n the movie “Inception,” Leonardo DiCaprio’s character interferes with others’ dreams to implant new ideas in their minds.

It’s a trippy plot premise. It’s also not entirely science fiction.

In the last few years, researchers have developed a way to “incept” in real life, a technique they hope could eventually help people with disorders from depression to autism. And a new study indicates we could do “inception” without the subject even being aware of what’s being learned.


The basic idea is this: Human subjects lie down in a functional magnetic resonance imaging, or fMRI, machine and play a game. During the game, they have their brains scanned, and the game gives them feedback. Afterward, they have developed new brain connections.

This process is called neurofeedback, which uses the data from the fMRI machine to tell someone what is going on in their brain. It’s similar to using your heart rate to set your pace while running (which then changes your heart rate), but instead, you’re controlling your own brain activity.

The most intriguing version, developed by Takeo Watanabe of Brown University and partners at the Advanced Telecommunications Research Institute in Japan, is called Associative Decoded fMRI Neurofeedback. The technique uses brain scanning with an fMRI machine to teach a skill or association — but without the person being aware of what they’re learning.

In Watanabe’s latest work, published in Current Biology on Thursday, subjects learned to see the color red while looking at a picture of black and white stripes — and they had no idea they were being trained to do this.

Fairly innocuous, then, but the idea is somewhat spooky — a kind of subliminal messaging, Watanabe said, “in the simplest form.”

Power of the mind

The idea for neurofeedback technique grew out of research from the 1960s showing that a person could regulate his heart rate or temperature just by thinking about it. Our brains regulate temperature and heart rate. Watanabe wanted to see if we could regulate other aspects of brain activity.

The process of getting incepted takes multiple days. First, the subjects have to have their brains scanned as a baseline so the software knows what kind of brain activity to look for. Then, the subjects undergo three days of serious neurofeedback training. While they’re lying in the brain scanner, the scientist running the experiment tells them to look at the center of the screen. When black vertical stripes come on the screen, they’re told to “try to somehow regulate your brain activity.” After the vertical stripes go away, the subjects are given a score of how well they did. They’re paid more money if they score high.

The subjects aren’t told to think about colors, or try to “see” anything in their minds. They’re told to try to regulate their brain activity, just that. Each person did this exercise over 500 times.

In fact, subjects scored highly when they managed to make their brains light up with “red” brain activity, even though they were looking at black and white vertical stripes. The goal of the neurofeedback training was to strengthen the connection in the subjects’ brains between seeing vertical stripes and seeing the color red.

After the experiment, the researchers asked the subjects what they were thinking about when they got high scores. Their answers were as different as could be.

“I imagined a zebra,” said one participant.

“I imagined a gymnastics match in which I performed well,” “I imagined a situation where I behaved violently,” others reported.

None of them mentioned colors. But the next day, when the researchers showed the subjects vertical stripes, the subjects were more likely to think something was tinted red than people who hadn’t done the neurofeedback training. For neurofeedback subjects, vertical stripes that were in reality greenish looked slightly reddish to them.

Subjects aren’t hallucinating the color red, Watanabe says. They’re experiencing something more akin to synesthesia, a condition in which people perceive colors when they look at printed numbers and letters.

And when the researchers tested the subjects again several months later, they were still biased toward red.

In an earlier study, Watanabe used the same technique to make people get better at seeing specific orientations of stripes flashed on a screen for just a short amount of time. There again, the subjects had no idea what they were learning.

Such a use of fMRI is “very novel and very clever,” said Thomas Naselaris, a psychology professor at the Medical University of South Carolina who was not involved in the study.

A treatment for autism?

Watanabe thinks neurofeedback could eventually be used to treat neurological and psychiatric disorders, such as depression or autism.

The idea would be a similar one. In the case of depression, people would get high scores when their brain activity looked less depressed. For autism, the therapy could theoretically help people deactivate areas associated with autism symptoms.

The latter is something Watanabe is actively studying. Based on a study this spring that looked at differences in the brain connections of normal and autistic people, he has begun testing neurofeedback therapy on people with Asperger’s syndrome. The therapy trains them to enhance the “normal” connections and weaken the autism-related connections in their brains. Right now, however, Watanabe says the therapy isn’t really working.

Dr. Charles Gilbert, a vision researcher at the Rockefeller University said in an email that “to extend it as a technique for autism therapy is quite a stretch.”

Naselaris is more optimistic. “I have no idea if it will work, but on the basis of this result I think it’s worth developing an application for this. This seems potentially very powerful,” he said.

Tricking the brain

Other scientists have used neurofeedback without the subliminal twist, using electroencephalogram to give people feedback about the speed of their brain waves. A person with anxiety, therefore, would be told to use the meter to try to consciously calm themselves, for instance. (Though some of the benefit may be placebo effect.)

But for most conditions, the subliminal kind of neurofeedback would probably be needed. The brain activity related to psychiatric conditions is complex and mysterious. Instructing someone to activate a specific brain connection or turn down a certain collection of brain regions would be confusing.

“It gets a little tricky, because with depression for example there’s a tendency to recall negative memories, and it actually might even worsen symptoms if you try to recall positive memories,” said Steve Ramirez, a researcher at Harvard University who studies depression in mice.

So you can’t just tell someone to “think happy,” you need to trick them out of negative brain states.

But that endeavor is contingent on scientists finding the precise brain signatures for different psychiatric conditions — no small task. Neurofeedback, for the time being, can make you see colors — but it may be some time before it can change your brain’s tint on the world.

Sign up for Morning Rounds

A daily dose of health and medicine news — and a finalist for Digiday’s best email newsletter.

Recommended Stories