Automated quantification of epicardial adipose tissue (EAT) in coronary CT angiography; comparison with manual assessment and correlation with coronary artery disease

Abstract

Background: Epicardial adipose tissue (EAT) is emerging as a risk factor for coronary artery disease (CAD).

Objective: The aim of this study was to determine the applicability and efficiency of automated EAT quantification.

Methods: EAT volume was assessed both manually and automatically in 157 patients undergoing coronary CT angiography. Manual assessment consisted of a short-axis-based manual measurement, whereas automated assessment on both contrast and non-contrast-enhanced data sets was achieved through novel prototype software. Duration of both quantification methods was recorded, and EAT volumes were compared with paired samples t test. Correlation of volumes was determined with intraclass correlation coefficient; agreement was tested with Bland-Altman analysis. The association between EAT and CAD was estimated with logistic regression.

Results: Automated quantification was significantly less time consuming than automated quantification (17 ± 2 seconds vs 280 ± 78 seconds; P < .0001). Although manual EAT volume differed significantly from automated EAT volume (75 ± 33 cm(³) vs 95 ± 45 cm(³); P < .001), a good correlation between both assessments was found (r = 0.76; P < .001). For all methods, EAT volume was positively associated with the presence of CAD. Stronger predictive value for the severity of CAD was achieved through automated quantification on both contrast-enhanced and non-contrast-enhanced data sets.

Conclusion: Automated EAT quantification is a quick method to estimate EAT and may serve as a predictor for CAD presence and severity.

Keywords: Automated quantitative analysis; Coronary CT angiography; Coronary artery disease; Epicardial adipose tissue; Quantification.