Scientists Develop Mind-Reading Brain Scans

Researchers have demonstrated that brain scans can reveal peoples' intentions.

Researchers at the Max Planck Institute for Human Cognitive and Brain Sciences' Bernstein Center for Computational Neuroscience in Berlin recently demonstrated that algorithms, coupled with an MRI (magnetic resonance imaging), can determine whether study participants planned to add or subtract groups of numbers.

"It has never before been possible to read out of brain activity how a person has decided to act in the future," a statement from the center explained.

The participants chose one of the two math functions and held the thought in their minds while getting an MRI. The images revealed fine-grained patterns of activity and computers read their covert intentions, before participants received sets of numbers and performed the math problems. The computers, programmed to recognize patterns that commonly occur with specific thoughts, determined participants' intentions 70% of the time. The computers use a method called multivariate pattern recognition to pick up brain activity across extended regions to determine what a person has decided to do.

Dr. John-Dylan Haynes, who has done similar research in the past, said during a recent interview that the findings could help paralyzed patients and criminal investigators someday. Technology already exists to help paralyzed patients move computer cursors a step at a time, but the recent findings indicate potential for great leaps ahead, he said.

"With the brain-scanning current technology, you can read out something like left or right," Haynes said. "So, you can control a cursor on a screen of letters. It's powerful and fast, but very restricted and very tedious to click your way through some string of letters in order to spell a word or a sentence. In the future, it will be possible to read even abstract thoughts and intentions out of patients' brains. One day, even the intention to "open the blue folder" or "reply to the e-mail" could be picked up by brain scans and turned into the appropriate action."

Haynes said that previous studies have shown areas of the brain involved in emotional reactions like jealousy and romantic love, but they didn't illuminate the details.

"It's like they point toward a book -- the book that has written in it all of your romantic love partners or the book with religious experiences in it -- but they can't look inside the book. They don't open the book. The thing is, we can really read them. It's like we shined a torch-light in the brain."

Haynes said it only took a month or two to program the computers and tweak old algorithms. Since researchers have demonstrated the ability to distinguish between two possible thoughts, it shouldn't be difficult to determine a "yes" or "no" answer to whether someone is lying or not, according to Haynes.

"That's easier to read out than if we have infinite alternatives," he said.

Haynes said that the technology could improve the accuracy of the legal system, but it also presents ethical dilemmas.

"We might want to protect someone's mental privacy and the ability to control what leaves their brains," he said.

Neuromarketing, pushing people's buttons quite literally, and measuring their attitudes about different things, could pose serious threats to mental privacy, Haynes said. He suggested that citizens, scientists, and policy makers begin thinking about when the use of the technology would be justified.

In the meantime, other methods, like voice analysis, also are under development and could be used for similar purposes.