With the help of tomography and computer analysis, Japanese scientists were able to almost correctly - in 60% of cases - determine the visual images that arise in a person in dreams. This is stated in an article published in the Indian journal Science Express.
A team of employees at the Institute of Science and Technology in Nara city tried to find a connection between visual images in sleep and functional magnetic resonance imaging of the brain of sleeping people.
The level of detail is far from what we saw in Leonardo DiCaprio's Inception, where people are manipulated by industrial espionage professionals using special techniques to "infiltrate" dreams.
Nevertheless, experts describe the results of the work as "amazing", according to the British newspaper The Daily Mail, which devoted an enthusiastic article to the experiment.
"This is probably the first real visual demonstration of brain function based on the content of dreams," - told reporters a neurologist and specialist in the study of sleep mechanisms Dr. Robert Stickgold of Harvard Medical School in Boston. Although, he noted, there is still a long way to go before the creation of a machine that can fully read our dreams, in an even more distant future on the basis of such technology it will be possible to create a program that “reads” the thoughts of people when they are awake.
200 times in 20 categories
In the first stage of the experiment, called the Dream Catcher, three volunteers slept until an encephalogram showed REM sleep (rapid eye movements). At this point, the participants in the experiment were woken up and asked to tell what they dreamed about. The process is repeated until each participant has recounted at least 200 dreams.
Promotional video:
The descriptions were analyzed and categorized into twenty categories according to key themes (men, women, tools, books, cars, etc.) for each participant.
The scientists then compared the distribution and dynamics of brain activity zones, obtained using functional magnetic resonance imaging, with a set of keywords in the subjects' stories. As a result, using the methods of self-learning computer programs, scientists have created a system capable of decoding the visual series of dreams based on tomography data.
The database compiled on this basis makes it possible to determine with an accuracy of 60-70% what image a person sees at a given moment in a dream. This is much higher than a simple coincidence.
The researchers believe that the data they obtained will allow to isolate the features of brain activity "linked" with visual images, not only during sleep, but also during wakefulness. In the future, they intend to try to decipher dreams during other phases of sleep.