Osteoarthritis (OA) is a widespread condition without a currently available treatment, significantly affecting individuals’ quality of life as this is a progressive condition that deteriorates with time, frequently leading to persistent Joint pain and stiffness can become severe enough to make daily tasks difficult. Magnetic resonance (MR) images are commonly employed in OA diagnosis, where medical experts assess changes, especially within the tibio-femoral cartilage compartment for knee OA. This study presents a new diagnostic technique for knee OA that detects the illness from MR images using a Support Vector Machine (SVM) algorithm. Our suggested method employs the Independent Component Analysis (ICA) technique on 3-D MR imaging data from a real-world cohort. The experimental results show that our ICA-SVM machine learning model obtained 86% testing accuracy and 72% specificity. and sensitivity of 100%, having been trained on a limited MR image dataset.