Quantification of left and right ventricular function and myocardial mass: Comparison of low-radiation dose 2nd generation dual-source CT and cardiac MRI

Abstract

To prospectively evaluate the accuracy of left and right ventricular function and myocardial mass measurements based on a dual-step, low radiation dose protocol with prospectively ECG-triggered 2nd generation dual-source CT (DSCT), using cardiac MRI (cMRI) as the reference standard. Twenty patients underwent 1.5T cMRI and prospectively ECG-triggered dual-step pulsing cardiac DSCT. This image acquisition mode performs low-radiation (20% tube current) imaging over the majority of the cardiac cycle and applies full radiation only during a single adjustable phase. Full-radiation-phase images were used to assess cardiac morphology, while low-radiation-phase images were used to measure left and right ventricular function and mass. Quantitative CT measurements based on contiguous multiphase short-axis reconstructions from the axial CT data were compared with short-axis SSFP cardiac cine MRI. Contours were manually traced around the ventricular borders for calculation of left and right ventricular end-diastolic volume, end-systolic volume, stroke volume, ejection fraction and myocardial mass for both modalities. Statistical methods included independent t-tests, the Mann-Whitney U test, Pearson correlation statistics, and Bland-Altman analysis. All CT measurements of left and right ventricular function and mass correlated well with those from cMRI: for left/right end-diastolic volume r=0.885/0.801, left/right end-systolic volume r=0.947/0.879, left/right stroke volume r=0.620/0.697, left/right ejection fraction r=0.869/0.751, and left/right myocardial mass r=0.959/0.702. Mean radiation dose was 6.2±1.8mSv. Prospectively ECG-triggered, dual-step pulsing cardiac DSCT accurately quantifies left and right ventricular function and myocardial mass in comparison with cMRI with substantially lower radiation exposure than reported for traditional retrospective ECG-gating.