PatchChestCT: A patch-level spatial annotation dataset for nine abnormalities in chest CT
PatchChestCT provides newly created binary 3D patch-level annotations for nine abnormality categories in 2,201 non-contrast chest CT studies selected from the public CT-RATE cohort (Hamamci et al., Nature Biomedical Engineering, 2026). Each annotation is stored as a binary NumPy array (.np
PatchChestCT provides newly created binary 3D patch-level annotations for nine abnormality categories in 2,201 non-contrast chest CT studies selected from the public CT-RATE cohort (Hamamci et al., Nature Biomedical Engineering, 2026).
Each annotation is stored as a binary NumPy array (.npz) with shape (24, 12, 12), corresponding to the 3D patch grid of the preprocessed CT volume.
Annotated abnormalities:
Arterial wall calcification · Coronary artery calcification · Pericardial effusion · Hiatal hernia · Lymphadenopathy · Atelectasis · Lung opacity · Consolidation · Bronchiectasis
IMPORTANT — This deposit contains PatchChestCT patch-level annotation files and the minimal CT-RATE volume identifiers required to link each annotation to its source volume. It does not include CT-RATE images, radiology reports, DICOM headers or subject-level clinical metadata. Source CT-RATE volumes must be obtained separately from the official CT-RATE repository and used under its Data Usage Agreement: https://huggingface.co/datasets/ibrahimhamamci/CT-RATE
Each subdirectory is named using the corresponding CT-RATE volume identifier (e.g., train_1003_a_1/) solely to enable unambiguous alignment with source volumes obtained separately by the user.
For alignment code and full usage instructions, see the companion code repository: https://github.com/SadVoxel/PatchChestCT
📤 Share this page
Found this useful? Share it with your network.
Files are hosted on the source repository. Click download to access the full dataset.