C3S3: Complementary Competition and Contrastive Selection for Semi-Supervised Medical Image Segmentation

TL;DR

C3S3 is a semi-supervised medical image segmentation model that enhances boundary accuracy by integrating contrastive learning and competitive pseudo-label generation, leading to significant performance improvements on MRI and CT datasets.

Contribution

The paper introduces C3S3, a novel model combining contrastive learning and competitive pseudo-labeling to improve boundary delineation in semi-supervised medical image segmentation.

Findings

Achieves at least 6% improvement on 95HD and ASD metrics.

Outperforms previous state-of-the-art methods on MRI and CT datasets.

Demonstrates robust boundary delineation and segmentation accuracy.

Abstract

For the immanent challenge of insufficiently annotated samples in the medical field, semi-supervised medical image segmentation (SSMIS) offers a promising solution. Despite achieving impressive results in delineating primary target areas, most current methodologies struggle to precisely capture the subtle details of boundaries. This deficiency often leads to significant diagnostic inaccuracies. To tackle this issue, we introduce C3S3, a novel semi-supervised segmentation model that synergistically integrates complementary competition and contrastive selection. This design significantly sharpens boundary delineation and enhances overall precision. Specifically, we develop an Outcome-Driven Contrastive Learning module dedicated to refining boundary localization. Additionally, we incorporate a Dynamic Complementary Competition module that leverages two high-performing sub-networks to generate pseudo-labels, thereby further improving segmentation quality. The proposed C3S3 undergoes rigorous validation on two publicly accessible datasets, encompassing the practices of both MRI and CT scans. The results demonstrate that our method achieves superior performance compared to previous cutting-edge competitors. Especially, on the 95HD and ASD metrics, our approach achieves a notable improvement of at least 6%, highlighting the significant advancements. The code is available at https://github.com/Y-TARL/C3S3.