TRANSFER LEARNING‐BASED CLASSIFICATION OF KNEE OSTEOARTHRITIS SEVERITY FROM X-RAY IMAGES

Knee osteoarthritis (KOA) a degenerative, long-term joint condition that, more often than not, affects the elderly and is characterized by articular cartilage degradation. Appropriate treatment and early analysis are essential for sickness control. However, traditional diagnostic methods for classifying KOA from X-ray images require laborious expertise and, unfortunately, have a large margin of error. This study presents an image processing-based solution for multi-classification KOA severity from X-ray images using the Bilateral filter, contrast-limited adaptive histogram equalization (CLAHE), and transfer learning models. The CLAHE method improved image quality, while the Bilateral filter enhanced details and minimized blurriness in X-ray images. KOA image dataset consists of 9786 knee images and five class labels. The performances of transfer learning models including AlexNet, ResNet101, DenseNet201, and VGG19 were compared. The ResNet101 model emerged as the most effective, achieving a kappa statistic of 0.970, weighted F1-score of 0.978, and an overall accuracy of 97.85%. This model’s high accuracy and precision make it a dependable and objective diagnostic solution.