发表机构
Duke University（杜克大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对细胞模拟导致的病理细胞分类歧义问题，提出Loki-OT方法，结合区域级组织推理与非平衡最优传输，在TCGA-BRCA队列上取得更优性能。
AI 中文摘要
细胞模拟是指不同细胞类型在形态上表现相似的现象，人类病理学家会借助周围组织上下文解决这种歧义，而当前视觉模型要么缺乏上下文推理能力（细胞基础模型），要么无法在细胞层面运行（病理多模态大语言模型MLLM）。本文提出Loki-OT，该方法通过非平衡最优传输将区域级组织推理传播至单个细胞预测，利用MLLM生成的密度先验作为模糊细胞重分配的软引导。Loki-OT的提出基于如下观察：预训练细胞基础模型的特征已编码包括组织上下文在内的判别信息，但标准细胞级监督无法有效利用组织上下文。生成的传输计划被蒸馏为轻量级学生多层感知机（MLP）分类器，该分类器在预训练特征空间中学习上下文感知的决策边界。在独立的TCGA-BRCA队列上，Loki-OT使用基于通用域细胞基础模型构建的278个弱区域级MLLM估计，取得了比全监督域内PanopTILs分类器更低的患者级平均绝对误差（MAE），并提升了富含上皮的模拟组织的F1值。
英文摘要
Cell mimicry arises when different cell types appear morphologically similar. Human pathologists resolve this ambiguity using surrounding tissue context, whereas current vision models either lack contextual reasoning (cell foundation models) or cannot operate at the cell level (pathology MLLMs). We present Loki-OT, which propagates region-level tissue reasoning to individual cell predictions via Unbalanced Optimal Transport, using MLLM-derived density priors as soft guidance for ambiguous cell reassignment. Loki-OT is motivated by the observation that pretrained cell foundation model features already encode discriminative information, including tissue context, but standard cell-level supervision fails to use tissue context effectively. The resulting transport plan is distilled into a lightweight student MLP classifier that learns context-aware decision boundaries within the pretrained feature space. On the independent TCGA-BRCA cohort, Loki-OT achieved lower patient-level MAE than the fully supervised in-domain PanopTILs classifier and improved F1 in epithelium-rich mimicry tissues, using 278 weak region-level MLLM estimates built on a general-domain cell foundation model. Code: https://github.com/xiangli980/Lymphocyte_Mimicry_Correction_via_Loki_OT
Comments13 pages, 3 figures. Accepted to the MICCAI 2026 COMPAYL Workshop
