Brain scans could help separate out dogs less suited to service work.

A study by researchers at the University of Emory tested whether an fMRI could predict whether a dog would be a successful service dog. The study was published online in Scientific Reports in March 2017.

They tested 49 dogs entering service training, all between 17 and 21 months old. Of the dogs included in the study, 33 completed service training. Of the others, 10 were released for behavioral reasons, 4 were selected for breeding, and 2 were released for medical reasons.

There have been some fMRI studies done on dogs in the past, but the researchers indicated those studies had small sample sizes, making them difficult to generalize. There are have also been some hints of individual differences between dogs’ brain responses relating to behavior, temperament, and personality.

Researchers were interested in answering whether there is “a neurobiological phenotype that contributes to a dog’s ability to perform tasks required of service dogs,” reasoning that if there is, they should be able to test for it. This could be significant considering the cost of training a service dog is between $20,000 and $50,000 and being able to predict whether a dog will fail a training program and pulling it out early could save money.

The dogs chosen for the study were either Labrador or golden retrievers or a mix of those breeds. This meant the dogs’ brain size and morphology were more similar than in previous fMRI studies using a wider range of breeds, allowing for consistent normalization.

They singled out three regions of the brain as biomarkers: caudate, amygdala, and a region of the temporal cortex. The caudate was to test reward sensitivity, the amygdala to test arousal, and the part of the temporal cortex that was responsive to faces.

Dogs were acclimated to the MRI scanner noise and enclosure as well as other steps in the scan process. While they were in the scanner, a person would give a signal, either indicating they would receive a reward or would not receive an award. The person doing the signaling with either their owner/handler or a stranger.

Successful dogs showed a strong caudate response for the signal they would receive a reward, which could represent a strong motivational signal no matter who is giving the signal. Dogs who showed more activity in the amygdala in response to the treat, especially if a stranger gave the signal, were more likely to fail the service training. According to the study, “These results suggest that, as indexed by caudate activity, successful service dogs generalize associations to hand signals regardless who gives them but without excessive arousal as measured in the amygdala.”

Looking at the success of predicting the dogs’ success or failure, in the best care scenario, 8 of the 10 dogs were correctly identified and 4 were flagged incorrectly. While there is a cost to the false negatives, the researchers estimated that pulling the dogs from training early would have resulted in a net savings.

The researchers concluded their results could hold significant predictive value for some types of service dog training. They also reasoned that identifying the caudate and amygdala activations could lead to studies that “could provide a window into the dog’s internal state independently of behavior.”

Photo credit: © iStock/TerryJ