Scientists can now “read” visual activity in the brain. MRI technology and computer models make it possible.
Magnetic resonance imaging (MRI) provides detailed images of the brain and all the necessary information related to the viewing of a chosen set of images. Then a computer database creates brain activity-image links, thus storing fresh patterns for further analysis.
Researchers have the right now to claim that “mind-reading" and reconstruction of what people see or think are possible in principle.
The following experiment was probed upon study co-authors, Kendrick N. Kay and Thomas Naselaris.
At first, they were shown 1,750 photographic images of animals, buildings, food, indoor scenes, outdoor scenes and people, while MRI recorded activity in the primary visual cortex region of their brains. All the received information was registered in computer database.
Later, Kay and Naselaris were offered 120 different photos. The computer model sifted through its previous store of brain activity-to-image patterns to "decode" the second data and find a correct match.
The whole procedure of decoding didn’t mean image reconstruction, but rather image identification.
Scientists received the following results: a set of 120 photo implied successful computer identification of the viewed images between 72 percent and 92 percent of the time. If taken 1,000 images, the success rate was 80 percent. 1 billion images resulted in about 20 percent of identification of the time.