Left Ventricular Function Can Reliably be Assessed From Dual-Source CT Using ECG-Gated Tube Current Modulation

Abstract

The aim of this study was to assess the reliability of global left ventricular (LV) function analysis using dual source CT (DSCT) with and without ECG-gated tube current modulation in comparison with magnetic resonance (MR) imaging. In 7 pigs with myocardial infarction retrospectively ECG-gated DSCT (2 x 64 x 0.6 mm; 120 kV; 380 mAs(rot.)) with and without tube current modulation (gating window: 40%-75%) as well as cine MR imaging of the heart were performed on the same day. From DSCT data 6 mm double-oblique short axis images were computed. DSCT as well as MR images were analyzed using Simpson method. LV volumes, ejection fraction and, myocardial mass were calculated. Wall motion was assessed visually. Standard deviations of attenuation values measured in the LV cavity were assigned to image noise and compared using Wilcoxon test. LV volumes were compared with Lin concordance-correlation coefficient (rho(c)). Regional wall motion was analyzed with weighted kappa statistics. Both, DSCT with (DSCT(DM)) and without (DSCT(norm)) dose modulation were in good agreement with MR imaging. For ejection fraction, there was an excellent concordance-correlation comparing DSCT(norm) with DSCT(DM) (rho(c) = 0.9749), DSCT(norm) with MR imaging (rho(c) = 0.9435), and DSCT(DM) with MR imaging (rho(c) = 0.9612). The corresponding kappa values comparing regional wall motion were kappa = 0.988, kappa = 0.901, and kappa = 0.888, respectively. During end-systole there was no significant difference in the image noise between DSCT(norm) and DSCT(DM) techniques (P = 0.8750), but the difference became significant when comparing end-diastolic DSCT images (P = 0.0156). ECG-gated tube-current modulation does not hamper the quantitative assessment of global and regional LV function using DSCT.