Machines may soon play a role in monitoring pain in horses admitted to veterinary clinics. Currently, it can be difficult to find an unbiased, quick way to determine if a horse is in pain. Drs. Nuray Kil, Katrin Ertelt and Ulrike Auer created a study that used an automated video tracker to detect and record daily equine activities. The end goal was to have the tracker create an algorithm that would be able to objectively assess pain and wellbeing of horses in a clinical setting. This ability would remove the guesswork of veterinarians and technicians determining if a horse was in pain.
Pain causes behavior changes in horses; understanding normal and pain-induced behaviors in horses is critical to properly evaluating pain levels. Though horses may work to mask pain in an unfamiliar surrounding like a clinic, even subtle variations become apparent when behavior is thoroughly analyzed.
Though there are multiple pain assessment scales available, they are all scored manually and can be skewed by multiple things, including inexperience and the amount of time spent viewing the horse.
To test their video tracker theory, researchers used 34 horses at the University of Veterinary Medicine in Vienna's teaching hospital. All the horses were housed in box stalls with water and were fed four times a day. The horses were recorded on an action camera and in time-lapse mode. The videos were then processed to look for an automated prediction of three body parts: the tail, nose and withers.
The technology was able to identify the horse's stance with an accuracy and sensitivity of more than 80 percent, meaning that it could more often than not detect when a horse was exhibiting pain behaviors. The research team concluded that this technology will improve the detection of equine pain and provide insight for equine behavioral research.
Read the study here.
Read more at HorseTalk.
The post Study Shows Machines Can Detect Equine Pain From Daily Behavior appeared first on Horse Racing News | Paulick Report.
