'Mind-Reading' Computer Debuts

Researchers at Carnegie Mellon University in Pittsburgh have developed a 'mind-reading' computer.

It is hoped the mind-reading machine, which can forecast the activity patterns a brain will create for a specific word, will offer a better understanding of how and where the brain stores information and even lead to improved treatments for language disorders and learning disabilities.

Researchers used nine volunteers to train the computer. They were given 58 words and asked to think about the meaning and properties of the words. Brain scans taken when the users were thinking about the different words were then captured using magnetic resonance imaging, which identifies real-time brain activity.

"If I show you the brain images for two words, the main thing you notice is that they look pretty much alike. If you look at them for a while you might see subtle differences," explains Tom Mitchell of the Machine Learning Department, which lead the study.

"We believe we have identified a number of the basic building blocks that the brain uses to represent meaning. These building blocks could be used to predict patterns for any concrete noun," added Mitchell.

The researchers then gave the computer two new words and images and asked it to pair them up correctly, which it successfully completed.

Mitchell said the next step would be to study brain activity for phrases rather than singular words.

"If I say 'rabbit' or 'fast rabbit' or 'cuddly rabbit', those are very different ideas," Mitchell said. "I want to basically use that as a kind of scaffolding for studying language processing in the brain."

Mitchell also highlighted a few problems with the research.

"It can be hard to focus. Somewhere in the middle of that their stomach growls. And all of sudden they think, 'I'm hungry - oops.' It's not a controllable experiment."