Detection of Knee Osteoarthritis Stages Using Convolutional Neural Network

Nowadays, many people suffer from the problem of knee osteoarthritis (KOA). As the age increases, the chances of suffering from knee osteoarthritis also increases simultaneously. Degeneration of the articular cartilage, the flexible, slick substance that typically shields bones from joint friction and impact, is what constitutes knee osteoarthritis. The disorder can also damage neighbouring soft tissues and result in alterations to the bone that lies beneath the cartilage. This study uses deep learning-based feature extraction and classification to demonstrate knee osteoarthritis detection at an early stage. At first, the x-ray images of different stages of knee osteoarthritis have taken for further processing. For classification, a mix of both healthy knee and defected has been taken. Following that, the deep convolutional neural network (CNN) model was utilised to determine if the individual had knee osteoarthritis or not. There are four grades in the Kellgren–Lawrence (KL) system: Grade I, Grade II, Grade III, and Grade IV. This study offers 95% accuracy for identifying knee osteoarthritis. The main objective of this study is to detect the different stages of knee osteoarthritis by performing image processing on x-ray images. This research helps to detect knee osteoarthritis at early stages and classify knee osteoarthritis stages using the KL grading system. © 2023, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.