Knee Osteoarthritis Detection And Classification Using Transfer Learning

Knee osteoarthritis (OA) is a common disease that causes pain, stiffness, and decreased mobility by wearing down the cartilage between the bones in the knee joint. Early and automatic detection of knee OA is important for ensuring effective treatment for patients. The aim of this article is to develop a deep learning model using ResNet50 architecture for the accurate classification of knee osteoarthritis (OA) from images of X-rays originally labeled using the Kellgren–Lawrence (KL) Scale. The dataset used in the current study involves images taken from 4796 participants of ages between 45 and 79 years, with approximately nine thousand knee X-ray images. The ResNet50 model is trained on this dataset, and various evaluation metrics were used to evaluate its performance. A flask-based web application was also developed for using the trained model. Our model achieved a high degree of accuracy (97.4%) in classifying knee OA, suggesting that it could be a valuable tool for diagnosing this condition quickly and accurately.