ChestXRay2017胸腔X射线透视图数据集

ChestXRay2017数据集共包含5856张胸腔X射线透视图，诊断结果（即分类标签）主要分为正常和肺炎，其中肺炎又可以细分为：细菌性肺炎和病毒性肺炎。 胸腔X射线图像选自广州市妇幼保健中心的1至5岁儿科患者的回顾性研究。所有胸腔X射线成像都是患者常规临床护理的一部分。为了分析胸腔X射线图像，首先对所有胸腔X光片进行了筛查，去除所有低质量或不可读的扫描，从而保证图片质量。然后由两名专业医师对图像的诊断进行分级，最后为降低图像诊断错误，还由第三位专家检查了测试集。 ChestXRay2017数据集的目录结构： 主要分为train和test两大子文件夹，分别用于模型的训练和测试。在每个子文件内又分为了NORMAL(正常)和PNEUMonIA(肺炎)两大类。在PNEUMONIA文件夹内含有细菌性和病毒性肺炎两类，可以通过图片的命名格式进行判别。

ChestXRay2017 dataset consists of 5856 chest X-ray radiographs. Its diagnostic classification labels are mainly divided into normal and pneumonia, where pneumonia can be further subdivided into bacterial pneumonia and viral pneumonia.

These chest X-ray images were selected from a retrospective study of pediatric patients aged 1 to 5 years at the Guangzhou Women and Children's Medical Center. All chest X-ray imaging was part of the routine clinical care for these patients.

To facilitate the analysis of the chest X-ray images, all radiographs were first screened to remove low-quality or unreadable scans, ensuring the overall image quality. Subsequently, the diagnostic grading of the images was performed by two professional physicians. To mitigate diagnostic errors, the test set was additionally reviewed by a third expert physician.

The directory structure of the ChestXRay2017 dataset is as follows: It is primarily split into two subfolders, train and test, which are used for model training and testing respectively. Each subfolder contains two main categories: NORMAL (normal) and PNEUMONIA (pneumonia). Within the PNEUMONIA folder, there are two subcategories: bacterial pneumonia and viral pneumonia, which can be distinguished via the naming format of the image files.

搜集汇总

背景与挑战

背景概述

ChestXRay2017数据集包含5856张儿科胸腔X射线图像，用于诊断正常和肺炎（细菌性和病毒性）。数据经过专业医师质量控制，结构分为训练集和测试集，适用于医学影像分析研究。