Man’s best friend is learning a new trick.
Sign up for our newsletter
Get The Plainsman straight to your inbox.
The department of psychology, the MRI Research Center and the Canine Performance Sciences program in the College of Veterinary Medicine are collaborating research on the best ways to train detection dogs. The research involves the dogs lying still and awake on an MRI scanner.
“The overall goal is to learn and understand better how dogs process information,” said Paul Waggoner, co-director of Canine Performance Sciences. “The particular interest for us is how dogs process odor information and human interaction.”
The data found could be used to help determine which dogs would be best suited for certain services, Waggoner said.
“If we can understand how dogs process information about human emotions and signals, we can train them to do the tasks we want them to do,” Waggoner said.
Gopikrishna Deshpande, associate professor at the MRI Research Center in the department of electrical and computer engineering, said he is responsible for the technicalities of the project.
“I’m the PI (principal investigator) for the project,” Deshpande said. “I make the imaging happen, coordinate all devices and stimuli and analyze data.”
Deshpande said the research meets two separate interests.
“The first is the basic science interest: understanding how dogs’ brains work and respond to human emotions,” Deshpande said. “And the second is the application aspect: how dogs respond to odors.”
MRI scanning on dogs that haven’t been placed under anesthesia is rare, but is essential in finding the most accurate results, Waggoner said.
“For example, when you dream, you do all sorts of things that aren’t possible when you are awake,” Waggoner said. “But when the dogs are awake, we are able to see a truer image of what the dog smells.”
Getting the dogs to hold still for five to 10 minutes with their head in a cone while on the MRI scanner is where the psychologist’s work comes in.
Jeff Katz, alumni professor of psychology, his graduate students and Waggoner have designed different tasks to train the dogs and to analyze their responses.
“We already know a lot about how humans process information, but we are now looking at how that differs or is similar to dogs,” Katz said. “Our role is scanning and developing data, using imaging and familiar and unfamiliar faces the dogs have never seen before, to see if there are different areas of the brain that are activated based on what they’ve seen.”
Part of the testing involves presenting the dogs with a positive, negative and neutral person and then judging their reactions to each. While completing the tasks, the dogs are also tested to see if they respond better to commands given from someone they are familiar with as opposed to someone they are unfamiliar with.
Although there are multiple phases of research and testing, Katz said the results so far have been interesting.
“It is a definite yes that dogs respond differently to positive feedback and to someone they are familiar with,” Katz said. “We notice a strong response in the amygdala.”
Lucia Lazarowski, second-year graduate student in psychology, has been working with Katz to train the dogs and analyze the data.
“One of the tests is called the ‘unsolvable task,’” Lazarowski said. “There is a box with a treat, and you stand to one side of the room and the dog flips the lid and gets the treat. Then it becomes harder. We measure to see how often they look up at the familiar person.”
It is important to distinguish service dogs from working dogs, Lazarowski said.
“Service dogs refers to dogs that are going to be placed with someone disabled or a veteran with PTSD," Lazarowski said. "They are more for performing daily tasks. But the detection dogs are the ones used to sniff out odors.”
Although the research is being used as of now to determine which dogs would make the best detection dogs, Katz said he is hoping they will soon branch out.
“We plan to work with this other cohort of service dogs that are more therapy dogs,” Katz said. “Because you may find that what makes one a good therapy dog would make another a bad detection dog.”
The team began research in 2011 after Deshpande received a grant to acquire funding from the Defense Advanced Research Projects Agency.
“The dogs cost around $25,000-$30,000 each,” Deshpande said. “In 2011, we received an internal grant to start the first pool of testing.”
Katz said they hope to publish some of their preliminary findings in the next couple years.
Do you like this story? The Plainsman doesn't accept money from tuition or student fees, and we don't charge a subscription fee. But you can donate to support The Plainsman.Support The Plainsman