datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
luna16-dp2d
LUNA16-DiagnosticPreservation-2D (LUNA16-DP2D)
Dataset for Anatomical Attention-Based Medical Image Compression
Dataset Description
LUNA16-DP2D is a curated 2D axial CT slice dataset derived from the complete
LUNA16 challenge (888 scans, subsets 0–9).
It is designed for research on diagnostic-preservation image compression —
systems that allocate more bits to anatomically important regions (lungs, mediastinum,
nodules) rather than optimizing purely for perceptual… See the full description on the dataset page: https://huggingface.co/datasets/nourbourai/luna16-dp2d.LUNA16
LUNA16
A mirror of the LUNA16 (LUng Nodule Analysis 2016) challenge data — 888 thoracic
LDCT scans derived from LIDC-IDRI — repackaged for use in the EasyMedSeg
medical segmentation framework.
Contents
subset0/ … subset9/ — 888 CT scans in MetaImage format (.mhd + .raw)
seg-lungs-LUNA16/ — official lung-field masks (.mhd + .zraw)
nodule-masks-spheres/ — derived nodule masks built from annotations.csv
by drawing a sphere of diameter_mm at each nodule centroid.… See the full description on the dataset page: https://huggingface.co/datasets/chungchi1278/LUNA16.reconstruction2_unetv2_luna16LUNA16
LUNA16
Curated release of LUNA16 public lung CT nodule analysis data for research use.
Dataset viewer (Hugging Face)
The Dataset viewer loads data/train-00000-of-00001.parquet: an index of subject_id and ct_path for each case. NIfTI is not previewed in-browser; fetch volumes via Files.
An empty root metadata.csv was removed because it broke the Hub loader; the index is rebuilt from file paths.
OpenVAE / SMILE project
This dataset is published alongside the… See the full description on the dataset page: https://huggingface.co/datasets/MitakaKuma/LUNA16.model2_luna_val_luna16LUNA16_VALIDATION_UNETLUNA16_mask_validationLUNA16_mask_validationLUNA16_singadkar_validationLUNA16_validation_unet2d