Kimia Path24

Kimia Path24数据集是由滑铁卢大学的KIMIA Lab创建，用于数字病理图像的分类和检索。该数据集包含24种不同组织纹理的全扫描图像，生成了1,325个大小为1000×1000的测试补丁。数据集的设计模拟了临床实践中的检索任务，用户可以根据需要创建从27,000到超过50,000个补丁的训练集。Kimia Path24数据集旨在通过机器学习算法支持更可靠的诊断决策，解决数字病理图像处理中的存储、传输和实时处理难题。

Kimia Path24 Dataset was developed by the KIMIA Lab at the University of Waterloo for digital pathology image classification and retrieval. This dataset encompasses whole-slide images of 24 distinct tissue textures, and includes 1,325 test patches each with a size of 1000×1000. Designed to simulate retrieval tasks encountered in clinical practice, the dataset enables users to construct training sets ranging from 27,000 to more than 50,000 patches according to specific requirements. The Kimia Path24 Dataset aims to support more reliable diagnostic decision-making via machine learning algorithms, while addressing the challenges in storage, transmission and real-time processing associated with digital pathology image handling.