Automated Assessment of Pelvic Longitudinal Rotation Using Computer Vision in Canine Hip Dysplasia Screening

Canine hip dysplasia is a painful condition common in large dog breeds, leading to joint instability and osteoarthritis. Accurate diagnosis is crucial for guiding breeding decisions to help reduce its prevalence. However, evaluating hip health status can be challenging, as minor positioning changes during X-rays can distort images and hinder proper hip joint assessment. In this study, we developed an artificial intelligence tool for the automatic evaluation of pelvic alignment in X-rays. By detecting subtle asymmetries in bone structure, the tool determines whether a dog’s hips are properly aligned. Our findings showed that the artificial intelligence tool performed as accurately as an expert human examiner in detecting misalignment. This automated approach could improve the reliability of canine hip dysplasia screenings, saving veterinarians time and reducing misdiagnosis risks due to human error. Ultimately, this technology has the potential to enhance medical care for dogs and support breeders in making more informed choices, contributing to canine health improvements and reducing canine hip dysplasia incidence in future generations.