Enhanced Diagnostic Precision in Knee Osteoarthritis: Leveraging MobileNetV2 for Deep Learning Classification

Osteoarthritis, a type of arthritis worldwide, affects millions of Americans by breaking down the protective cartilage that cushions the ends of the bones, most especially in joints of the hands, knees, hips, and spine. In this progressive condition of joint wear and tear, one usually experiences pain, stiffness, swelling, loss of function and sometimes disability that interferes with activities and work performance. The study talks about the identification of osteoarthritis through MobileNetV3. It is a state-of-the-art neural network structure based on device blocks. The model gives an accuracy of 82.94% for the loss of 0.37498 in the test. Precision Depending upon the case, it was 0.75, and for osteoarthritis cases, it was 0.89. Its overall recall is 0.83, as its F1 score is 0.79 in cases and 0.86 for osteoarthritis cases. The average precision scorings, both the macro and weighted recall rates, and F1 scores were approximately 0.82 to 0.83, respectively, demonstrating the effectiveness of MobileNetV3 concerning the classifications of osteoarthritis that might be used early in detection and treatment management.