warwick

腺体是重要的组织学结构，存在于大多数器官系统中，是分泌蛋白质和碳水化合物的主要机制。研究表明，由腺上皮引起的恶性肿瘤 (也称为腺癌) 是最普遍的癌症形式。病理学家通常使用腺体的形态来评估几种腺癌的恶性程度，包括前列腺，乳腺，肺和结肠。
腺体的准确分割通常是获得可靠的形态学统计数据的关键步骤。尽管如此，由于不同组织学等级的腺体形态变化很大，因此这项任务本质上非常具有挑战性。到目前为止，大多数研究都集中在健康或良性样本中的腺体分割上，但很少针对中等或高级别癌症，而且通常，它们针对特定数据集进行了优化。
在此挑战中，鼓励参与者在苏木精和曙红 (H & E) 染色的幻灯片图像上运行腺体分割算法，这些幻灯片由各种组织学等级组成。该数据集与专业病理学家提供的基本事实注释一起提供。要求参与者在提供的训练数据集上开发和优化他们的算法，并在测试数据集上验证他们的算法。

Glands are critical histological structures present in most organ systems, serving as the primary mechanism for secreting proteins and carbohydrates. Studies have shown that malignant tumors arising from glandular epithelium (also known as adenocarcinomas) are the most prevalent form of cancer. Pathologists typically use glandular morphology to assess the malignancy of several adenocarcinoma types, including those of the prostate, breast, lung, and colon.

Accurate segmentation of glands is often a critical step in obtaining reliable morphometric statistics. Nevertheless, this task is inherently highly challenging due to the substantial morphological variations of glands across different histological grades. To date, most studies have focused on gland segmentation in healthy or benign specimens, with few targeting intermediate or high-grade cancers, and they are often optimized for specific datasets.

In this challenge, participants are encouraged to run gland segmentation algorithms on hematoxylin and eosin (H&E) stained slide images, which encompass a range of histological grades. This dataset is provided alongside ground truth annotations generated by expert pathologists. Participants are required to develop and optimize their algorithms on the provided training dataset, and validate their performance on the test dataset.

搜集汇总

背景与挑战

背景概述

warwick数据集由华威大学于2015年发布，专注于腺体分割算法的开发与优化，提供专业病理学家的注释，用于评估腺癌的恶性程度。数据集包含苏木精和曙红染色的幻灯片图像，旨在支持中等或高级别癌症的研究。