发表机构
NPI University of Bangladesh; Towson University; University of Asia Pacific（NPI孟加拉大学; 陶森大学; 亚洲太平洋大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出SpatialUQ，一种仅用输出概率的后验不确定性量化方法，通过空间一致性（Jensen-Shannon散度）在多个数据集上超越MC-Dropout，实现高效、校准良好的失败检测，并随模型质量提升性能。
AI 中文摘要
临床视觉模型通常作为冻结的黑盒部署，在推理时无法访问内部信息、进行重训练或获取真实标签。我们提出了SpatialUQ，一种仅使用输出概率的后验不确定性方法。它通过六次确定性前向传播，测量全局预测与五个固定空间裁剪均值之间的Jensen-Shannon散度。其前提简单：可信的预测在空间上是一致的。在NIH ChestX-ray14（DenseNet-121，N=25,596）上，我们的多裁剪不确定性评分（MUS）在五分之一计算量下达到0.784的故障检测AUC，而MC-Dropout为0.664（p<10^-6），且具有原生校准（SCE=0.049，对比ℓ1的0.127），是在AUC高于0.78的方法中校准最佳的。MUS与熵、置信度和ℓ1的监督融合达到0.832，优于五成员集成（0.813）。MUS随模型质量扩展，使用BiomedCLIP达到0.899（ρ=0.846），而该关系在分布内仍具意义（ρ=0.523），但在严重分布偏移下失效（VinBigData，ρ=0.027）。MUS适用于弥漫性病变，但对小结节等局灶性小病变的可靠性较低。代码和实验材料公开于该https URL。
英文摘要
Clinical vision models are often deployed as frozen black boxes with no access to internals, retraining, or ground truth at inference time. We introduce \textbf{SpatialUQ}, a post-hoc uncertainty method using only output probabilities. It measures the Jensen-Shannon divergence between the global prediction and the mean of five fixed spatial crops in six deterministic forward passes. The premise is simple, trustworthy predictions are spatially consistent. On NIH ChestX-ray14 (DenseNet-121, $N{=}25{,}596$), our Multicrop Uncertainty Score (MUS) reaches $0.784$ failure-detection AUC versus $0.664$ for MC-Dropout ($p{<}10^{-6}$) at one-fifth the compute, with native calibration ($\text{SCE}{=}0.049$ vs.\ $0.127$ for $\ell_1$), the best-calibrated among methods above 0.78 AUC. A supervised fusion of MUS with entropy, confidence, and $\ell_1$ reaches $0.832$, outperforming a five-member ensemble ($0.813$). MUS scales with model quality, reaching $0.899$ with BiomedCLIP ($ρ= 0.846$), while this relationship remains meaningful in-distribution ($ρ= 0.523$) but breaks down under severe distribution shift (VinBigData, $ρ= 0.027$). MUS is well-suited to diffuse findings but is less dependable for small focal lesions such as nodules. Code and experimental materials are publicly available at https://huggingface.co/datasets/kawsher11/SpatialUQ.
