AI Model Decodes Pain in Goats, With Potential Human Benefits

Researchers at the University of Florida's College of Veterinary Medicine, led by Ludovica Chiavaccini, have developed an **artificial intelligence model** that recognizes pain in goats through facial expressions, a revolutionary step for veterinary science. This method promises to address the challenges of subjective pain assessment in animals—a persistent hurdle for veterinarians. The AI model, trained and tested on a dataset of 40 goats, achieved an accuracy range of 62% to 80% in identifying pain by facial cues alone. This groundbreaking study, published in *Scientific Reports*, paves the way for developing reliable pain scales for various animal species, potentially extending to non-verbal human patients like children. The initiative emphasizes both animal welfare and productivity in farming, as animals in pain exhibit reduced weight gain and output. Despite the current reliance on experience and inconsistent pain scales in veterinary medicine, AI could offer a consistent, objective tool for pain management. The findings underscore the need for more broad-based systems applicable across different species, especially as the current goat pain scale is limited to male goats undergoing castration.