显著性图在致密型乳腺钼靶上会失效吗？一项机会归一化、混杂因子感知的解释公平性审计
Do Saliency Maps Fail Dense Breasts? A Chance-Normalized, Confounder-Aware Audit of Explanation Fairness in Mammography
AI 导读
一项新审计在 1,345 个恶性 CBIS-DDSM 视图上检验了三种 CNN 的 Grad-CAM、Integrated Gradients 和遮挡图在致密型乳腺中的定位能力：Grad-CAM 峰值仅命中 44% 的致密病灶，低于非致密的 58%。
正文 · 原文
Abstract:Saliency maps, offered as evidence that mammography classifiers use the cancer, are untested in dense breasts, which can hide cancers. We tested whether they locate cancers worse there, and why. On 1,345 malignant CBIS-DDSM views, we scored three CNNs' Grad-CAM, Integrated Gradients and occlusion maps by whether their peak hits the lesion and a chance-normalized lesion-to-background saliency ratio (bits; +1 doubles it). Unadjusted gaps were large: Grad-CAM peaks hit 44% of dense vs. 58% of non-dense lesions. Case mix, chiefly more calcifications in dense breasts, explained about half or more of each gap. Occlusion kept a significant average of -0.56 bit gap. Controls also implicate the classifiers: unadjusted gaps exceeded an image-only conspicuity map's, and models relied less on dense-breast lesions. We recommend reporting density gaps before and after case-mix adjustment, per classifier, with conspicuity and lesion-reliance controls.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.05909 [cs.CV] |
| (or arXiv:2611.05909v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.05909 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ahmed Jaheen [view email]
[v1]
Wed, 4 Nov 2026 04:01:32 UTC (1,027 KB)
来源：arXiv cs.AI · arxiv.org