Abstract
While deep learning approaches are mostly developed for single images, in real world applications, images are often taken as a series to inform decision making. Due to hardware (memory) and software (algorithm) limitations, so far, no methods have been developed to integrate multiple images. In this study, we present an approach that seamlessly integrates deep learning and traditional machine learning, to study multiple images and score rheumatoid arthritis joint damage. This method allowed the quantification of joining space narrowing to approach the clinical upper limit. Beyond predictive performance, we integrated the multilevel interconnections across joints and damage types into the machine learning model and revealed the cross-regulation map of joint damages in rheumatoid arthritis.