NCI SEER-LINKED PEDIATRIC WHOLE SLIDE IMAGING (POP: 8/17/2020 - 8/16/2021)

NCI SEER 关联儿科全切片成像（POP：2020 年 8 月 17 日 - 2021 年 8 月 16 日）

• 批准号: 10272821
• 负责人: THOMAS TUCKER
• 金额: $ 20.36万
• 依托单位: UNIVERSITY OF KENTUCKY
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-08-17 至 2021-08-16
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10272821
• 关键词: Cancer Patient; Cancer Surveillance Research Program; Childhood; Computer software; Development; Diagnosis; Goals; Hematoxylin and Eosin Staining Method; Histologic; Image; Informatics; Kentucky; Logistics; Machine Learning; Malignant Childhood Neoplasm; Metadata; Modeling; National Cancer Institute; Nuclear; Operative Surgical Procedures; Pathology; Research Institute; Resources; Slide; Technology; Testing; Tissue Banks; Tumor-Infiltrating Lymphocytes; cancer diagnosis; digital; experience; microscopic imaging; neoplasm registry; patient population; tumor; virtual; whole slide imaging

Project Abstract: The development of high-quality digital scanners has made it possible to produce detailed images of hematoxylin and eosin (H&E) slides from cancer patient histological material. Emerging new informatics technologies also make it possible to develop machine learned models to extract additional important features from the digitized slides such as tumor infiltrating lymphocytes, delineating tumor regions and identifying nuclear segmentation. The Kentucky Cancer Registry’s Virtual Tissue Repository (VTR) has experience obtaining the H&E slides for specific patient populations from pathology labs across the state, having these slides digitized, submitting the anonymized digitized images to the National Cancer Institute (NCI) for research, and returning the original slides to the contributing pathology labs.

项目摘要：高质量数字扫描仪的发展使得从癌症患者的组织学材料中产生苏木精和伊红（H&E）玻片的详细图像成为可能。新兴的新信息学技术也使开发机器学习模型成为可能，以从数字化幻灯片中提取其他重要特征，如肿瘤浸润淋巴细胞，描绘肿瘤区域和识别核分割。肯塔基州癌症登记处的虚拟组织库（VTR）拥有从全州病理实验室获取特定患者群体的H&E玻片的经验，将这些玻片数字化，将匿名的数字化图像提交给国家癌症研究所（NCI）进行研究，并将原始玻片返回给贡献的病理实验室。