100,000 Histological Images Of Human Colorectal Cancer And Healthy Tissue

<strong>Data Description "NCT-CRC-HE-100K"</strong> This is a set of 100,000 non-overlapping image patches from hematoxylin &amp; eosin (H&amp;E) stained histological images of human colorectal cancer (CRC) and normal tissue. All images are 224x224 pixels (px) at 0.5 microns per pixel (MPP). All images are color-normalized using Macenko's method (http://ieeexplore.ieee.org/abstract/document/5193250/, DOI 10.1109/ISBI.2009.5193250). Tissue classes are: Adipose (ADI), background (BACK), debris (DEB), lymphocytes (LYM), mucus (MUC), smooth muscle (MUS), normal colon mucosa (NORM), cancer-associated stroma (STR), colorectal adenocarcinoma epithelium (TUM). These images were manually extracted from N=86 H&amp;E stained human cancer tissue slides from formalin-fixed paraffin-embedded (FFPE) samples from the NCT Biobank (National Center for Tumor Diseases, Heidelberg, Germany) and the UMM pathology archive (University Medical Center Mannheim, Mannheim, Germany). Tissue samples contained CRC primary tumor slides and tumor tissue from CRC liver metastases; normal tissue classes were augmented with non-tumorous regions from gastrectomy specimen to increase variability. <strong>Ethics statement "NCT-CRC-HE-100K"</strong> All experiments were conducted in accordance with the Declaration of Helsinki, the International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS), the Belmont Report and the U.S. Common Rule. Anonymized archival tissue samples were retrieved from the tissue bank of the National Center for Tumor diseases (NCT, Heidelberg, Germany) in accordance with the regulations of the tissue bank and the approval of the ethics committee of Heidelberg University (tissue bank decision numbers 2152 and 2154, granted to Niels Halama and Jakob Nikolas Kather; informed consent was obtained from all patients as part of the NCT tissue bank protocol, ethics board approval S-207/2005, renewed on 20 Dec 2017). Another set of tissue samples was provided by the pathology archive at UMM (University Medical Center Mannheim, Heidelberg University, Mannheim, Germany) after approval by the institutional ethics board (Ethics Board II at University Medical Center Mannheim, decision number 2017-806R-MA, granted to Alexander Marx and waiving the need for informed consent for this retrospective and fully anonymized analysis of archival samples). <strong>Data set "CRC-VAL-HE-7K"</strong> This is a set of 7180 image patches from N=50 patients with colorectal adenocarcinoma (no overlap with patients in NCT-CRC-HE-100K). It can be used as a validation set for models trained on the larger data set. Like in the larger data set, images are 224x224 px at 0.5 MPP. All tissue samples were provided by the NCT tissue bank, see above for further details and ethics statement. <strong>Data set "NCT-CRC-HE-100K-NONORM"</strong> This is a slightly different version of the "NCT-CRC-HE-100K" image set: This set contains 100,000 images in 9 tissue classes at 0.5 MPP and was created from the same raw data as "NCT-CRC-HE-100K". However, no color normalization was applied to these images. Consequently, staining intensity and color slightly varies between the images. Please note that although this image set was created from the same data as "NCT-CRC-HE-100K", the image regions are not completely identical because the selection of non-overlapping tiles from raw images was a stochastic process. <strong>General comments</strong> Please note that the classes are only roughly balanced. Classifiers should never be evaluated based on accuracy in the full set alone. Also, if a high risk of training bias is excepted, balancing the number of cases per class is recommended.<strong>数据描述 “NCT-CRC-HE-100K”</strong> 该数据集包含100,000张无重叠图像块，均提取自人类结直肠癌（colorectal cancer, CRC）及正常组织的苏木精-伊红（hematoxylin & eosin, H&E）染色组织学图像。 所有图像分辨率为224×224像素（px），采样倍率为0.5微米每像素（microns per pixel, MPP）。所有图像均采用Macenko方法进行色彩标准化（引用文献：http://ieeexplore.ieee.org/abstract/document/5193250/，DOI: 10.1109/ISBI.2009.5193250）。 组织类别包括：脂肪组织（Adipose, ADI）、背景（Background, BACK）、碎屑（Debris, DEB）、淋巴细胞（Lymphocytes, LYM）、黏液（Mucus, MUC）、平滑肌（Smooth muscle, MUS）、正常结肠黏膜（Normal colon mucosa, NORM）、癌相关间质（Cancer-associated stroma, STR）、结直肠腺癌上皮（Colorectal adenocarcinoma epithelium, TUM）。 这些图像均手动从86例H&E染色的人类癌组织切片中提取，切片样本来自德国海德堡国家肿瘤疾病中心（National Center for Tumor Diseases, NCT）生物样本库以及德国曼海姆大学医学中心（University Medical Center Mannheim, UMM）病理档案的福尔马林固定石蜡包埋（formalin-fixed paraffin-embedded, FFPE）样本。组织样本涵盖结直肠癌原发肿瘤切片以及结直肠癌肝转移癌组织；正常组织类别通过补充胃切除术标本的非肿瘤区域以提升数据多样性。 <strong>伦理声明 “NCT-CRC-HE-100K”</strong> 所有实验均符合《赫尔辛基宣言》、涉及人类受试者的生物医学研究国际伦理指南（CIOMS）、《贝尔蒙报告》以及美国联邦《共同规则》。 匿名存档组织样本的获取符合德国海德堡国家肿瘤疾病中心生物样本库的相关规定，并获得海德堡大学伦理委员会的批准（样本库决策编号：2152与2154，授予Niels Halama与Jakob Nikolas Kather；所有患者均已按照NCT生物样本库方案签署知情同意书，伦理委员会批准编号S-207/2005，并于2017年12月20日更新）。 另有一组组织样本由德国曼海姆大学医学中心（海德堡大学附属）病理档案提供，该获取流程已获得机构伦理委员会批准（曼海姆大学医学中心第二伦理委员会，决策编号2017-806R-MA，授予Alexander Marx，且对于该存档样本的回顾性且完全匿名的分析豁免知情同意要求）。 <strong>数据集 “CRC-VAL-HE-7K”</strong> 该数据集包含7180张图像块，来自50例结直肠腺癌患者（与NCT-CRC-HE-100K数据集中的患者无重叠），可作为基于更大规模数据集训练的模型的验证集。与前述大规模数据集一致，所有图像分辨率为224×224像素，采样倍率为0.5微米每像素（microns per pixel, MPP）。所有组织样本均来自NCT生物样本库，详细信息与伦理声明参见前文。 <strong>数据集 “NCT-CRC-HE-100K-NONORM”</strong> 该数据集为“NCT-CRC-HE-100K”的变体版本：其包含100,000张图像，涵盖9种组织类别，采样倍率为0.5微米每像素（microns per pixel, MPP），且与“NCT-CRC-HE-100K”采用相同的原始数据构建。但该数据集未应用色彩标准化步骤，因此不同图像间的染色强度与色彩存在细微差异。请注意，尽管该数据集与“NCT-CRC-HE-100K”源自同一批原始数据，但图像区域并非完全一致，这是因为从原始图像中选取无重叠图像块的过程为随机过程。 <strong>通用说明</strong> 请注意，各组织类别的样本数量仅大致均衡。切勿仅基于全数据集的准确率对分类器进行评估。此外，若存在较高的训练偏差风险，建议对每类样本的数量进行均衡处理。