Eden.fm

http://www.dailytech.com/article.aspx?newsid=16358

.. A previous study had looked at using fMRI scans of parts of the brain linked to shape identification to correctly guess the a viewed image from a series of stock images.  Jack Gallant, a University of California, Berkeley neuroscientist who led the current effort, describes this previous work as similar to “the magician’s card trick where you pick a card from a deck, and he guesses which card you picked. The magician knows all the cards you could have seen.”

The new study expands this approach greatly by also scanning parts of the brain used for general classifications like “person”, “car”, or “building”.  Utilizing Bayesian probabilistic math, researchers armed with a database of over 6 million possible results and the new scans were able to go beyond identification into the realm of reconstruction, coming up with an image corresponding to what the person was thinking of, after an initial calibration to adjust for mental differences.