CLoE: Curriculum Learning on Endoscopic Images for Robust MES Classification

TL;DR

This paper introduces CLoE, a curriculum learning framework that improves endoscopic image-based ulcerative colitis severity classification by accounting for label noise and ordinal structure, enhancing robustness and accuracy.

Contribution

The paper presents a novel curriculum learning approach that incorporates image quality as a proxy for label confidence and combines it with augmentation to improve MES classification.

Findings

CLoE improves accuracy and robustness over baseline models.

ConvNeXt-Tiny achieves 82.5% accuracy on LIMUC.

The method effectively handles label noise in ordinal classification.

Abstract

Estimating disease severity from endoscopic images is essential in assessing ulcerative colitis, where the Mayo Endoscopic Subscore (MES) is widely used to grade inflammation. However, MES classification remains challenging due to label noise from inter-observer variability and the ordinal nature of the score, which standard models often ignore. We propose CLoE, a curriculum learning framework that accounts for both label reliability and ordinal structure. Image quality, estimated via a lightweight model trained on Boston Bowel Preparation Scale (BBPS) labels, is used as a proxy for annotation confidence to order samples from easy (clean) to hard (noisy). This curriculum is further combined with ResizeMix augmentation to improve robustness. Experiments on the LIMUC and HyperKvasir datasets, using both CNNs and Transformers, show that CLoE consistently improves performance over strong supervised and self-supervised baselines. For instance, ConvNeXt-Tiny reaches 82.5\% accuracy and a QWK of 0.894 on LIMUC with low computational cost. These results highlight the potential of difficulty-aware training strategies for improving ordinal classification under label uncertainty. Code will be released at https://github.com/zeynepozdemir/CLoE.