用患者级验证发现病理大模型信号主要来自细胞计数而非几何结构。
What Carries the Signal in Pathology Foundation-Model Atlases? A Patient-Level Controlled Benchmark in Breast Cancer
- 以患者为单位评估模型，发现细胞计数特征贡献最大。
- 嵌入向量预测分子程序得分达Spearman rho 0.25-0.56，显著优于随机。
- 几何结构实际未带来增益，因拓扑由欧氏距离决定。
病理基础模型据称能从组织形态中编码分子程序，但证据多为群体层面的基因排序列表，而非对独立患者的预测。本研究以患者为分析单位，重建该分析：在285例有配对切片与RNA-seq的乳腺癌患者（TCGA-BRCA）上，使用11个冻结骨干网络、4个预设基因程序和分组交叉验证（按患者分组），均在折叠内完成所有预处理。对均值池化嵌入进行岭回归，对保留患者程序得分的预测达到Spearman相关系数0.25–0.56，其中UNI2在所有四个程序中表现最优（免疫程序0.556）。匹配置换检验显示每细胞原始p≈1e-4（10,000次置换），校正后p=0.0044。信号真实存在但并非全由形态决定。在相同患者与分组下，嵌入向量优于组织成分指标，对ER/腔面、增殖和免疫程序分别提升+0.280、+0.284、+0.479（p≤0.003），但在基底型中，仅分区分数达0.469，嵌入得分为0.493（p=0.77）。54个可解释的细胞计数特征在所有程序上接近嵌入性能（差距0.043–0.085）。几何机制无显著贡献，因测地线图通过欧氏最近邻搜索选邻点，仅重加权已选定边，拓扑本质仍为欧氏（黎曼减欧氏=+0.0010，95%CI [-0.0007, +0.0029]）。一致应用几何结构反而更差（-0.0117）。岭回归优于图与度量解码器+0.097（95%CI [+0.069, +0.127]）。文献中常见的驱动基因计数指标在此任务中几乎无信息量：91.8%的随机六基因面板能恢复≥5/6个驱动基因。
原文摘要 · Abstract (English)
Pathology foundation models are reported to encode molecular programmes in tissue morphology, but the evidence is usually a cohort-wide ranked gene list rather than a prediction for a held-out patient. We rebuild such an analysis with the patient as the unit of evidence and ask which pipeline component carries signal. Across 11 frozen backbones, four pre-specified gene programmes and 285 TCGA-BRCA patients with paired slides and RNA-seq (44 cells; GroupKFold by patient, all preprocessing fitted inside the fold), ridge regression on mean-pooled embeddings predicts held-out programme scores at Spearman rho = 0.25-0.56, UNI2 strongest on all four (immune 0.556). A matched permutation null gives raw p ~ 1e-4 at 10,000 permutations for every cell; Holm-adjusted p = 0.0044. The signal is real but not uniformly morphological. Against competing models on the same patients and folds, embeddings beat tissue composition for ER/luminal, proliferation and immune (+0.280, +0.284, +0.479; p <= 0.003) but not basal, where compartment fractions alone reach 0.469 against the embedding's 0.493 (p = 0.77). Fifty-four interpretable cell-count features come within 0.043-0.085 on every programme. The geometric machinery contributes nothing measurable, and we identify why: the geodesic graph selects neighbours by Euclidean nearest-neighbour search and only reweights edges already chosen, so the topology is Euclidean by construction (Riemannian minus Euclidean = +0.0010, 95% CI [-0.0007, +0.0029]). Applied consistently the geometry is worse (-0.0117). Ridge regression beats the graph-and-metric decoder by +0.097 (CI [+0.069, +0.127]). The driver-count metric common in this literature is near-uninformative here: 91.8% of random six-gene panels recover >=5/6 drivers.
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。