完全注释的犬乳腺癌全切片图像数据集

本数据集名为‘完全注释的犬乳腺癌全切片图像数据集’，由德国弗里德里希-亚历山大-埃尔兰根-纽伦堡大学模式识别实验室和柏林自由大学兽医病理学研究所等机构共同创建。数据集包含21个全切片图像，全部注释了有丝分裂象（MF），总计13,907个MF和36,379个硬负例。创建过程中，通过专家病理学家的多次审查和机器学习的辅助，确保了注释的准确性和一致性。该数据集主要应用于提高人类乳腺癌诊断的准确性，特别是在有丝分裂计数方面，旨在解决现有数据集注释不完整和专家间不一致的问题。

This dataset is named "Fully Annotated Whole-Slide Image Dataset of Canine Breast Cancer". It was jointly developed by institutions including the Pattern Recognition Laboratory at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) in Germany and the Institute of Veterinary Pathology at the Free University of Berlin, as well as other relevant organizations. The dataset contains 21 whole-slide images, all annotated for mitotic figures (MF), with a total of 13,907 MF instances and 36,379 hard negatives. During its development, the accuracy and consistency of the annotations were ensured through multiple reviews by expert pathologists and assistance from machine learning techniques. This dataset is primarily intended to improve the accuracy of human breast cancer diagnosis, particularly in mitotic counting, aiming to address the problems of incomplete annotations and inter-expert inconsistency in existing datasets.