The patient shuffled in, a dull ache clearly evident in his eyes. This patient wasn’t a typical hospital visitor—he was a goat, grimacing with discomfort from a painful bladder stone. While veterinarians at the University of Florida treated his immediate health issue, the goat was also making an unexpected contribution to groundbreaking research: developing an AI model that can "see" pain in animals.

Bladder stones are a common issue in goats and other small ruminants, often resulting in intense discomfort. But recognizing pain in animals is a challenging task, traditionally relying on the subjective experience of veterinarians. Enter Dr. Ludovica Chiavaccini, D.M.V., D.E.S., M.S., and her team, who are working to create an objective, AI-driven pain scale for goats and, eventually, other animals and even humans.

Chiavaccini, a clinical associate professor of anesthesiology at the University of Florida’s College of Veterinary Medicine, saw a unique opportunity: if AI could learn to recognize pain in goats, it could open the door for measuring pain in non-verbal patients across species. “If we solve the problem with animals, we can also solve the problem for children and other non-verbal patients,” Chiavaccini explained.

To develop the model, researchers filmed the faces of goats in both pain and comfort, feeding this data into an artificial intelligence system. The AI, trained and tested on 40 goats so far, has achieved an accuracy rate of 62% to 80% in identifying painful expressions. With continued training and expansion to other animal species, this model may help veterinarians and medical professionals manage pain in patients who cannot express their discomfort.

The implications go beyond animal welfare. Chiavaccini noted that farm animals in pain not only suffer but also experience reduced weight gain and productivity, making pain management an essential component of modern farming practices.

The use of AI in this setting could revolutionize how veterinarians and caregivers assess pain. Until recently, pain assessment was highly subjective, relying on a veterinarian’s experience and the few species-specific pain scales available. While some standardized pain scales have been developed, they are often limited in scope. When Chiavaccini and her team began their research, there was no pain scale for goats. Today, a single pain score exists but is validated only for male goats undergoing castration. The need for a more generalizable system is clear.

For now, this AI model represents a significant step forward. The study’s findings were published on November 7 in Scientific Reports, marking an exciting milestone in the field of veterinary anesthesiology. With continued research and refinement, AI-powered pain scales may soon become an essential tool in both animal and human healthcare, offering a new level of empathy and care for all non-verbal patients.