LUNA16 (LUng Nodule Analysis 2016): 888 low-dose chest CT scans with nodule annotations and lung masks

Provider
Bram van Ginneken, Colin Jacobs et al. (LUNA16 organizers, Radboudumc); mirrored by MedOtter
License
CC-BY-4.0 (source)
Size
5,335 files
Format
CSV, Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/MedOtter/LUNA16)
Access
Free and public. No account with the data provider is needed.
DOI
10.5281/zenodo.3723295
Category
Health & Biomedical
Source
https://doi.org/10.1016/j.media.2017.06.015

About this dataset

LUNA16 (LUng Nodule Analysis 2016) is the lung nodule detection challenge built from the LIDC-IDRI collection: 888 thoracic CT scans with 1,186 annotated nodules. Scans are in MetaImage format (.mhd plus .raw) split into the ten official folds subset0 to subset9; annotations.csv gives each nodule's world-coordinate centroid and diameter in mm, candidates.csv and candidates_V2.csv list candidate locations for the false-positive reduction track, and seg-lungs-LUNA16/ holds the official automatic lung-field masks. The mirror adds nodule-masks-spheres/, approximate binary sphere masks drawn from the annotation diameters, which are not voxel-level ground truth. Read with SimpleITK or itk. Use for pulmonary nodule detection, false-positive reduction, lung cancer screening CAD and lung segmentation. Mirrored by MedOtter from the organizers' two Zenodo records (3723295 and 4121926), both CC BY 4.0, Zenodo DOI 10.5281/zenodo.3723295; cite Setio et al., Medical Image Analysis 2017 (doi 10.1016/j.media.2017.06.015) and the LIDC-IDRI collection.

How to load this dataset in Python

import pandas as pd
df = pd.read_csv("hf://datasets/MedOtter/LUNA16/annotations.csv")
df.head()

Data structure

File types in sampled files: CSV, Parquet

Files

Example file: annotations.csv

Analyze this dataset without downloading it

On Scigantic, LUNA16 (LUng Nodule Analysis 2016): 888 low-dose chest CT scans with nodule annotations and lung masks opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Health & Biomedical datasets