Automated AI-Based Aortic Measurements From Attenuation Correction CT as an Adjunctive Cardiovascular Risk Biomarker:...

Area of Science:

• Cardiology
• Radiology
• Artificial Intelligence

Background:

• Aortic enlargement predicts dissection and rupture but is often overlooked in myocardial perfusion imaging.
• Computed tomography (CT) attenuation correction scans are widely available during these procedures.

Purpose of the Study:

• To determine if automated, AI-derived aortic measurements from CT scans correlate with adverse outcomes.
• To assess the prognostic value of opportunistic aortic measurements in a large patient cohort.

Main Methods:

• Utilized CT attenuation correction scans from 29,339 patients across 10 centers.
• Employed a deep learning model for automated thoracic aorta segmentation and diameter measurement.
• Calculated aortic size index (ASI) by normalizing diameters to body surface area.

Main Results:

• Elevated ASI (ascending >2.2 cm/m², descending >1.6 cm/m²) was significantly associated with increased all-cause mortality.
• Ascending ASI showed an adjusted hazard ratio of 1.16 (P<0.001), and descending ASI showed 1.23 (P<0.001).
• The prognostic value of abnormal ASI remained significant, independent of age, sex, and perfusion abnormalities.

Conclusions:

• Artificial intelligence can extract valuable aortic size information from routine CT scans.
• Opportunistic aortic measurements from CT attenuation correction scans can serve as an adjunctive risk biomarker.
• This method adds prognostic value to myocardial perfusion imaging without additional radiation or imaging.