MHIST (Minimalist Histopathology image analysis dataset)

该数据集包括3,152来自达特茅斯-希区柯克医学中心 (DHMC) 病理和实验室医学系的大肠息肉的苏木精和曙红 (H & E) 染色的福尔马林固定石蜡包埋 (FFPE) 固定尺寸图像 (按224像素224)。经达特茅斯-希区柯克健康 (d-hh) 机构审查委员会 (IRB) 许可，该数据集被取消身份并发布。所有图像均根据七位病理学家drs的意见进行标记。DHMC病理和实验室医学系的Arief suriainata，Bing Ren，Xiaoying Liu，Mikhail Lisovsky，Louis Vaickus，Charles Brown和Michael Baker，用于结直肠息肉的类型。该数据集及其相关注释旨在促进与研究界的合作，并促进开发和评估用于数字病理学中准确组织学图像分析的新方法。有关此数据集的更多信息，请参阅 “用于组织病理学图像分析的培养皿”。

This dataset comprises 3,152 fixed-size (224×224 pixels) hematoxylin and eosin (H&E)-stained formalin-fixed paraffin-embedded (FFPE) images of colorectal polyps sourced from the Department of Pathology and Laboratory Medicine at Dartmouth-Hitchcock Medical Center (DHMC). This dataset was de-identified and released with approval from the Institutional Review Board (IRB) of Dartmouth-Hitchcock Health (D-HH). All images were labeled based on the consensus of seven pathologists from the Department of Pathology and Laboratory Medicine at DHMC: Drs. Arief Suriainata, Bing Ren, Xiaoying Liu, Mikhail Lisovsky, Louis Vaickus, Charles Brown, and Michael Baker, for the classification of colorectal polyps. This dataset and its associated annotations are designed to facilitate collaboration with the research community and advance the development and evaluation of innovative approaches for accurate histopathological image analysis in digital pathology. For additional information about this dataset, please refer to "Cultivation Plate for Histopathological Image Analysis".

搜集汇总

背景与挑战

背景概述

MHIST数据集包含3,152张固定尺寸的大肠息肉病理图像，由专业病理学家标注，用于数字病理学图像分析研究。数据集发布于2021年，旨在促进该领域的方法开发与评估。