COVIDx CT-3
收藏资源简介:
COVIDx CT-3是由滑铁卢大学创建的一个大规模跨国开放源胸部CT图像数据集,旨在辅助COVID-19的计算机辅助筛查。该数据集包含来自至少17个国家的6,068名患者的431,205张CT切片,是目前最大的开放访问COVID-19 CT图像数据集。数据集通过多种CT扫描仪和协议从全球多个数据源精心处理和策划而成。COVIDx CT-3主要用于支持机器学习驱动的筛查系统开发,帮助临床医生区分COVID-19肺炎与其他病因的肺炎。尽管数据集在多样性上有所努力,但仍存在显著的地理和类别不平衡问题。
COVIDx CT-3 is a large-scale multinational open-source chest CT image dataset created by the University of Waterloo, aiming to assist computer-aided screening of COVID-19. It contains 431,205 CT slices from 6,068 patients across at least 17 countries, making it the largest publicly accessible COVID-19 CT image dataset to date. The dataset was meticulously processed and curated from multiple global data sources using a diverse range of CT scanners and imaging protocols. Primarily intended to support the development of machine learning-driven screening systems, it helps clinicians distinguish COVID-19 pneumonia from pneumonia caused by other etiologies. Despite efforts to improve diversity, the dataset still exhibits significant geographic and class imbalance issues.




