Evaluating the systemic right ventricle by CMR: the importance of consistent and reproducible delineation of the cavity

Background: The method used to delineate the boundary of the right ventricle (RV), relative to the trabeculations and papillary muscles in cardiovascular magnetic resonance (CMR) ventricular volume analysis, may matter more when these structures are hypertrophied than in individuals with normal cardiovascular anatomy. This study aimed to compare two methods of cavity delineation in patients with systemic RV. Methods: Twenty-nine patients (mean age 34.7 +/- 12.4 years) with a systemic RV (12 with congenitally corrected transposition of the great arteries (ccTGA) and 17 with atrially switched (TGA) underwent CMR. We compared measurements of systemic RV volumes and function using two analysis protocols. The RV trabeculations and papillary muscles were either included in the calculated blood volume, the boundary drawn immediately within the apparently compacted myocardial layer, or they were manually outlined and excluded. RV stroke volume (SV) calculated using each method was compared with corresponding left ventricular (LV) SV. Additionally, we compared the differences in analysis time, and in intra- and inter-observer variability between the two methods. Paired samples t-test was used to test for differences in volumes, function and analysis time between the two methods. Differences in intra- and inter-observer reproducibility were tested using an extension of the Bland-Altman method. Results: The inclusion of trabeculations and papillary muscles in the ventricular volume resulted in higher values for systemic RV end diastolic volume (mean difference 28.7 +/- 10.6 ml, p < 0.001) and for end systolic volume (mean difference 31.0 +/- 11.5 ml, p < 0.001). Values for ejection fraction were significantly lower (mean difference -7.4 +/- 3.9%, p < 0.001) if structures were included. LV SV did not differ significantly from RV SV for both analysis methods (p = NS). Including structures resulted in shorter analysis time (p < 0.001), and showed better inter-observer reproducibility for ejection fraction (p < 0.01). Conclusion: The choice of method for systemic RV cavity delineation significantly affected volume measurements, given the CMR acquisition and analysis systems used. We recommend delineation outside the trabeculations for routine clinical measurements of systemic RV volumes as this approach took less time and gave more reproducible measurements.