Mining Chest CT Metrics to Improve Patient Outcomes
Non-contrast chest CT has numerous imaging biomarkers that can help improve patient outcomes. The primary one is coronary artery calcium score (CAC). While the outcomes data using CAC are from cardiac gated CT exams, non-gated chest CT ordinal calcium score also correlates with major adverse cardiac events. CT of bone mineralization is rarely commented on in this age of DEXA scanning. Recent work shows that the L1 Hounsfield unit density correlates with risk of fracture. Ordinal scoring of emphysema correlates with mortality risk. These simple metrics, based on HU density, can be used by Artificial Intelligence to mine patient imaging and clinical data to create nomograms and survival curves. The radiology report of the future will better inform patients about their physiologic age, etc.
Department: Radiology
Contact: Tricia Hatcliff | 650-498-7359 | thatcliff@stanford.edu
Presenter(s):
- Mark Schiebler, MD Professor of Cardiothoracic Radiology
