Over the years, scientists have tried to uncover the mysteries of man’s most complicated organ: the brain. Today’s attempts incorporate modern technology with brain activity to obtain digital re-creations of thoughts, visual experiences or dreams.

The study was led by Jack Gallant, neuroscientists at the University of California, Berkeley, and published on March in the journal Nature. The technique involves brain scanning using functional magnetic resonance imaging (fMRI), which monitors blood flow patterns within the brain and associates them with images shown to the subjects.

It’s not the best of this method so far, but it’s a step forward towards understanding and predicting brain activity. The experiment submitted two of Gallant’s team members, Kendrick Kay and Thomas Naselaris, to a series of 1,750 different pictures. Afterwards, the team of scientists selected 120 pictures the two hadn’t seen before and tried to predict which one they will be looking at by using brain scanning.

The predictions proved accurate in 72 percent of the time with one of the subjects, and 92 percent for the other subject. It’s a new accomplishment in accurately decoding brain activity, but scientists are just at the beginning of the road. The challenge ahead now is to decode brain responses from a whole new range of images, without knowing them first however.

“That is in principle a much harder problem,” said Gallant. You’d need a very good model of the brain, a better measure of brain activity than fMRI, and a better understanding of how the brain processes things like shapes and colours seen in complex everyday images, the report said. “And we don’t really have any of those three things at this time.”

The extension of future experiments could prove useful in all kinds of applications, but there is still a long way to go. For the time being, it is about “reading” visual activity, in the future, could be about reconstructing thoughts and dreams.