Equine scientists are using artificial intelligence to recognise pain in horses from subtle changes in their facial expressions.
A team at Nottingham Trent University in England has launched an international project, amid hopes it will result in a better understanding of how painful certain conditions are, aid the development of better treatments, and improve monitoring of recovery.
The researchers are training a computer to automatically identify and track key features of a horse’s face that have been shown to change in response to pain.
The machine will then analyse videos of horses that have recently undergone surgery or are recovering from illness to see how these facial features change as their condition changes and they’re given pain relief.
As the system can track much more subtle fluctuations than the human eye, the researchers hope to learn much more about how horses express their pain and how this differs depending on their personality.
It is hoped that the system may be able to detect other emotions in the future and potentially help people to better understand how their horses feel about what they’re experiencing.
“The immediate application with regards to pain is for research and veterinary facilities to use the system to scan videos,” said Dr Carrie Ijichi, senior lecturer in the university’s School of Animal, Rural and Environmental Sciences.
“This way, we can learn about how truly painful certain conditions are and how effective pain medication or treatments are. It would also help guide us in ensuring horses are given the most appropriately dosed treatment.
“We want to find markers that are sensitive to pain, better than the human eye can see, and which will provide us with an accurate grading for pain.
“As the technology improves, hopefully this is something that anyone could have at home and use to assess pain in a range of settings, including potentially while the horse is being ridden.”