Largest annotated lung cancer screening dataset from Québec enables fair AI testing
LUCQUE: A lung cancer screening dataset from Québec for reliable AI evaluation
DatabasesArtificial Intelligence
Summary
Lung cancer screening uses low-dose CT scans to find potentially dangerous nodules early. The authors created LUCQUE, a large set of these scans from people in Québec, with detailed markings from experts showing the nodules and other features. This well-organized dataset is openly accessible using special software that keeps patient data private while letting AI developers test their tools fairly and reliably. This helps improve and compare AI tools for lung cancer detection using real-world data.
What this means in practice
- •For ai developers: Evaluate lung cancer detection algorithms on a large, well-annotated, real-world dataset with privacy safeguards.
- •For medical imaging platform teams: Integrate the LUCQUE dataset to benchmark and improve lung screening tools while keeping patient data secure.
Authors
Johanna Topalis, Yannick Lemaréchal, Gabriel Couture, Hanife Ersay, Loran Nas, Balthasar Schachtner, Bastian Oliver Sabel, Michael Ingrisch, Philippe Després, Katharina Jeblick
Abstract
BACKGROUND: AI-based tools are expected to play a key role in lung cancer screening (LCS). The evaluation of these tools often relied on private datasets or fragmented public datasets with inconsistent availability of clinical data and expert nodule annotations. This has hindered reproducibility, comparability, and independent assessment of AI tools. To address these gaps, we present LUCQUE, a novel, large, comprehensively annotated dataset derived from LDCT lung cancer screening in Québec, which is accessible for reliable AI evaluation. METHODS: LUCQUE was curated and is hosted on PARADIM, a digital infrastructure supporting DICOM data de-identification, annotation, long-term storage and analysis. LDCT scans and clinical information were combined to a curated collection and complemented by voxel-level, expert-derived pulmonary nodule and cyst segmentations, and Lung-RADS-based lesion characteristics. Key dataset characteristics and statistics were reported, including participant demographics, imaging parameters, radiological findings, and segmented lesion characteristics. Dataset analyses were conducted according to PARADIM's model-to-data approach, in which user code is executed on the platform such that data remains on premise. RESULTS: LUCQUE currently comprises 1,678 LDCT baseline studies from 1,678 participants. Most studies were classified as Lung-RADS 2. A total of 1,776 pulmonary nodules and 60 cysts were annotated, with a median (interquartile range) volume of $89\,(56,\,159)\,\mathrm{mm}^3$ and $1,384\,(589,\,3793)\,\mathrm{mm}^3$ respectively. CONCLUSIONS: Compared to publicly available LCS datasets, LUCQUE currently represents the largest accessible annotated real-world LCS dataset. Access to LUCQUE is provided via a model-to-data approach, enabling privacy-preserving, reproducible, and independent real-world evaluation of AI tools in LCS.