NuCLS (Nucleus Classification, Localization and Segmentation)

NuCLS数据集包含来自TCGA的乳腺癌图像的220,000多个标记核。通过病理学家，病理学居民和医学生的共同努力，使用数字幻灯片存档对这些核进行了注释。这些数据可以以多种方式用于开发和验证用于核检测，分类和分割的算法，也可以用作开发和评估间分析方法的资源。
提供了来自单评分者和多评分者研究的数据。对于单评分者数据，我们提供了病理学家审查和未更正的注释。对于多评分者数据集，我们提供在有弱分割和分类算法的建议和没有建议的情况下生成的注释。

NuCLS dataset contains over 220,000 labeled nuclei from breast cancer images derived from The Cancer Genome Atlas (TCGA). These nuclei were annotated using digital slide archives through the collaborative efforts of pathologists, pathology residents, and medical students. This dataset can be utilized to develop and validate algorithms for nuclear detection, classification, and segmentation in multiple scenarios, and can also serve as a valuable resource for developing and evaluating inter-rater analysis methods.

Data from both single-rater and multi-rater studies are provided. For the single-rater dataset, we provide both pathologist-reviewed and uncorrected annotations. For the multi-rater dataset, we offer annotations generated both with and without the guidance of weak segmentation and classification algorithms.

搜集汇总

背景与挑战

背景概述

NuCLS数据集是一个包含220,000多个乳腺癌图像标记核的生物医学数据集，用于核检测、分类和分割算法的开发与验证。数据集提供了单评分者和多评分者的注释数据，适用于多种研究和评估场景。