Cancer Data and Image Database for AI (CANDID-4-AI)

The American College of Radiology® (ACR®) is working in partnership with the National Cancer Institute  (NCI) and Booz Allen Hamilton to develop a central public repository of medical images and associated clinical data from lung cancer screening subjects for use in artificial intelligence (AI) algorithm development and validation.

This repository will provide an opportunity for the development of new AI and machine learning algorithms within the research community and may enable enhancements to clinical care, leading to improvements in the diagnosis, prevention and treatment of lung cancer.

The ACR Center for Research and Innovation™ (CRI) and ACR Data Science Institute® (DSI) will facilitate collection and harmonization of images and corresponding data from a total of 15,000 lung cancer screening subjects across up to 15 clinical sites within the United States. Socioeconomic status variables will also be collected, which will expand the breadth and useability of the dataset, provide downstream AI and social science research application, and lay groundwork for future data collection efforts and collaborations

CANDID-4-AI is a cooperative effort of the ACR Center for Research and Innovation and the ACR Data Science Institute .

Data collection for CANDID-4-AI is now live.


If your institution is interested in becoming a participating site for CANDID-4-AI, please contact the project management team: