A Comprehensive Review on Advances in Detection of Knee Osteoarthritis

One joint ailment that frequently affects the knee is osteoarthritis (KOA). Osteoarthritis is primarily caused by the deterioration of joint cartilage, which leads to contact between adjacent bones and is a contributing factor to stiffness, pain, and limited movement. The ongoing accumulation of medical data pertaining to KOA has led researchers to consider artificial intelligence in the diagnosis of KOA. Numerous techniques for automatically determining the KL grade from the radiographs have been devised. The standard for measuring the severity of osteoarthritis is the Kellgren–Lawrence (KL) grading system, which divides the condition into five categories. For individuals with osteoarthritis (OA), measuring and visualising cartilage thickness is helpful in identifying the disease's early stages. For the purpose of segmenting images of the knee joint, numerous algorithms are available. They fall into two categories: model-based and pixel-based techniques. A number of machine learning-based approaches as well as the sophisticated deep learning-based convolutional neural network (CNN) are examined and contrasted. This paper discusses and provides a brief overview of joint space narrowing, volume measurement, and techniques for segmenting the articular cartilage of the knee joint.