Impact of bi-planar localization of the tricuspid valve on the evaluation of right ventricular functional parameters in the short axis plane.

Abstract

Primary objective was to evaluate by cardiac MRI the accuracy of right ventricular stroke volume (RVSV) measurement in the short-axis (SA) plane with cross-referencing of the tricuspid plane. 2D phase-contrast measurement at the main pulmonary artery (PSV) was the reference. Secondary objective was to analyze the reproducibility of RV functional parameters. In this single-center retrospective study, 41 patients (mean age 40 ± 18years; age range 16-71years; M/F sex ratio 51%) referred for various acquired and congenital cardiopathies underwent CMR including SA balanced steady state free precession imaging (b-SSFP). Right ventricular vertical long-axis and four chamber views were used for cross-referenced localization of the tricuspid valve. Right ventricular functional parameters were measured on three occasions by two observers using Syngo Via® (Siemens Healthineers, Erlangen, Germany). The Student t-test and Bland Altman plot were used to test for differences between RV stroke volumes derived from cine b-SSFP (RVSV) or 2D PC (PSV). Bland Altman plots, coefficient of variation (COV) and intraclass correlation coefficient (ICC) were used to evaluate intra- and inter-observer reproducibility of RVSV, RVED and RVES volumes, and RV ejection fraction. There was high correlation (r = 0.94) and no significant difference between RVSV and PSV (83 ± 20mL vs. 81 ± 21mL p > 0.05). Intra- (ICC: 0.95; COV: 6.2) and inter-observer reliability (ICC: 0.91; COV: 8.9) of RVSV measurements were excellent. Finally, intra- and inter-observer reproducibility was excellent for RVEF, RVEDV and RVESV. Right ventricular stroke volumes can be routinely derived from SA analysis using cross-referenced localization of the atrioventricular plane. Moreover, all right ventricular systolic function parameters are highly reproducible when using this technique.