Scholarlink

<p>We present a label-efficient pipeline for CT auto-segmentation in resource-constrained settings. The framework couples a semi-supervised segmentation backbone with a foundation model-guided regularizer to strengthen the learning from scarce annotations. To better exploit volumetric context, we introduce a multi-view collaborative learning procedure that performs view-specific inference to form a unified supervision signal that suppresses view-dependent noise and improves mask fidelity. We evaluate on a public CT benchmark with varying numbers of labeled scans. In the highly label-limited regime, the approach yields strong accuracy with average Dice 83.79% for the liver and 60.08% for the tumor using 20 labeled cases, outperforms existing segmentation methods. By reducing contouring from hours to seconds, improving small-structure recovery and boundary fidelity, and requiring no interactive prompts, the method offers a plug-and-play path to deployment and a reliable basis for downstream radiomics and longitudinal monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Foundation model-guided multi-view semi-supervised CT segmentation of liver tumors in resource-constrained settings

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月29日

查看全文
  • Yuchuan Jiang,
  • Yao Du,
  • Kai Xiong,
  • Kuiyuan Huang,
  • Tengzheng Li,
  • Zhangyun Li,
  • Morang Zhang,
  • Xiaoning Gan,
  • Qiang Li,
  • Junjie Liang,
  • Mingrong Cao,
  • Jian Sun,
  • Junfu Wang,
  • Jinzhong Duanmu,
  • Xueqin Li,
  • Zhili Wen,
  • Qunguang Jiang,
  • Xin Yu,
  • Shangxiang Chen

摘要

We present a label-efficient pipeline for CT auto-segmentation in resource-constrained settings. The framework couples a semi-supervised segmentation backbone with a foundation model-guided regularizer to strengthen the learning from scarce annotations. To better exploit volumetric context, we introduce a multi-view collaborative learning procedure that performs view-specific inference to form a unified supervision signal that suppresses view-dependent noise and improves mask fidelity. We evaluate on a public CT benchmark with varying numbers of labeled scans. In the highly label-limited regime, the approach yields strong accuracy with average Dice 83.79% for the liver and 60.08% for the tumor using 20 labeled cases, outperforms existing segmentation methods. By reducing contouring from hours to seconds, improving small-structure recovery and boundary fidelity, and requiring no interactive prompts, the method offers a plug-and-play path to deployment and a reliable basis for downstream radiomics and longitudinal monitoring.