arXiv 2409.013302025-08-13cs.CVcs.AI

使用数字组织病理学和深度学习进行儿童脑肿瘤分类：在多中心瑞典队列上评估SOTA方法

Pediatric brain tumor classification using digital histopathology and deep learning: evaluation of SOTA methods on a multi-center Swedish cohort

Department of Biomedical Engineering, Linköping University, Sweden（生物医学工程系，利厄普大学）；Center for Medical Image Science and Visualization, Linköping University, Sweden（医学影像科学与可视化中心，利厄普大学）；Crown Princess Victoria Children’s Hospital and Department of Health, Medicine and Caring Sciences, Linköping University, Sweden（维多利亚公主儿童医院和健康、医学与关怀科学系，利厄普大学）；Department of Radiology and Department of Health, Medicine and Caring Sciences, Linköping University, Sweden（放射科和健康、医学与关怀科学系，利厄普大学）；Department of Oncology-Pathology, Karolinska Institutet, Solna, Sweden（肿瘤病理学系，卡罗林斯卡研究所，索纳）；Department of Clinical Pathology and Cancer Diagnostics, Karolinska University Hospital, Sweden（临床病理学和癌症诊断系，卡罗林斯卡大学医院）；Department of Radiation Physics and Department of Medical and Health Sciences, Linköping University, Sweden（放射物理系和医学与健康科学系，利厄普大学）

Iulian Emil Tampu, Per Nyman, Christoforos Spyretos, Ida Blystad, Alia Shamikh, Gabriela Prochazka, Teresita Díaz de Ståhl, Johanna Sandgren, Peter Lundberg, Neda Haj-Hosseini

发表信息：: Tampu IE et al. Pediatric brain tumor classification using digital pathology and deep learning: Evaluation of SOTA methods on a multi-center Swedish cohort. Brain Pathology. 2025. e70029
DOI：: 10.1111/bpa.70029

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本研究在多中心瑞典队列上，利用UNI等基础模型提取特征并结合ABMIL等弱监督多示例学习方法对儿童脑肿瘤WSIs进行分层分类，发现UNI特征表现最佳且模型具备良好的跨中心泛化能力。

摘要：

脑肿瘤是儿童和年轻成人中最常见的实体瘤，但大型组织病理学数据集的匮乏限制了计算病理学在这一群体中的应用。本研究在从最先进的组织病理学基础模型提取的图像块特征上，实现了两种弱监督多示例学习（MIL）方法，以对来自多中心瑞典队列的苏木精-伊红全视野数字切片（WSIs）中的儿童脑肿瘤进行分类。研究收集了来自瑞典六家大学医院540名受试者（年龄8.5±4.9岁）的脑肿瘤WSIs。使用三种预训练特征提取器从WSIs中获取实例（图像块）级特征：ResNet50、UNI和CONCH。使用基于注意力的MIL（ABMIL）或聚类约束注意力MIL（CLAM）聚合实例，进行患者级分类。基于儿童脑肿瘤的分层分类，在三个分类任务上评估模型：肿瘤类别、肿瘤家族和肿瘤类型。通过在两个中心的数据上训练并在其他四个中心的数据上测试，来评估模型的泛化能力。通过注意力映射评估模型的可解释性。使用UNI特征和ABMIL聚合取得了最高的分类性能，在肿瘤类别、家族和类型分类任务上的Matthew's相关系数分别为0.76±0.04、0.63±0.04和0.60±0.05。在评估泛化能力时，使用UNI和CONCH特征的模型优于使用ResNet50的模型。然而，从中心内测试到中心外测试的性能下降幅度在不同特征提取器间是相似的。这些结果表明，最先进的计算病理学方法在诊断不同分层级别的儿童脑肿瘤方面具有潜力，并在多中心国家数据集上展现出良好的泛化能力。

英文摘要：

Brain tumors are the most common solid tumors in children and young adults, but the scarcity of large histopathology datasets has limited the application of computational pathology in this group. This study implements two weakly supervised multiple-instance learning (MIL) approaches on patch-features obtained from state-of-the-art histology-specific foundation models to classify pediatric brain tumors in hematoxylin and eosin whole slide images (WSIs) from a multi-center Swedish cohort. WSIs from 540 subjects (age 8.5$\pm$4.9 years) diagnosed with brain tumor were gathered from the six Swedish university hospitals. Instance (patch)-level features were obtained from WSIs using three pre-trained feature extractors: ResNet50, UNI, and CONCH. Instances were aggregated using attention-based MIL (ABMIL) or clustering-constrained attention MIL (CLAM) for patient-level classification. Models were evaluated on three classification tasks based on the hierarchical classification of pediatric brain tumors: tumor category, family, and type. Model generalization was assessed by training on data from two of the centers and testing on data from four other centers. Model interpretability was evaluated through attention mapping. The highest classification performance was achieved using UNI features and ABMIL aggregation, with Matthew's correlation coefficient of 0.76$\pm$0.04, 0.63$\pm$0.04, and 0.60$\pm$0.05 for tumor category, family, and type classification, respectively. When evaluating generalization, models utilizing UNI and CONCH features outperformed those using ResNet50. However, the drop in performance from the in-site to out-of-site testing was similar across feature extractors. These results show the potential of state-of-the-art computational pathology methods in diagnosing pediatric brain tumors at different hierarchical levels with fair generalizability on a multi-center national dataset.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。