Lung cancer – the leading cause of cancer-related deaths in the United States – kills more people than cancers of the breast, prostate, and colon combined. Researchers sought to develop better methods to screen for lung cancer in high-risk populations before symptoms appear, but as with many technologies, refinements are needed.
Low-dose computerized tomography screening has been shown to reduce lung cancer mortality. The NCI-supported National Lung Screening Trial (NLST), found about 20% fewer lung cancer deaths among trial participants randomized to the low-dose CT screening arm compared to those randomized to the chest X-ray arm. However, a high false-positive rate is a barrier to wider adoption of low-dose CT screening. In NLST the false positive rate was on average about 23% in the low-dose CT arm across the three screening rounds.
It is anticipated that artificial intelligence (AI) – the use of computer programs, or algorithms, that use data to make decisions or predictions – can substantially reduce the false positive rate of low-dose CT screening while minimally affecting test sensitivity, thereby reducing diagnostic uncertainty. However, AI research is limited by the availability of high-quality image databases. To address this gap, NCI awarded a three-year project to create a library of low-dose CT images and corresponding data related to screening for lung cancer. The images and corresponding data will be collected to support AI research in low-dose CT lung cancer screenings.
During a three-year period, participating medical facilities across the United States will collect low-dose CT screening and diagnostic chest CT images related to lung cancer screening and diagnostic follow-up. The images will be linked to up-to-date demographic and clinical data and then transferred to an NCI designated organization for long-term storage and dissemination to the research community.
Overall coordination is being supported by Booz Allen Hamilton under a larger 5-year umbrella contract with NCI for collection of medical images and data related to cancer screening. This contract will focus on supporting AI research in broad areas including cancer detection, diagnosis, risk prediction, and clinical decision making.