A novel method to Predict knee osteoarthritis

Abstract

Osteoarthritis (OA) is the most prevalent form of arthritis, affecting the knee in particular and leaving many severely disabled around the globe. A degenerative musculoskeletal illness that affects a osteoarthritis (OA). About six percent of the global population lives with this condition. Articular cartilage at the ends of bones gradually wears away in the knee, the joint most often affected by osteoarthritis (O-A). The knee joint is affected by osteoarthritis (O-A), a degenerative joint disease. Despite its inefficiency and sensitivity to user variables, manual knee joint segmentation and annotation are still employed in clinical practice for OA diagnosis. Therefore, to address the shortcomings of the traditional method and enhance the efficiency of the clinical workflow, a solution based on the VGG16 model is suggested. Clinical examination and plain radiography are now the methods used to diagnose knee osteoarthritis. We require a solution to the clinically important concerns of OA development prediction and early change detection. One clinically essential issue in OA is the prediction of the disease's progress and the detection of early abnormalities; nevertheless, imaging-based diagnostic procedures may be improved using innovative quantitative approaches. We provide a new Deep Learning (DL) approach to predict the evolution of osteoarthritis (OA) using medical images of the knee in this research. The system's model is something we build. The fit function is then used to achieve this goal. Each batch will have a maximum of six. Graphs of accuracy and loss will be plotted thereafter. While 87% of validation attempts were successful, 95% of training attempts were successful.

Keywords