用合成数据提前判断医学模型能否迁移到真实肺部CT，无需真实标签。
When Does Synthetic CT Transfer? A Label-Free Donor/Host Diagnostic for Medical Vision-Language Model Routing on Real Lung CT
- 基于合成数字孪生数据，区分可迁移的病灶特征与不可迁移的解剖特征。
- 病灶大小和存在性排序在真实数据中转移率高达R2 ≥ 0.96，叶级信息不迁移。
- 仅靠合成数据即可构建无监督模型路由系统，适合缺乏标注的医疗场景。
合成数据对模型能力的评估只有在迁移到真实数据后仍有效才有意义，但真实标签恰恰是医学影像中最缺乏的。我们提出一个无需标签的预测机制：在合成数字孪生数据上，由供体（病灶本身）决定的能力能跨越合成到真实数据的转变，而由宿主（周围解剖结构）决定的能力则未必。我们在三个肺部CT视觉语言任务上验证该机制，涵盖五种公开的视觉语言模型、四种引导条件和七个真实数据集。结果在所有情况下均成立：病灶存在性和大小排序可成功迁移（R2 ≥ 0.96），叶级信息则不能；该判别能力在留源外校准下依然稳定，且在任何真实标签出现前即已明确。TrialCouncil——一种仅在合成CT上校准的无训练集成系统——在预测可迁移处精确匹配最佳固定模型的表现。核心贡献并非路由器本身，而是发现模型迁移性可仅从合成数据中无标签地预先预测。
原文摘要 · Abstract (English)
A synthetic measurement of model competence is useful only if it survives the move to real data, yet the real labels that would verify it are exactly what medical imaging lacks. We ask whether transfer can be predicted in advance, label-free, and answer with a mechanism: on synthetic digital twins, competence that is donor-driven (a property of the transplanted nodule) survives the synthetic to real change of host, while host-driven competence (a property of the surrounding anatomy) need not. We test this on three lung CT vision-language tasks chosen to span that axis, across five public VLMs, four guidance conditions, and seven real datasets. The prediction holds in every case: presence and size orderings transfer (R2 >= 0.96), lobe does not; the split survives leave-source-out calibration, and the diagnostic names that boundary before any real label. TrialCouncil, a training-free council calibrated only on synthetic CT, confirms it by matching the best fixed model exactly where transfer is predicted. The contribution is not the router but the finding that transfer itself is predictable, label-free, from synthetic data alone.
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。