IGNITE数据工具包
收藏资源简介:
IGNITE数据工具包是一个多染料、多中心和多扫描仪的数据集,用于分析非小细胞肺癌的H&E和PD-L1免疫组化染色的全切片图像。该数据集包含来自155名独特患者的887个完全注释的兴趣区域,涵盖了三个互补的任务:多类语义分割、核检测和PD-L1阳性肿瘤细胞检测。IGNITE数据工具包旨在为开发深度学习模型提供资源,以改进非小细胞肺癌中的组织分割和细胞检测,并促进新型生物标志物的发展。
The IGNITE toolkit is a multi-dye, multi-center, and multi-scanner dataset for analyzing whole-slide images of hematoxylin and eosin (H&E) and PD-L1 immunohistochemically stained sections derived from non-small cell lung cancer (NSCLC) patients. It contains 887 fully annotated regions of interest (ROIs) from 155 unique patients, covering three complementary tasks: multi-class semantic segmentation, nucleus detection, and PD-L1 positive tumor cell detection. The IGNITE toolkit aims to provide a valuable resource for developing deep learning models to improve tissue segmentation and cell detection in NSCLC, and to promote the development of novel biomarkers.

- 1A tissue and cell-level annotated H&E and PD-L1 histopathology image dataset in non-small cell lung cancer拉德堡德大学医学中心病理学系, 荷兰 · 2025年



