Multi-Scanner Canine Cutaneous Squamous Cell Carcinoma Histopathology Dataset

本数据集名为‘Multi-Scanner Canine Cutaneous Squamous Cell Carcinoma Histopathology Dataset’，由埃尔朗根-纽伦堡大学模式识别实验室等机构创建。该数据集包含44个样本，每个样本通过五种不同的扫描系统数字化，旨在研究扫描系统间的域移问题。数据集提供了图像间的局部对应关系，有助于开发域无关算法。此外，数据集还附带了1,243个多边形注释，涵盖七种组织类型。数据集的应用领域主要集中在通过深度学习解决病理图像分析中的域移问题，提高模型的泛化能力。

This dataset, named 'Multi-Scanner Canine Cutaneous Squamous Cell Carcinoma Histopathology Dataset', was curated by the Pattern Recognition Lab of Friedrich-Alexander University Erlangen-Nuremberg (FAU) and other institutions. It contains 44 samples, each digitized using five distinct scanning systems, and is designed to investigate the domain shift problem across scanning platforms. The dataset provides local correspondence relationships between images, which aids in the development of domain-agnostic algorithms. Additionally, the dataset is accompanied by 1,243 polygonal annotations covering seven tissue types. Its primary application focuses on addressing domain shift issues in histopathological image analysis via deep learning, with the goal of enhancing the generalization capability of models.