EBHI

EBHI数据集是由东北大学MBIE学院显微图像与医学图像分析组创建，包含5532张电子显微镜图像，涵盖五种肿瘤分化阶段，包括正常、息肉、低级别上皮内瘤变、高级别上皮内瘤变和腺癌。数据集支持四种放大倍数（40×, 100×, 200×, 400×），用于评估计算机辅助诊断技术在结直肠癌早期检测中的应用。创建过程中，由两名中国医科大学肿瘤医院的病理学家提供图像，并由四名生物医学研究人员组织和生产数据集。EBHI旨在吸引研究人员探索新的分类算法，以帮助临床医生和患者进行自动化诊断。

The EBHI dataset was created by the Microscopic and Medical Image Analysis Group of the School of MBIE at Northeastern University. It contains 5532 electron microscopy images covering five tumor differentiation stages: normal, polyp, low-grade intraepithelial neoplasia, high-grade intraepithelial neoplasia, and adenocarcinoma. The dataset supports four magnification levels (40×, 100×, 200×, 400×) and is used to evaluate computer-aided diagnosis technologies for early detection of colorectal cancer. During its creation, the images were provided by two pathologists from the Cancer Hospital of China Medical University, and the dataset was organized and produced by four biomedical researchers. The EBHI dataset aims to attract researchers to explore novel classification algorithms to assist clinicians and patients in automated diagnosis.

搜集汇总

背景与挑战

背景概述

EBH-HE-IDS是一个包含5532张结直肠癌组织病理学电子显微镜图像的公开数据集，涵盖五种肿瘤分化阶段（正常、息肉、低度上皮内瘤变、高度上皮内瘤变和腺癌），并提供40×至400×四种放大倍数的图像。