CytoFormer: A Molecularly Supervised Cell Foundation Model for Histopathology Cell Classification

📄 arXiv: 2608.16718v1 📥 PDF

作者: Jialu Yao, Songhao Li, Alina Yu, Zhi Huang

分类: cs.CV

发布日期: 2026-08-17

备注: 20 pages, 5 figures, 2 extended data figures

💡 一句话要点

提出CytoFormer以解决细胞分类中的标注效率问题

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 细胞分类 空间转录组学 分子监督 主动学习 病理学

📋 核心要点

  1. 现有细胞分类方法依赖人工标注，效率低且不可靠，限制了大规模单细胞分析的应用。
  2. 本文提出CytoFormer模型，通过分子监督形态，结合空间转录组学与H&E图像，实现高效的细胞分类。
  3. CytoFormer在16个器官的测试中达到85%的准确率，且在主动学习设置中显著提高了标注效率，F1分数提升0.13。

📝 摘要（中文）

直接从常规的苏木精-伊红（H&E）组织学中识别细胞类型可以实现大规模的单细胞分析，但现有模型训练依赖于人工病理学家的标注，速度慢、成本高且对许多细胞类型不可靠。本文提出了一种通过分子监督形态的方法，利用成像基础的空间转录组学在切片上对单个细胞进行分析，并将分子身份与形态观察结合。我们组装了81个配对的Xenium切片，涵盖16个器官，生成了1540万个细胞的标注数据，训练了CytoFormer模型。该模型在空间保留的组织上达到了0.85的准确率和0.78的宏F1分数，且在交互式主动学习环境中表现出更高的标注效率。

🔬 方法详解

问题定义：本文旨在解决传统细胞分类方法中对人工标注的依赖，导致的效率低下和不可靠性问题。现有方法在处理多种细胞类型时面临挑战，限制了大规模应用。

核心思路：论文提出通过分子监督形态的方法，利用空间转录组学技术获取细胞的分子特征与形态信息，从而实现更高效的细胞分类。这样的设计使得模型能够在同一物理细胞上同时观察到分子身份和形态特征。

技术框架：整体架构包括数据采集、细胞标注、模型训练和评估四个主要阶段。首先，通过81个配对的Xenium切片获取1540万个细胞数据；其次，利用聚类和标记基因注释生成细胞标签；最后，训练CytoFormer模型并进行性能评估。

关键创新：CytoFormer的创新在于其通过分子与形态的结合，提供了一种新的细胞级分析表示，显著提高了标注效率和分类准确性。这与传统依赖人工标注的方法形成了本质区别。

关键设计：模型采用多任务学习框架，针对每个器官设计分类头，损失函数结合了分类损失与重建损失，确保模型在多种细胞类型上均能有效学习。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

CytoFormer在16个器官的空间保留组织测试中达到了85%的准确率和78%的宏F1分数，且在主动学习环境中，F1分数达到0.82，仅需少量标注即可超越现有基线0.13，显示出显著的标注效率提升。

🎯 应用场景

该研究的潜在应用领域包括医学诊断、病理学研究和生物医学数据分析。CytoFormer模型能够在临床环境中快速、准确地进行细胞分类，降低人工标注的需求，从而提高病理分析的效率和可靠性。未来，该模型可能推动个性化医疗和精准医学的发展。

📄 摘要（原文）

Identifying cell types directly from routine haematoxylin and eosin (H&E) histology would enable single-cell analysis at scale, but training such models has relied on manual pathologist annotations, which are slow, expensive and unreliable for many cell types. We instead supervise morphology with molecules. Imaging-based spatial transcriptomics profiles individual cells in situ on a section that can afterwards be stained with H&E, so that molecular identity and morphology are observed for the same physical cell. We assembled 81 such paired Xenium sections spanning 16 organs, derived per-cell labels by clustering, marker-gene annotation, organ-wise human review and quality control, and mapped them onto the cell types commonly reported in each organ. This yielded 15.4 million cells, each with a paired H&E image patch and one of 23 cell types, on which we trained CytoFormer, a cell foundation model with a multi-task, per-organ classification head. On spatially held-out tissue CytoFormer reached an accuracy of 0.85 and a macro-F1 of 0.78 across all 16 organs, and its predictions reproduced the tissue architecture of an entire held-out section. The representation also transfers: with the encoder frozen, a linear head on CytoFormer features performed better than six pathology foundation models on four expert-annotated benchmarks, including on organs and cell types that were not part of pretraining. Finally, in an interactive active-learning setting, CytoFormer's embeddings are markedly more label-efficient than existing pathology foundation models, detecting normal epithelium amid look-alike tumour with an F1 of 0.82 from only a few annotations and leading the strongest baseline by 0.13 in F1. CytoFormer turns paired H&E and spatial transcriptomics into a reusable, label-efficient representation for cell-level analysis of routine histology.