SEP 22, 2022 7:24 AM PDT

Machine Learning Algorithm Offers Insights into a Dog's Neural Activity

WRITTEN BY: Kerry Charron

Emory University researchers decoded images from a canine brain using functional magnetic resonance imaging (fMRI) to reconstruct what and how a dog might see. The researchers found that dogs are attuned to what is happening in the immediate environment rather than to who or what is doing the action. The study published in Journal of Visualized Experiments provides insights into different ways of thinking used by humans and other animals. 

The researchers recorded and analyzed functional magnetic resonance imaging (fMRI) neural data for two awake, unrestrained dogs as they watched videos in three different 30-minute sessions. In order to obtain suitable fMRI imaging, the researchers trained the dogs to walk into a fMRI scanner and remain still and unrestrained. Recent advancements in machine learning and fMRI have been successful in decoding visual stimuli from the human brain, but the technique has only been used with humans and some primates. fMRI allows scientists to detect within brain-data patterns the different objects, movements, or actions that an individual observes while watching a video.

The researchers created video content to hold a dog’s attention for an extended period. The Emory research team used a video recorder attached to a gimbal and selfie stick to film scenes from a dog’s perspective. Canine-oriented scenes showed dogs sniffing, playing, eating, or walking on a leash. Activity scenes included a cat walking in a house, or people interacting, eating, or throwing a ball. The video data was categorized by time stamps into object-based classifiers (such as dog, car, human, cat) and action-based classifiers (such as sniffing, playing, or eating). 

Two humans participated in the same experiment. They also watched the same 30-minute video in three separate sessions and while fMRI recorded brain patterns. The brain data was then mapped onto the video classifiers using time stamps. 

The results showed major differences in how human and dog brains function. for the two human subjects found that the model developed using the neural net showed 99% accuracy in mapping the brain data onto both the object- and action-based classifiers. Decoding video content from the dogs revealed that the model did not work for the object classifiers, but it was 75% to 88% accurate at decoding the action classifications for the dogs. 

The study highlighted major differences in the visual systems of dogs and humans. Dogs only detect shades of blue and yellow, but their visual processing is better at detecting motion due to higher density of vision receptors. Scientists believe there may be evolutionary reasons for this difference. According to lead author Dr. Gregory Burns, “It makes perfect sense that dogs’ brains are going to be highly attuned to actions first and foremost. Animals have to be very concerned with things happening in their environment to avoid being eaten or to monitor animals they might want to hunt. Action and movement are paramount.” The study’s findings have implications for animal behavior research and ecological initiatives such as predator reintroduction programs. In addition, the study offers a foundation for future research using neural modeling.  

Sources: Emory University, Eureka Science Alert, Journal of Visualized Experiments

 

About the Author
BA and MA in English, MPS in Human Relations, and Ed.D. in Higher Education Administration
Kerry Charron writes about medical cannabis research. She has experience working in a Florida cultivation center and has participated in advocacy efforts for medical cannabis.
You May Also Like
OCT 09, 2022
Genetics & Genomics
Small Trial of Down Syndrome Treatment Sees Positive Results
OCT 09, 2022
Small Trial of Down Syndrome Treatment Sees Positive Results
Down syndrome is the most common genetic type of intellectual disability. It occurs when a person carries an extra copy ...
OCT 05, 2022
Health & Medicine
SARS-CoV-2 Infects Cells in the Brain & May Cause Structural Changes
OCT 05, 2022
SARS-CoV-2 Infects Cells in the Brain & May Cause Structural Changes
Researchers have now published a study that has revealed how SARS-CoV-2, the virus that causes COVID-19, can infect brai ...
NOV 14, 2022
Neuroscience
How COVID-19 Can Disrupt the Brain
NOV 14, 2022
How COVID-19 Can Disrupt the Brain
The virus that causes COVID-19 does not infect neurons, or if it does, that only happens in rare cases. But COVID-19 can ...
NOV 14, 2022
Drug Discovery & Development
Experimental Opioid Vaccine Could Prevent Overdose Deaths
NOV 14, 2022
Experimental Opioid Vaccine Could Prevent Overdose Deaths
Researchers have developed a vaccine that may be able to block the opioid fentanyl from reaching the brain. They note th ...
NOV 28, 2022
Neuroscience
Modified 'Bath Salts' Hold Potential to Treat Depression
NOV 28, 2022
Modified 'Bath Salts' Hold Potential to Treat Depression
Experimental synthetic drugs may treat depression and anxiety-related disorders with few adverse side effects. The corre ...
NOV 29, 2022
Genetics & Genomics
Eliminating 'Toxic' RNA Molecules to Treat ALS & Dementia
NOV 29, 2022
Eliminating 'Toxic' RNA Molecules to Treat ALS & Dementia
Researchers have developed a way to target RNA molecules that can have a harmful result. Instead of targeting the genome ...
Loading Comments...