peternasser99/MED

PADCHEST数据集是一个包含160K高分辨率胸部X光图像及其对应标注报告的数据集。图像分布在54个压缩文件中，总大小为1TB。标注信息包括图像ID、患者信息、DICOM标准字段、投影信息、报告文本、标签、定位信息等。数据集还提供了标签和定位的UMLS Metathesaurus CUIs。标签分为三类：鉴别诊断、放射学发现和解剖位置。数据集的使用和检索方法也在README中进行了说明。

The PADCHEST dataset is a labeled dataset containing 160K high resolution chest x-ray images with their corresponding labeled reports. The detailed description and labeling methods are described in [1]. The images are distributed in 54 zip files adding up to 1 TB. Each images information is recorded in the PADCHEST_chest_x_ray_images_labels_160K.csv file, including image ID, image directory, study date, patient information, image parameters, etc. Additionally, it includes report, labels, localization information, etc. The labels are classified into three types: differential diagnosis, radiological findings, and anatomical locations. The dataset is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.