Low-Contrast-Enhanced Contrastive Learning for Semi-Supervised Endoscopic Image Segmentation

TL;DR

This paper introduces LoCo, a semi-supervised endoscopic image segmentation framework that uses low-contrast-enhanced contrastive learning to improve accuracy and robustness, especially for low-contrast and minority classes.

Contribution

The paper proposes a novel semi-supervised segmentation method with ICE, BCE, and CDF strategies to better utilize unlabeled data and enhance low-contrast pixel segmentation.

Findings

Achieves state-of-the-art segmentation accuracy on public datasets.

Effectively segments low-contrast pixels among different tissue types.

Significantly outperforms previous methods in robustness and accuracy.

Abstract

The segmentation of endoscopic images plays a vital role in computer-aided diagnosis and treatment. The advancements in deep learning have led to the employment of numerous models for endoscopic tumor segmentation, achieving promising segmentation performance. Despite recent advancements, precise segmentation remains challenging due to limited annotations and the issue of low contrast. To address these issues, we propose a novel semi-supervised segmentation framework termed LoCo via low-contrast-enhanced contrastive learning (LCC). This innovative approach effectively harnesses the vast amounts of unlabeled data available for endoscopic image segmentation, improving both accuracy and robustness in the segmentation process. Specifically, LCC incorporates two advanced strategies to enhance the distinctiveness of low-contrast pixels: inter-class contrast enhancement (ICE) and boundary contrast enhancement (BCE), enabling models to segment low-contrast pixels among malignant tumors, benign tumors, and normal tissues. Additionally, a confidence-based dynamic filter (CDF) is designed for pseudo-label selection, enhancing the utilization of generated pseudo-labels for unlabeled data with a specific focus on minority classes. Extensive experiments conducted on two public datasets, as well as a large proprietary dataset collected over three years, demonstrate that LoCo achieves state-of-the-art results, significantly outperforming previous methods. The source code of LoCo is available at the URL of \href{https://github.com/AnoK3111/LoCo}{https://github.com/AnoK3111/LoCo}.