标题： 在跨数据集评估中对异常与正常分裂的深度学习和视觉基础模型进行基准测试 显示英文标题

标题： Benchmarking Deep Learning and Vision Foundation Models for Atypical vs. Normal Mitosis Classification with Cross-Dataset Evaluation

摘要： 非典型有丝分裂标志着细胞分裂过程中的偏差，已被证明是肿瘤恶性程度的独立预后标志。 然而，由于发生率低，有时与正常有丝分裂形态差异细微，病理学家之间的评分者间一致性低，以及数据集中的类别不平衡，非典型有丝分裂分类仍然具有挑战性。 基于乳腺癌非典型有丝分裂数据集（AMi-Br），本研究提出了一项全面的基准测试，比较了用于自动非典型有丝分裂图像（AMF）分类的深度学习方法，包括端到端训练的深度学习模型、使用线性探测的基础模型，以及通过低秩适应（LoRA）微调的基础模型。 为了进行严格的评估，我们进一步引入了两个新的保留AMF数据集——AtNorM-Br，这是一个来自TCGA乳腺癌队列的有丝分裂图像数据集，以及AtNorM-MD，这是一个来自MIDOG++训练集子集的多领域有丝分裂图像数据集。 我们在域内AMi-Br和域外AtNorm-Br和AtNorM-MD数据集上分别获得了平均平衡准确率值0.8135、0.7788和0.7723。 我们的工作表明，尽管非典型有丝分裂图像分类是一个具有挑战性的问题，但可以通过利用迁移学习和模型微调技术的最新进展来有效解决。 我们在此GitHub仓库中提供了本文中使用的所有代码和数据：https://github.com/DeepMicroscopy/AMi-Br_Benchmark。 显示英文摘要

摘要： Atypical mitosis marks a deviation in the cell division process that has been shown be an independent prognostic marker for tumor malignancy. However, atypical mitosis classification remains challenging due to low prevalence, at times subtle morphological differences from normal mitotic figures, low inter-rater agreement among pathologists, and class imbalance in datasets. Building on the Atypical Mitosis dataset for Breast Cancer (AMi-Br), this study presents a comprehensive benchmark comparing deep learning approaches for automated atypical mitotic figure (AMF) classification, including end-to-end trained deep learning models, foundation models with linear probing, and foundation models fine-tuned with low-rank adaptation (LoRA). For rigorous evaluation, we further introduce two new held-out AMF datasets - AtNorM-Br, a dataset of mitotic figures from the TCGA breast cancer cohort, and AtNorM-MD, a multi-domain dataset of mitotic figures from a subset of the MIDOG++ training set. We found average balanced accuracy values of up to 0.8135, 0.7788, and 0.7723 on the in-domain AMi-Br and the out-of-domain AtNorm-Br and AtNorM-MD datasets, respectively. Our work shows that atypical mitotic figure classification, while being a challenging problem, can be effectively addressed through the use of recent advances in transfer learning and model fine-tuning techniques. We make all code and data used in this paper available in this github repository: https://github.com/DeepMicroscopy/AMi-Br_Benchmark.