Scientists have learnt to "read" human thoughts: how they did it

Remote mind reading is considered to be absolutely fantastic and unattainable. However, it seems that scientists have come close to achieving this.

According to Live Scene, a recent study has managed to establish quite precisely what kind of information passed through the brains of its participants. The description of the experiment was published at the end of September, it has not yet been peer-reviewed, but it has already aroused great scientific interest.

The researchers used the functional magnetic resonance imaging (fMRI) method to "read minds". It does not involve implanting electrodes into the subject's brain, as was done in earlier studies. A person is simply placed in a CT scanner and the flow of oxygenated blood through the brain is monitored. Since active brain cells require more energy and oxygen, this indirectly measures the activity of the organ.

It is important to note that the activity of the brain is not recorded directly - the electrical signals emitted by its cells move much faster than the blood flow. But even such imperfect indirect evidence allows us to decode the semantic meaning of people's thoughts. Of course, this is not a literal "translation" of their thoughts, but the results can be used to determine in some detail what the subject was thinking.

One woman and two men aged between 20 and 30 took part in the trial. Each participant listened to a total of 16 hours of different podcasts and radio programmes in several sessions in the scanner. The team then uploaded these scans to a computer, where a decoder algorithm compared the patterns in the audio to those in the recorded brain activity.

This algorithm can create stories based on the data, and the scientists claim that the conclusions it draws from it match the original story of the podcast or radio programme "pretty well". In other words, the decoder could deduce what story each participant was listening to based on the person's brain activity.

Mistakes did occur. For example, the algorithm got confused about pronouns or sometimes swapped the first person with the third person. Nevertheless, its conclusions were quite accurate.

In additional tests, the algorithm could quite accurately explain the plot of a silent film that participants watched while lying in the CT scanner. It could even retell the story that participants imagined in their heads. In the long term, the research team aims to develop this technology so that it can be used in brain-computer interfaces designed for people who cannot speak or type.

As OBOZREVATEL previously reported, scientists have identified 1,000 genes responsible for differences in brain function between men and women.

/Life/Scientists have learnt...