用轮廓感知位移提升医学图像少样本分割精度
UCAD: Uncertainty-guided Contour-aware Displacement for semi-supervised medical image segmentation
- 基于超像素生成与解剖结构对齐的区域，保持边界一致性
- 通过不确定度引导选择难样本进行位移，提升模型稳定性
- 适合标注数据稀缺的医学图像分割任务
现有半监督分割中的位移策略仅作用于矩形区域，忽略解剖结构，导致边界失真和语义不一致。为此，我们提出UCAD——一种不确定性引导的轮廓感知位移框架，可在保持轮廓语义的同时增强一致性学习。该方法利用超像素生成与解剖边界对齐的区域，并设计不确定性引导的选择机制，有选择地对困难区域进行位移以促进一致性学习。进一步提出动态不确定性加权一致性损失，自适应稳定训练过程，在未标注区域有效正则化模型。大量实验表明，UCAD在标注有限条件下持续优于现有最先进方法，显著提升分割精度。代码已公开：https://github.com/dcb937/UCAD。
原文摘要 · Abstract (English)
Existing displacement strategies in semi-supervised segmentation only operate on rectangular regions, ignoring anatomical structures and resulting in boundary distortions and semantic inconsistency. To address these issues, we propose UCAD, an Uncertainty-Guided Contour-Aware Displacement framework for semi-supervised medical image segmentation that preserves contour-aware semantics while enhancing consistency learning. Our UCAD leverages superpixels to generate anatomically coherent regions aligned with anatomy boundaries, and an uncertainty-guided selection mechanism to selectively displace challenging regions for better consistency learning. We further propose a dynamic uncertainty-weighted consistency loss, which adaptively stabilizes training and effectively regularizes the model on unlabeled regions. Extensive experiments demonstrate that UCAD consistently outperforms state-of-the-art semi-supervised segmentation methods, achieving superior segmentation accuracy under limited annotation. The code is available at:https://github.com/dcb937/UCAD.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。