ChestX-Det is an improved version of ChestX-Det10 https://github.com/Deepwise-AILab/ChestX-Det10-Dataset. On the basis of ChestX-Det10, we add three new categoires and 35 new images (from NIH ChestX-14). Besides, the segmentation annotations are also provided.
ChestX-Det consists of 3578 images from NIH ChestX-14. We invite three board-certified radiologists to annotate them with 13 common categories of diseases or abnormalities.
The 13 categories are Atelectasis, Calcification, Cardiomegaly, Consolidation, Diffuse Nodule, Effusion, Emphysema, Fibrosis, Fracture, Mass, Nodule, Pleural Thickening, Pneumothorax.
The annotation files are ChestX_Det_train.json and ChestX_Det_test.json. The format in annotation files are:
{
"file_name": "xxx.png",
"syms": [s1, s2, ...],
"boxes": [[x1, y1, x2, y2], [x1, y1, x2, y2], …],
"polygons": [[[x3, y3], [x4, y4], …], [[x3, y3], [x4, y4], …], ...]
}
syms: The categories of thoracic abnormalities or diseases.
boxes: x1, y1, x2, y2 are left-top, right-bottom coordinates of the bounding box.
polygons: x3, y3, x4, y4 are point set coordinates of mask contour.
For image downloading, please visit http://resource.deepwise.com/ChestX-Det/train_data.zip and http://resource.deepwise.com/ChestX-Det/test_data.zip.
For any question, please contact
lianjie@deepwise.com
In the paper "A Structure-Aware Relation Network for Thoracic Diseases Detection and Segmentation", we propose a SAR-Net which makes use of the anatomical information. We release the pre-trained PSPNet so that others can use the same amount of information as we used in the paper. The code is pre-trained_PSPNet. For pkl file downloading, please visit http://resource.deepwise.com/ChestX-Det/pspnet_chestxray_best_model_4.pkl.