on top of it,” says NextMind founder Sid Kouider. NextMind does not currently offer access to raw data, but Kouider says the team would consider doing so in the future. “By addressing this broader market we expand the potential of neuro-technology,” he writes in an email, adding that it “should help us to reach a scale never achieved so far” and thereby benefit research.
While these latest devices have yet to undergo stringent tests by independent labs, they’re already being adopted by researchers. Kouider says that most buyers so far are developers, but some are academic researchers or consumers. And the University of Minnesota’s Burwell was all too happy to try out Notion after Neurosity cofounder and CEO AJ Keller, formerly of OpenBCI, reached out to him.
Burwell is launching a company based on software he’s developing to detect neural biomarkers of drug craving, and had been using Muse to monitor patients’ brain activity. But now he says he’s leaning toward taking his work forward with Notion. He says Muse has done well for him, but it has some limitations—if wearers move their heads or scrunch their foreheads, for example, the electrical activity of the facial muscles can “swamp out the EEG signals.” Notion sits on the crown of the head, where there are fewer muscles to interfere. Burwell is now on the list to receive the Notion 2 headset, which began shipping this summer, and Neurosity has written support letters for Burwell’s grant applications for research using their equipment.
“Up until maybe five years ago, this stuff wasn’t ready for prime time yet,” says Burwell. Even now, many neuroscientists remain cautious, largely sticking to EEG setups designed for researchers. But consumer-oriented technology is starting to prove itself, and Burwell says he expects that as it continues to improve, the research community will come to accept at least some of these devices as valid research tools. “We’re kind of at a really critical period here.”
Consumer BCI Products
In addition to several companies that offer mobile electroencephalography (EEG) equipment specifically designed for researchers, a number are developing EEG-based brain-computer interface (BCI) devices for the consumer market. Here are a few:
Year company launched
Access to raw data
San Jose, California
MindWave Mobile 2 headband with a single electrode
Gaming and development
New York City
Various BCI components sold a la carte or as kits for creating simple wearables
Broadly accessible BCI technology
Ultracortex headset sells for $350
Muse 2 and Muse S headbands with four electrodes, plus sensors to measure heart rate, movement, and breathing
Meditation and mindfulness
$4/month (but can be hacked for free)
Dreem (formerly Rythm)
Paris, New York City, and Taiwan
Dreem 2 headband with six electrodes, plus sensors to measure
Free with research version
New York City
Notion 2 headset with eight electrodes, plus a sensor to measure movement and breathing and two haptic motors to generate vibrational feedback
Decode motor intention; productivity, flow for software developers
Free; processed on device for security
NextMind headset with eight electrodes
Decode visual attentional focus; gaming, AR/VR