A Comparative Analysis Of Convolutional Reinforcement Learning Technique To Detect Rheumatoid Arthritis

Abstract

Rheumatoid Arthritis (RA) is an autoimmune disease that damages body tissue and affects the joints. Early detection of rheumatoid arthritis by hand and wrist necessitates an efficient system analysis. The hand and wrist joints are the first to be affected by Rheumatoid Arthritis. The most commonly used imaging modality for rheumatoid arthritis is magnetic resonance imaging (MRI) (RA). The earliest symptom of RA is synovitis (inflammation of synovial fluid). The human eye fails to detect very subtle changes during diagnosis, making it a difficult task for medical specialists. To detect RA automatically, we created a Convolutional Reinforcement Learning Technique (CRLT) based on the KCP algorithm in this work. The model is created using 6200 MRI images of the hand and wrist. The model is found to be very efficient when compared to various CNN Architectures (VGG-16, ResNet-50, and Inception V3), with Test Accuracy of 79.35 percent, Loss of 0.51578545, F1 of Score 0.76, and an Error rate of 0. 0.00055. As a diagnostic tool, this model would be useful.