An Investigative Approach to Knee Osteoarthritis Detection Using Deep Learning on X-ray Radiographs

Le résumé fourni par la source

A chronic joint condition called osteoarthritis affects the ligaments, capsule, synovium, subchondral bone, and articular cartilage. Osteoarthritis most frequently develops in the knee joints. A scoring system developed by Kellgren and Lawrence (KL) is used to measure the severity of osteoarthritis. The range of the KL score is 0–4. Grades 0 through 4 represent normal, doubtful mild, moderate, and severe, respectively. In this work, we classified 3 levels of Osteoarthritis by modifying the KL scoring system. This research identifies grade 0 for normal and doubtful, grade 1 for mild and moderate, and grade 2 for severe. To identify these classes from knee radiograph images, we used a deep-learning architecture namely ResNet-50, and customized it. Using our proposed model, we achieved 97% training accuracy. Clinical researchers will find this research useful in effectively determining the degree of osteoarthritis in the knee.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.