What dog lover hasn’t looked at his or her trusty mutt and tried to figure out what they’re thinking? Researchers from National Autonomous University of Mexico in Mexico City have gotten a bit further toward answering that question than most. Or, at the very least, the team appears to have gained some insight into why dogs seem to understand us so well.
In a recently published research paper, the team demonstrates that it’s possible to work out what a dog is looking at by analyzing a scan of its brain. Using functional magnetic resonance imaging (fMRI) technology, the researchers scanned the brains of four border collies, trained to sit on a scanner without moving. The dogs were shown happy, sad, angry and fearful facial expressions — all made by people they were unfamiliar with — and had their brain patterns recorded.
Analyzing these patterns using machine-learning algorithms allowed the researchers to figure out which face the dog had seen. The most distinctive brain pattern was associated with a happy face, which triggered a particular activity in the temporal cortex of the brain, used for processing complex visual information.
Interestingly, the study mirrors findings from a more human-centric experiment staged earlier this year. In that case, researchers in Japan were able to get an A.I. to caption images describing what a person had seen, based solely on an fMRI brain scan image. Accurate captions included the likes of, “A dog is sitting on the floor in front of an open door” or “a group of people standing on the beach.”
While the researchers in this latest experiment limited their study to only a few emotional states to identify, it certainly suggests an impressively high level of human emotional recognition on the part of our dogs. Now if only someone could figure out how to convert this into some kind of portable brain-reading tech so we can see more information about how dogs interpret the world around them on a daily basis!
A paper describing the work, titled “Decoding Human Emotional Faces in the Dog’s Brain,” is available to read online on scientific paper repository bioRxiv.
