电脑能读懂你的心思?

电脑能懂读心术？能解码你的想法？这个曾只局限科幻小说中的神秘武器现在已真真切切地存在于我们生活当中。《当代生物学》杂志的某份研究报告表示：一项新研发的计算机程序在对脑部进行扫描及分析后，能够真正读懂人们的内心所想。

Mind reading machines are no longer just science fiction. For several years, scientists have been experimenting with devices that can detect patterns of brain activity associated with simple things like a carrot or a hammer . And now a team in the U.K. reports success in decoding much more complicated thoughts.

NPR's Jon Hamilton has the story.

JON HAMILTON: Our brains devote a lot of space to our own autobiographies and Eleanor Maguire of University College London says autobiographies are built on memories of important events.

Professor ELEANOR MAGUIRE (Cognitive Neuroscience, University College London): You know, when you got married or when you first went to college or when this or that happened. So these are very much the events that shape our lives.

HAMILTON: Maguire was part of the team that wanted to know whether traces of these so-called episodic memories could be detected in brain scans. So they found 10 volunteers who agreed to watch several very short films. The films served as proxies for real-life experiences.

Prof. MAGUIRE: In one case, a woman came on the screen, rummaged in her handbag, took out a letter and then posted the letter in the mailbox. In another one, a woman came on screen, finished a cup of coffee and then put the disposable cup in the trash can.

HAMILTON: Maguire says the volunteers watched the films over and over until they had formed a strong memory of each episode .

Prof. MAGUIRE: And then we popped the people in the scanner and then had them recall these movies.

HAMILTON: The scans reported activity in a part of the brain called the hippocampus , which seems to help organize episodic memories. Then the team used a computer program to search for patterns of activity associated with each film, and Maguire says it found them.

Prof. MAGUIRE: In every single case, it was able to predict with high accuracy which of the memories those 10 participants were recalling.

HAMILTON: Each person's brain had a unique pattern associated with each film clip. That's what the scientists expected to find. But Maguire says a comparison of all 10 brains showed something else that was really surprising. The pattern associated with, say, the woman mailing a letter didn't vary much from brain to brain.

Prof. MAGUIRE: It was really similar, the parts of the hippocampus that were involved in being able to discriminate between these memories. So it was incredibly consistent across individuals.

HAMILTON: Other scientists say that seems to be true for lots of thoughts.

Dr. MARCEL JUST (Director, Center for Cognitive Brain Imaging, Carnegie Mellon University): We all have very similar patterns for a given concept.

HAMILTON: Marcel Just directs the Center for Cognitive Brain Imaging at Carnegie Mellon University in Pittsburgh. Just says it's pretty easy to put someone in a scanner and tell whether they are thinking about a face, for example. But he says that by studying a particular brain in particular situations, it's also possible to get much more specific information.

Dr. JUST: We asked people to think about a face of various members of their family and to some extent we can decode which one they're thinking about.

HAMILTON: Just says mind reading experiments like these rely on the cooperation of the person being scanned. They have to think about something over and over. So far there's no way to use a brain scanner to forcibly search someone's memories. But Just says the ability to detect what a person is thinking is progressing with remarkable speed.

Dr. JUST: At the extreme, maybe we could decode somebody's dream while they're dreaming. Imagine that. Having somebody in the scanner go to sleep, and they're dreaming, and you can say, okay, ice cream cones, you know, white horses, whatever. Is that possible? Not this year. Not next year. But I think that's doable.

HAMILTON: And Just says that is likely to touch off a discussion about who will get to see the thoughts in our brains. The new research appears in the journal Current Biology.

Jon Hamilton, NPR News.