Feasibility of epicardial adipose tissue quantification in non-ECG-gated low-radiation-dose CT: comparison with prospectively ECG-gated cardiac CT

Background: Epicardial adipose tissue (EAT) is an important indicator of cardiovascular risk. This parameter is generally assessed on ECG-gated computed tomography (CT) images. Purpose: To evaluate feasibility and reliability of EAT quantification on non-gated thoracic low-radiationdose CT examinations with respect to prospectively ECG-gated cardiac CT acquisition. Material and Methods: Sixty consecutive asymptomatic smokers (47 men; mean age 64 +/- 9.8 years) underwent low-dose CT of the chest and prospectively ECG-gated cardiac CT acquisitions (64-slice dual-source CT). The two examinations were reconstructed with the same range, field of view, slice thickness, and convolution algorithm. Two independent observers blindly quantified EAT volume using commercially available software. Data were compared with paired sample Student t-test, concordance correlation coefficients (CCC), and Bland-Altman plots. Results: No statistically significant difference was observed for EAT volume quantification with low-dose-CT (141.7 +/- 58.3 mL) with respect to ECG-gated CT (142.7 +/- 57.9 mL). Estimation of CCC showed almost perfect concordance between the two techniques for EAT-volume assessment (CCC, 0.99; mean difference, 0.98 +/- 5.1 mL). Inter-observer agreement for EAT volume estimation was CCC: 0.96 for low-dose-CT examinations and 0.95 for ECG-gated CT. Conclusion: Non-gated low-dose CT allows quantifying EAT with almost the same concordance and reliability as using dedicated prospectively ECG-gated cardiac CT acquisition protocols.