New mind-reading technique could help the paralyzed speak

Thanks to a new technique from the Netherlands, those who can't speak or move may soon be able to communicate using only their minds.
Written by Sarah Korones, Contributor

A new “mind-reading” technique from the Netherlands could give a voice to those who have lost both the ability to speak and move.

Developed by scientists at Maastricht University, the process uses functional magnetic resonance imaging (fMRI) to read the minds of those who suffer from Locked-in Syndrome, a condition that occurs when patients are completely paralyzed and unable to communicate but still retain a fully active mind.

In the new study, researchers manipulated brain-scanning technology so that individuals could answer questions by “typing” 27 symbols (26 letters and a space) using only their minds—no muscles or vocal chords necessary.

Here’s how it worked: Participants first underwent one hour of training in which they learned how to answer questions by selecting letters on a computer screen using only their minds.

While lying inside an fMRI machine—a device that measures activity changes in the brain based on changes in blood flow—participants stared at a screen that contained the 26 letters and the space arranged into three rows. Each row was assigned to a different mental task: motor imagery (such as tracing shapes in one’s mind), mental calculation (such as rehearsing multiplication tables) or inner speech (such as silently reciting a poem).

Throughout the experiment, different groups of letters were highlighted at different times. To choose a particular letter, volunteers waited until it was highlighted and then performed the mental task associated with its row for a specific amount of time.

Based on the unique mental task performed and the corresponding length of its performance, the system could associate the movements with each of the 26 letters and the blank space, allowing participants to spell out responses to the researchers’ basic questions.

While the software couldn’t always pick up on the participants’ responses perfectly, it was able to determine the first letter of each response with 82 percent accuracy. When using context-dependent text recognition software (like the kind that makes texting so much easier), the system became even more precise. The software could correctly identify the first letter 95 percent of the time when the second letter was taken into account and 100 percent of the time once the third letter was considered.

The table on the screen isn’t limited to letters, either. Researchers could also fill the screen with pictures of foods and drinks, allowing patients to communicate their dietary preferences to doctors and nurses.

Since none of the volunteers were actually paralyzed, the current study is only a proof of concept. Its results, however, are a tremendous development in spelling devices for those who can’t speak or move.

The results of the experiment are published in the journal Current Biology.

[via Scientific American, Smithsonian]

This post was originally published on Smartplanet.com

Editorial standards