COVIDGR-1.0

COVIDGR-1.0是由安达卢西亚数据科学与计算智能研究所与格拉纳达的圣塞西利奥临床大学医院合作创建的高质量CXR图像数据集。该数据集包含426张阳性（COVID-19）和426张阴性（非COVID-19）的PA视图CXR图像，涵盖从正常到严重的所有疾病严重程度。数据集的创建遵循统一的临床协议，并由专家放射科医生进行标注。COVIDGR-1.0旨在支持开发用于COVID-19早期检测的分类模型，特别关注提高模型对中度和重度病例的识别能力。

COVIDGR-1.0 is a high-quality CXR image dataset developed in collaboration between the Andalusian Institute of Data Science and Computational Intelligence and the San Cecilio University Clinical Hospital in Granada. This dataset contains 426 positive (COVID-19) and 426 negative (non-COVID-19) PA-view CXR images, covering all disease severity levels from normal to severe. It was created following a unified clinical protocol and annotated by expert radiologists. COVIDGR-1.0 aims to support the development of classification models for early COVID-19 detection, with a particular focus on improving the models' ability to identify moderate and severe cases.