An AI model has successfully identified goats experiencing pain, which could enhance treatment methods for animals and individuals who cannot communicate verbally.
A goat was brought in with a bladder stone, showing signs of discomfort and sadness.
This goat wouldn’t even chew its cud.
The patient in this scenario was indeed a goat. While he received treatment for his bladder stone—a frequent issue for these small ruminants—he also played a key role in a new study aimed at accurately determining pain levels not only in goats but also in various domestic animals and potentially even humans in the future.
“By addressing pain in animals, we can also tackle pain assessment in children and other individuals who are unable to express their discomfort,” explained Ludovica Chiavaccini, D.M.V., D.E.S., M.S., a clinical associate professor of anesthesiology at the University of Florida’s College of Veterinary Medicine.
Chiavaccini and her team recorded the facial expressions of both pained and comfortable goats. They then utilized this data to train an artificial intelligence model that learned to recognize when goats were in pain simply by analyzing their facial features.
This AI system, which has been trained on 40 goats so far, displayed accuracy rates ranging from 62% to 80% in recognizing pained expressions, depending on the testing methods used. As more data about additional goats and other animal species becomes available, these AI tools could significantly improve how clinicians manage pain in patients who cannot verbally communicate.
“This issue goes beyond just animal welfare,” said Chiavaccini. “We know that animals suffering from pain struggle to gain weight and are less productive. Farmers are increasingly recognizing the importance of managing both acute and chronic pain in their animals.”
While incorporating AI-based pain assessment tools in veterinary practices will necessitate further investigation, it promises to address a challenge that has long persisted in animal healthcare. Historically, measuring pain in animals has been subjective and challenging. Veterinarians have traditionally depended on years of experience for assessment. In recent years, researchers have created standardized pain scales for various species to minimize subjectivity, though the reliability of these measures can vary significantly.
When Chiavaccini and her team initiated their study—sparked by a graduate student’s passion for goats—there was no pain scale designed specifically for goats. Currently, there is a single pain scale for goats, but it has only been validated for male goats undergoing castration, highlighting the necessity for a more widely applicable system, Chiavaccini noted.
The research team shared their results on November 7 in the journal Scientific Reports.