Abstract 13154: Isotropic Calcium Scoring sing 0.5 mm Reconstructions from 320-Detector Row CT Datasets Identifies More Patients with Non-Zero Agatston Score and Hence More Subclinical Atherosclerosis, Compared to Standard 3.0 mm Reconstructions

Abstract

Background: Coronary artery calcium (CAC) can be quantified using 0.5 mm isotropic reconstructions from 320-detector row CT without increased radiation dose. Little is known about re-classification of patients with non-zero calcium scores and quantitative measures of CAC using 0.5 isotropic reconstructions, compared to standard 3.0 mm reconstructions.

Methods: Prospectively, 104 clinically referred patients (age 54±12 years; 50% male) were imaged on 320-detector row CT using prospectively triggered volumetric mode. Images were reconstructed using both standard 3.0 mm and 0.5 mm axial slices. Agatston score, lesion number and MESA percentile rank were determined by two independent observers. Noise was measured as standard deviation of attenuation values in the ascending aorta. Proportions of zero vs. non-zero CAC patients were compared and inter-rater agreement was determined by kappa statistics. Non-normally distributed values were compared by ranked Wilcoxon test. Continuous variables were correlated using linear regression.

Results: Proportion of non-zero calcium score was 23% higher on 0.5 mm datasets (76.9 vs. 53.9; p=0.0008). Kappa value was 0.94 and 0.52 for the proportion of non-zero calcium scores on 3.0 and 0.5 mm datasets, respectively. Agatston score, lesion number, calculated MESA percentile rank and noise were significantly higher on 0.5 mm vs. 3.0 mm reconstructions (Figure).The Agatston score had no association with noise (p=0.76). Median radiation dose was 1.31 mSv [1.31-1.45].

Conclusion: The 0.5 mm isotropic reconstruction reclassified 23% non-zero calcium score patients on 320-detector row CT. MESA percentile rank was also higher. Thus, more patients with sub-clinical atherosclerosis could potentially be identified on 0.5 mm reconstruction. However, with greater image noise, the results may be less reproducible. The clinical utility needs to be validated in large, prospective, population-based longitudinal studies.