Long-Term Prognostic Implications of Thoracic Aortic Calcification on CT Using Artificial Intelligence-Based...

Area of Science:

• Cardiovascular Imaging
• Artificial Intelligence in Medicine
• Preventive Cardiology

Background:

• Assessing thoracic aortic calcification (TAC) for prognostic purposes is challenging due to difficulties in standardization.
• Coronary artery calcification (CAC) is a known cardiovascular risk marker.
• The incremental prognostic value of TAC, especially when quantified by AI, requires further investigation in screening populations.

Purpose of the Study:

• To evaluate the long-term prognostic implications of AI-quantified TAC on routine chest CT scans.
• To assess if AI-based TAC improves cardiovascular risk prediction beyond established factors like CAC.

Main Methods:

• Retrospective analysis of 7404 asymptomatic individuals undergoing non-contrast chest CT for health screening.
• AI program quantified TAC and CAC (Agatston scores); manual quantification served as a reference.
• Multivariable Cox models assessed the independent association of AI-based TAC categories with major adverse cardiovascular events (MACE) and all-cause mortality (ACM), adjusting for CAC, clinical, and laboratory variables.

Main Results:

• Excellent agreement between AI-based and manual quantification for both TAC and CAC.
• AI-based TAC was not independently associated with MACE risk after adjusting for CAC and other variables.
• A high AI-based TAC score (1001-3000) was independently associated with increased ACM risk; however, AI-TAC did not improve prognostic model fit for MACE or ACM.

Conclusions:

• The addition of AI-quantified TAC to risk prediction models offers limited incremental value over CAC in asymptomatic individuals.
• AI-based quantification of TAC provides a standardized method for evaluating its potential as an imaging biomarker.
• Further research may clarify the specific role of TAC in cardiovascular risk stratification.