Published Paper Details:

Knee osteoarthritis (KOA) is an ordinary degenerative joint sickness that extensively affects the pleasant of life. Early and accurate diagnosis is critical for effective management and treatment. This project, titled "Predicting and Classifying Knee Osteoarthritis with Deep Learning on X-ray Images," aims to develop an automated system for accurate prediction and classification of knee osteoarthritis (KOA) severity from X-ray images using deep learning, addressing the limitations of subjective visual assessment. Current KOA diagnosis relies heavily on visual interpretation of X-rays by radiologists, often using the Kellgren-Lawrence (KL) grading system. This method is subjective, leading to inter-observer variability and potential diagnostic inconsistencies. This model classifies knee osteoarthritis from normal to a severe stage. We propose a deep learning-based system employing Convolutional Neural Networks (CNNs) to automatically analyse knee X-ray images and classify KOA severity into different KL grades. The core algorithm employed is a Convolutional Neural Network (CNN) with developing a modified VGG16 and MobileNetV2 as the foundational model. This study underscores the transformative potential of deep learning in automating KOA diagnosis and classification. The system's high accuracy and scalability, especially with the MobileNetV2 model, make it a viable solution for real-world clinical applications, promising significant improvements in diagnostic efficiency and patient outcomes.