Retrospective Evaluation of CT Based Biomarkers on Cancer and Non-Cancer Related Patient Outcomes
Principal Investigator
Name
Steven Rothenberg
Degrees
M.D.
Institution
The University of Alabama at Birmingham
Position Title
Assistant Professor
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1044
Initial CDAS Request Approval
Apr 13, 2023
Title
Retrospective Evaluation of CT Based Biomarkers on Cancer and Non-Cancer Related Patient Outcomes
Summary
CT scans contain valuable information that may be prognostic for mortality or conditions beyond the diagnosis of lung cancer. For example, artificial intelligence (AI) quantification of aortic calcifications on abdominal CT is a more accurate predictor of future CVD events than the Framingham risk score. Similarly, AI quantification of a coronary artery calcification (CAC) score on chest CT stratifies the risk of coronary artery disease (CAD). This investigation will explore a variety of automated CT based biomarkers to investigate differences in cancer and non-cancer related outcomes.
Aims
Retrospectively validate the accuracy of a variety of chest CT biomarkers for predicting mortality in the NLST dataset
- Compare accuracy and determine independence for the following or a subset of metrics: cardiothoracic ratio, coronary artery calcification (CAC-DRS), emphysema quantification, bone mineral density, etc
- Create a predictive model for mortality using a combination of features and biomarkers
- Compare accuracy and determine independence to lung cancer diagnosis for predicting mortality
Collaborators
Andrew Smith M.D. Ph.D. (University of Alabama at Birmingham)
Seth Lirette Ph.D. (Blackshear & Lirette, LLC)