让病理切片分析更高效：仅用少量关键区域即可准确判断癌症类型。
ReaMIL: Reasoning- and Evidence-Aware Multiple Instance Learning for Whole-Slide Histopathology
- 通过轻量级门控头筛选关键组织块，实现精准定位。
- 在肺癌数据集上仅需8.2个切片块即达0.983的准确率，且置信度快速上升。
- 无需额外标注，自动输出可解释的可视化热图，适合临床辅助决策。
我们提出ReaMIL（推理与证据感知多实例学习），一种用于全切片病理图像的多实例学习方法。该方法在强大MIL主干网络上添加轻量级选择头，生成软性每块门控，并采用带预算的充分性目标进行训练：在保留切片数量受限的前提下，确保真实类别概率达到τ ≥ 0.90。该目标使所选证据集小而空间紧凑，同时不牺牲基线性能。在TCGA-NSCLC（LUAD vs. LUSC）、TCGA-BRCA（IDC vs. Others）和PANDA数据集上，ReaMIL匹配或略微提升基线AUC，提供量化证据效率诊断。在NSCLC任务中，于τ=0.90时达到AUC 0.983，平均最小充分数MSK≈8.2，AUKC≈0.864，表明一旦保留少量关键切片，分类置信度迅速上升并趋于稳定。方法无需额外监督，可无缝集成标准MIL训练流程，并自然生成滑片级可视化叠加图。我们报告准确率、MSK、AUKC及连通性指标，以严格评估模型在全切片图像上的行为表现。
原文摘要 · Abstract (English)
We introduce ReaMIL (Reasoning- and Evidence-Aware MIL), a multiple instance learning approach for whole-slide histopathology that adds a light selection head to a strong MIL backbone. The head produces soft per-tile gates and is trained with a budgeted-sufficiency objective: a hinge loss that enforces the true-class probability to be $\geq τ$ using only the kept evidence, under a sparsity budget on the number of selected tiles. The budgeted-sufficiency objective yields small, spatially compact evidence sets without sacrificing baseline performance. Across TCGA-NSCLC (LUAD vs. LUSC), TCGA-BRCA (IDC vs. Others), and PANDA, ReaMIL matches or slightly improves baseline AUC and provides quantitative evidence-efficiency diagnostics. On NSCLC, it attains AUC 0.983 with a mean minimal sufficient K (MSK) $\approx 8.2$ tiles at $τ= 0.90$ and AUKC $\approx 0.864$, showing that class confidence rises sharply and stabilizes once a small set of tiles is kept. The method requires no extra supervision, integrates seamlessly with standard MIL training, and naturally yields slide-level overlays. We report accuracy alongside MSK, AUKC, and contiguity for rigorous evaluation of model behavior on WSIs.
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。