COVIDx CXR-3
收藏资源简介:
COVIDx CXR-3是由滑铁卢大学视觉与图像处理实验室创建的大型开放源代码基准数据集,包含30,386张来自17,026名患者的胸部X光图像,覆盖至少51个国家,旨在支持COVID-19计算机视觉研究。该数据集通过与专家放射科医生和临床医生的合作,精心筛选,包括不同年龄、性别和影像视图的多样性,以减少偏见和数据集不平衡。COVIDx CXR-3的应用领域主要集中在通过深度学习驱动的计算机辅助诊断,以检测和评估COVID-19病例的严重性,从而在疫情期间提供有效的筛查和患者分流策略。
COVIDx CXR-3 is a large open-source benchmark dataset created by the Vision and Image Processing Lab at the University of Waterloo. It encompasses 30,386 chest X-ray images from 17,026 patients across at least 51 countries, with the goal of supporting computer vision research focused on COVID-19. This dataset was carefully curated through collaboration with expert radiologists and clinicians, incorporating diversity in age, gender and radiographic views to reduce bias and address dataset imbalance. The primary applications of COVIDx CXR-3 center on deep learning-driven computer-aided diagnosis for detecting COVID-19 cases and evaluating their severity, thereby providing effective screening and patient triage strategies during the COVID-19 pandemic.

- 1COVIDx CXR-3: A Large-Scale, Open-Source Benchmark Dataset of Chest X-ray Images for Computer-Aided COVID-19 Diagnostics滑铁卢大学 · 2022年



