By Sydney Johnson Oct 26, 2017
If Blade Runner had a classroom scene, it might look something like the promotional video by BrainCo, Inc. Students sit at desks wearing electronic headbands that report EEG data back to a teacher’s dashboard, and that information purports to measure students’ attention levels. The video’s narrator explains: “School administrators can use big data analysis to determine when students are better able to concentrate.”
BrainCo just scored $15 million in venture funding from Chinese investors, and has welcomed a prominent Harvard education dean, who will serve as an adviser. The company says it has a working prototype and is in conversations with a Long Island school to pilot the headset.
The headband raises questions from neuroscientists and psychologists, who say little evidence exists to support what the device-and-dashboard combination aims to do. It also raises legal questions, like what BrainCo will do with students’ biometric data.
BrainCo has some big ideas. The company’s CEO has said that BrainCo aims to develop a tool that can translate thoughts directly into text, or “brain typing.” To support that work, the company plans to use data collected from students using its headsets to compile “the world’s largest brainwave database.”
Theodore Zanto, a professor of neurology at the University of California at San Francisco, had two words when he first read through the company’s website: “Holy shit.”
The brains behind BrainCo
The founder and CEO of BrainCo is Bicheng Han, a PhD candidate at the Center for Brain Science at Harvard University. In 2015, his Somerville, Mass.-based startup was incubated in the Harvard Innovation Lab, and last year the company received $5.5 million in seed funding in a round led by the Boston Angel Club, with participation from Han Tan Capital and Wandai Capital, to develop BrainCo’s first product: Focus 1.
Focus 1 is a headband that aims to detect and report brain activity through EEG, or electroencephalography, which measures in the brain. To advertise the device to schools, BrainCo packages the headset as Focus EDU, which essentially is the headset plus a dashboard where teachers can view all of their students’ EEG data. According to the video, a high numerical score for the EEG signal suggests that a student is paying attention; a low score is interpreted as a distracted or unfocused student.
Max Newlon, a research scientist at BrainCo, adds the company is also studying if the headset could help students and families “train their brain” to improve attention skills.
BrainCo is hardly the first company to sell so-called “brain-training”—or even EEG headsets. Similar devices include Muse, a “personal meditation” headband intended to guide relaxation based on real-time EEG readings. There are also Neurocore Brain Performance Centers, clinics that “empower you to train your brain” also using EEG readings. (Education Secretary Betsy DeVos is among Neurocore’s investors.)
Focus EDU, by contrast, is among the first EEG products that will be marketed directly to teachers and schools.
“We are trying to be the first company to quantify this invisible metric” of student engagement, says Newlon. “Teachers have an innate ability to know when their students are engaged, but we want to give them a superpower so they can track and quantify that over time.”
The idea was enough for BrainCo to win awards including “Most Innovative” at a pitchfest during the 2017 International Society for Technology in Education (ISTE) national conference.
But the company has also faced less enthusiastic reviews. At the 2016 CES conference, an electronics and consumer tech tradeshow, BrainCo’s Focus 1 device flopped in a live demo, which attempted to use human brainwaves detected by the headband to control a robotic hand. The Daily Dot called it the most “cringeworthy demonstration” at the event. “That’s a mishap that calls into question the overall function of the device,” the reporter wrote. “Was it ever actually reading the brainwaves at all?”
When BrainCo returned to CES in 2017, the company arrived with an even bigger robot—which the site WearableZone reported was a success—along with a strategic “pivot” towards education.
More recently, BrainCo has chalked up some big wins: It signed education superstar, James Ryan, Harvard’s dean of education, as an adviser. And now it’s closed a $15 million Series A funding round, bringing the the company’s total funding to nearly $20.5 million. The funding was led by Chinese investors Decent Capital and the China Electronics Corporation, which on its website describes itself as “one of the key state-owned conglomerates directly under the administration of central government, and the largest state-owned IT company in China.”
Making it ‘official’
Increasing engagement in class isn’t the only way BrainCo plans to sell its product. According to Newlon, the startup hopes to secure approval from the US Food & Drug Administration to use the headset for ADHD therapy.
BrainCo aims to rely on neurofeedback (using EEG metrics) to do “brain-training.” According to UCSF’s Theodore Zanto, brain-training involves repeating a specific task, such as playing a videogame, until the individual is considered focused. Advocates contend that improving focus on a specific task can help individuals better their ability to pay attention to other activities.
Newlon says BrainCo is also working on a smartphone game that will rely on EEG measurements from the headset to help people improve focus. Students could play the game for 20- to 30-minute sessions and earn points when they are in a focused state.
Companies that claim to offer similar brain-training options, like Neurocore, frequently cost more than $250 just for an evaluation. BrianCo wants to offer a cheaper, take-home version. “We are hoping to provide an alternative to lower the cost to make it happen in home,” says Newlon.
Several scientists have concerns about at-home “brain-training,” however, as well as about asking teachers to use EEG to inform pedagogical practices.
Sandra Loo, professor of psychiatry at the University of California, Los Angeles, says EEG technology is not sophisticated enough to accurately factor in important variables such as neurodiversity. That, she says, can yield misleading assumptions about how an individual student is doing or what he or she needs.
“Even in resting EEG, there are different subgroups [of brain activity]. It’s not just ADHD; there is variability in normal kids,” says Loo, who also directs pediatric neuropsychology at UCLA. “No one is paying attention to that, [instead] the approach is one-size-fits-all—’Do this and it will work for your child.’”
Zanto, who directs UCSF’s neuroscience program, doesn’t believe there is any physical danger in wearing the headset and administering EEG without a doctor present. What does concern him are errors in recording EEG signals—which are sensitive to muscle movement—as well as incorrect interpretations of the data.
“I haven’t seen any data indicating you can dissociate [in an EEG scan] if someone is paying attention to the teacher or their phone or just their own internal thoughts and daydreaming,” says Zanto. Students wearing the headset “might be incredibly focused, but focused on the wrong thing, and you could get the same EEG measures.”
Even if the EEG is accurate, Loo worries that teachers without medical training might not know how to analyze the data—something that could lead to improper uses of it. “Mental-health professionals undergo a lot of training to make an accurate diagnosis,” says Loo. “I wouldn’t go in and try to do what the teachers are doing without their training.”
When EdSurge asked to see some of the scientific findings that BrainCo was using as evidence of their efforts, the company pointed to a meta-analysis(a study of other studies) suggesting neurofeedback can be an effective treatment for ADHD.
To be sure, Zanto, Loo and other EEG experts who spoke with EdSurge say there is scientific evidence that correlates EEG activity with attention. But using EEG data to parse what a student is actually paying attention to or using neurofeedback to improve attention is more difficult to confirm. What’s more, there’s scant evidence that neurofeedback can be used to treat learning disabilities, say scientists.
“The scientific evidence doesn’t support neurofeedback as a treatment for ADHD,” says Loo. “I have done research on EEG and ADHD for my whole career, and we just don’t know yet.”
“I haven’t seen anything indicating neurofeedback of this sort can improve outcomes [for people with] ADHD,” echoes Zanto. “It’s not out of the realm of possibility, but that just means more research needs to be done.”
Marketing claims related to neurofeedback products have attracted regulatory concern before: Last year, the Federal Trade Commission leveled a $2-million fine on Lumosity for “deceptive advertising” about the effectiveness of its brain-training games.
Researchers at BrainCo seem to be at least aware of the research gap. Included in the email with the meta-analysis, BrainCo shared a WebMD article noting “that we still need more research.”
Big data plans
BrainCo’s goal of treating ADHD—and helping educators in general—is related to another interest for company, as stated by its CEO in an interview that ran on the Forbes website in May. In the Q&A, Han said that the company wants to use artificial intelligence to create a brain typing tool—something that turns thoughts into text. Data collected from students using the devices would be key to informing the development of that technology.
BrainCo says that a company in China has ordered 20,000 devices already.
“Our goal with the first 20,000 devices, each of which will be used by multiple students in schools, is to capture data from 1.2 million people,” Han said in the interview. “This will enable us to use artificial intelligence on what will be the world’s largest database to improve our algorithms for things like attention and emotion detection.”
But BrainCo has not yet established any policies that guide (or prevent) the company from using data collected from U.S. students. The company intends to “use [headset] data for a number of different things,” according to Newlon.
“The hope is when we have this really big database, we will do some analysis that others haven’t been able to do yet,” Newlon adds. “If we have a big database, we might do more interesting research along the lines of something like brain typing.”
The headband’s potential for ADHD therapy isn’t currently mentioned on BrainCo’s website. To date, the company hasn’t yet gone through a clinical trial to approve the product for medical purposes.
But a lack of transparency on the company website about data privacy policies could trigger other troubles.
For instance, Bill Fitzgerald, director of privacy review programs at Common Sense Media, points out that BrainCo does not have any publicly-available privacy statement, ethical review or any information about how students’ data will be stored or used. If a parent wishes to have their child’s information deleted, BrainCo doesn’t offer information on how that happens or how it will prove to families that the biometric information is indeed removed. There is also no public-facing process on how families will be notified if their child’s biometric data is breached.
In many cases, safeguarding student privacy is required by law. Under the Children’s Online Privacy Protection Act (COPPA), Emily Tabatabai, a privacy lawyer, says “a provider needs to have procedures in place to address the collection of data, the deletion of data, how it will be used and shared.”
Newlon says that BrainCo’s privacy policies are still being “ironed out.” He stresses that the company has not yet started collecting data—though it plans to do so.
While FERPA, the federal student privacy law, doesn’t mandate that privacy policies are posted on a company’s website, some state laws do. Tabatabai says “at the very least, the company must have data protections in place and describe its privacy policies to the school.” The lawyer also underscores that determining which privacy laws may apply to any company is dependent upon factors such as whether or not information is collected online, or if a student’s educational records are being shared with the company.
BrainCo declined to name the Long Island school that it says is considering to pilot its device. When EdSurge began reporting for this story, BrainCo claimed that the pilot was slated to begin on Oct. 31. Since then, the company has backed off that timeline and says the trial will be delayed.
Newlon wrote in an email: “We will not be running any pilot study without the proper data privacy procedures in place.”
Newlon also points out that the company is an Institutional Review Board (IRB) organization, which gives it permission “to review and monitor biomedical research involving human subjects,” according to FDA rules. But an IRB doesn’t make BrainCo immune to data breaches, or malpractice. “If they are looking at the details of a pilot and they don’t have these [privacy] questions answered, they should stop every single thing they are doing until they do,” says Fitzgerald.
Loo, the psychologist at UCLA, adds that by using medical records—namely, EEG recordings made in a classroom—may also mean that BrainCo will have to comply with the Health Insurance Portability and Accountability Act, or HIPAA, which lays out privacy regulations for medical data.
“Any time you have medical biological materials, a lot of safety measures come into play,” says Loo. “Is there confidentiality there? Can the company release [data] to insurance companies? Who knows.”
Ryan, the company’s adviser and Harvard’s dean of education, is aware that BrainCo wants to pilot Focus EDU on students. But as for any data and privacy protections underway at the company, he says “I don’t know, I haven’t had a conversation with Max [Newlon] about it.”
Meanwhile, BrainCo’s CEO did not respond when asked in an email about the lack of privacy regulations in place. And when EdSurge asked to try out one of the Focus EDU headsets, BrainCo also declined.
The wild ambition of selling brain-reading headbands to schools—which are charged with protecting students—seems to raise a range of concerns, including deeply philosophical ones. How far should schools go to “train” students?