Medical Image Analysis of Knee Osteoarthritis using Modified Deep CNN

Abstract

Kellgren and Lawrence (KL) grading method is mainly used by the clinicians for grading the X-ray images. However, grading individual images are prone to errors. This study proposes an approach for automated knee osteoarthritis classification based on deep neural networks. Diagnosis of Osteoarthritis involves splitting the knee X-ray into the healthy knee or unhealthy knee with KL grading. Here we use a 20-layer deep residual networks ie ResNet 20 for automated knee osteoarthritis classification. ResNet 20 has 19 convolutional layers and 1 fully-connected layer.We analysed the dataset with different CNN models. And got the different performance.