Effect of Dual-Source Radiofrequency Transmission on Left Ventricular Measurements and Measurement Reproducibility at 3.0 T Cardiac MR Imaging: Comparison with Conventional Single-Source Transmission Reference

Abstract

To prospectively assess effect of dual-source radiofrequency (RF) transmission on left ventricular (LV) measurements and measurements reproducibility at 3.0 T MR using balanced steady-state free precession (b-SSFP) cine imaging, compared to the conventional single-source RF transmission reference approach. Cardiac b-SSFP cine imaging was performed in 19 subjects at 3.0 T MR equipped with dual-source RF transmission. All images were analyzed to obtain LV end-diastolic volume, end-systolic volume, stroke volume, ejection fraction, mass, LV end-diastolic inferior wall thickness, and interventricular septal thickness. The difference of all LV measurements between the two imaging techniques was tested with the paired t test and the intertechnique agreement was tested through linear regression and Bland-Altman analyses. Additionally, repeated LV measurements were performed to determine intra and interobserver variability with the Bland-Altman method, the 95% limits of agreement, the coefficient of variation (CV) and the intraclass correlation coefficient. Compared to conventional single-source, dual-source slightly overestimated end-diastolic volume, end-systolic volume, and stroke volume (mean differences, 3.9 mL ± 9.7, 1.1 mL ± 2.6, and 2.8 mL ± 9.1, respectively; p > 0.05), resulting in a small but significant positive bias in ejection fraction (1.5% ± 2.6; p = 0.021). Mass was significantly smaller with dual-source than with single-source (-4.0 g ± 6.5, p = 0.001). Dual-source slightly underestimated interventricular septal thickness (-0.29 mm ± 0.6, p = 0.067) and significantly underestimated LV end-diastolic inferior wall thickness (-0.55 mm ± 0.4, p < 0.0001). The two techniques in measurements correlated highly (r2 = 0.81 to 0.96, p < 0.0001). Intra and interobserver variability in dual-source measurements was much lower than that in single-source, and variability values were <14.0%. Improved image quality of b-SSFP cine imaging at 3.0 T MR with dual-source RF transmission may provide more reproducible LV measurements compared to conventional single-source approach. Dual-source RF transmission also provides a reasonable estimate of the LV measurements.