Order-Guided Disentangled Representation Learning for Ulcerative Colitis Classification with Limited Labels

Ulcerative colitis (UC) classification, which is an important task for endoscopic diagnosis, involves two main difficulties. First, images with the annotation about UC (positive or negative) are usually limited. Second, they show a large variability in their appearance due to location colon. Especially, second difficulty prevents us from using existing semi-supervised learning techniques, common remedy first difficulty. In this paper, we propose practical method classification by newly exploiting additional features, colon (e.g., left colon) and image capturing order, both of often attached individual sequences. The proposed can extract essential information efficiently disentanglement process those features. Experimental results demonstrate that outperforms several methods task, even small number annotated images.