Diagnosis of Degenerative Joint Disease of The Knee

Osteoarthritis, the predominant form of degenerative joint disease, imposes a considerable global health burden. The proposed innovative diagnostic approach not only meets the critical need for accurate detection but also holds promise in unraveling the intricate complexities of osteoarthritis. By leveraging clinical assessments, imaging technologies, and laboratory tests, the proposed system brings forth a transformative methodology for the early detection of degenerative knee joint disease. Through the incorporation of deep learning algorithms that consider risk elements, the system facilitates timely interventions. In the medical field, this project aims to pioneer a groundbreaking diagnostic system for the precise identification of degenerative joint disease in the knee. This study presents a novel approach for the automated diagnosis of DJD using knee X-ray images. The objective is to develop a classification model capable of accurately categorizing knee X-ray images into five distinct classes representing different stages of DJD severity: class 0 (normal), class 1 (doubtful), class 2 (mild), class 3 (moderate), and class 4 (severe). By seamlessly integrating clinical assessments, advanced imaging techniques, and state-of-the-art laboratory tests, This comprehensive system employs deep learning algorithms to analyze a wealth of data, including risk factors. The ultimate goal is to significantly enhance accuracy in detecting degenerative knee joint disease and enabling early intervention by grading their severity. Dataset contains 3341 images of different categories like normal,doubtful,mild,moderate,severe. A custom CNN is used in this paper, optimizer Adam, loss function spatial cross entropy, batch size 32 and epochs 300.This model gives the accuracy of 82%.