Related Experiment Videos
Lung image database consortium: developing a resource for the medical imaging research community
Samuel G Armato1, Geoffrey McLennan, Michael F McNitt-Gray
1Department of Radiology, MC 2026, University of Chicago, 5841 S Maryland Ave, Chicago, IL 60637, USA. s-armato@uchicago.edu
Radiology
|August 31, 2004
Summary
The Lung Image Database Consortium (LIDC) established a robust database for computer-aided diagnosis (CAD) research on lung nodules in CT scans. This resource facilitates the development and evaluation of CAD methods for improved lung nodule detection.
Area of Science:
- Medical Imaging
- Radiology
- Computer-Aided Diagnosis
Background:
- Advancing computer-aided diagnosis (CAD) for lung nodules in thoracic computed tomography (CT) requires a standardized research resource.
- The National Cancer Institute initiated the Lung Image Database Consortium (LIDC) to address this need.
Purpose of the Study:
- To describe the foundational consensus process undertaken by the LIDC.
- To outline the key issues addressed in establishing a robust database for lung nodule CAD research.
Main Methods:
- A consensus process involving five academic institutions was employed.
- Key issues addressed include mission statement, eligibility criteria, nodule definition, truth requirements, population process, and statistical framework.
Main Results:
- The LIDC has established a framework for a scientifically robust database.
- Consensus was reached on critical technical and clinical issues for database development.
Conclusions:
- The LIDC database, built on careful planning and consensus, will serve as a valuable international research resource.
- Sharing these foundational issues aims to guide future development and evaluation of lung nodule detection CAD methods.