基于扩散模型的肿瘤修复用于临床数据稀缺下的肾脏分割
Diffusion-Based Tumor Inpainting for Renal Segmentation under Clinical Data Scarcity
浏览论文内容
中文总结 AI 辅助
针对临床数据稀缺下肾脏肿瘤分割问题,提出基于扩散模型的肿瘤修复框架,2.5D合成在降低假阳性同时保持Dice,且计算成本远低于3D。
中文摘要 AI 辅助
深度学习在肾脏肿瘤分割中的应用需要大量标注数据集,然而临床部署通常只能从目标站点获得少量肿瘤阳性病例。我们提出了一种基于扩散模型的修复框架,能够在健康CT扫描中合成解剖学上合理的肾脏肿瘤,无需额外标注,并首次系统比较了2D、2.5D和全3D(MAISI)合成策略在此任务中的表现。在公共数据(KiTS23、KIRC)上训练扩散模型,并在内部队列的三种低数据场景下评估nnU-Net分割性能,我们发现2.5D和3D增强显著减少了假阳性(从约18-20%降至约3-6%),同时保持了Dice系数,而2D增强未带来一致益处。关键在于,所提出的2.5D方法在每项指标上均与全3D合成相当,但计算成本大幅降低,表明仅局部体积一致性就足以在数据和资源稀缺的临床环境中实现有效增强。
英文摘要
Deep learning segmentation of renal tumors requires large annotated datasets, yet clinical deployments typically offer only a handful of tumor-positive cases from the target site. We propose a diffusion-based inpainting framework that synthesizes anatomically plausible renal tumors within healthy CT scans, requiring no additional annotation, and provide the first systematic comparison of 2D, 2.5D, and full 3D (MAISI) synthesis strategies for this task. Training the diffusion model on public data (KiTS23, KIRC) and evaluating nnU-Net segmentation on a internal cohort across three low-data regimes, we find that 2.5D and 3D augmentation substantially reduce false positives (from $\sim$18--20\% to $\sim$3--6\%) while maintaining Dice, whereas 2D provides no consistent benefit. Crucially, the proposed 2.5D method matches full 3D synthesis on every metric at substantially lower computational cost, indicating that local volumetric consistency alone is sufficient for effective augmentation in data- and resource-scarce clinical settings.
发表机构
- University of Tartu(塔尔图大学)
- University of Bern(伯尔尼大学)
- McGill University(麦吉尔大学)
- Mila – Québec Artificial Intelligence Institute(米拉-魁北克人工智能研究所)
- Better Medicine OÜ(Better Medicine 公司)
- STACC OÜ(STACC 公司)
机构由 AI 辅助整理,请以论文原文为准。