杏吧原创

REMOTE measurements of a person鈥檚 brain activity can show what they are thinking 鈥� even when they are not aware of it themselves.

So far, the technique has only been used to identify visual patterns. But it may eventually be possible to probe awareness, focus of attention, memory and even someone鈥檚 intentions. In the meantime, it could help detect whether people who appear to be in a coma are in fact conscious.

In previous studies, scientists have trained monkeys to move robotic arms with the power of thought. Others have managed to recreate scenes moving in front of cats by recording information direct from the animals鈥� neurons (New 杏吧原创, 2 October 1999, p 17). Both these experiments required electrodes to be inserted into creatures鈥� brains though.

鈥淚t might be possible to probe awareness, focus of attention, memory or even someone鈥檚 intentions鈥�

But now Yukiyasu Kamitani at ATR Computational Neuroscience Laboratories in Kyoto, Japan, and Frank Tong at Princeton University in New Jersey have achieved similar 鈥渕ind-reading鈥� feats using functional MRI scanning. The pair showed four volunteers patterns of parallel lines in one of eight orientations. Each orientation produced a different pattern of activity. Then, by focusing on brain regions responsible for visual perception they were able to recognise which orientation the subjects were observing.

What鈥檚 more, when two sets of lines were superimposed, and a subject was asked to focus on one, the researchers could work out which one they were thinking about (Nature Neuroscience, DOI: 10.1038/nn1445).

In a separate study John-Dylan Haynes and Geraint Rees at University College London showed six volunteers two patterns in quick succession. The first appeared for just 15 milliseconds 鈥� too fleeting for the viewer to consciously perceive it. But by viewing fMRI images of the brain, the team could say which image the subjects had seen (DOI: 10.1038/nn1444). The information was lodged in the brain even if the volunteers weren鈥檛 unaware of it.

The study probed a part of the visual cortex that detects but does not perceive visual stimuli. Haynes thinks that farther along the visual pathway, brain regions consciously take note that there has been a stimulus. But for the 鈥渋nvisible鈥� stimulus, this does not happen.

By working out the point at which consciousness kicks in, it ought to be possible to construct what Haynes has dubbed a 鈥渃onsciousness-meter鈥�. He says such a toll would 鈥渁llow us to assess whether a patient is consciously perceiving his or her environment鈥�.

Yang Dan, a neurobiologist at the University of California, Berkeley, agrees this might be possible. But she points out that there is little agreement about what consciousness actually is. And subtler forms of mind-reading, such as working out intentions or beliefs, are even more speculative, she argues.

Even if such information could be gleaned from brain scans, both studies suggest the patterns are unique to individuals. And if anyone is considering the technique as an alternative to the polygraph, they should think again, says Dan. 鈥淭he relationship between brain patterns and lies may be very loose.鈥�