Background: The position of the right ventricle (RV), often partly behind the sternum, implies difficulties to image the RV free wall using transthoracic echocardiography (TTE) and consequently limits the possibilities of stroke volume calculations. The aim of this study was to evaluate whether the volume of the right ventricle (RV) can be determined by combining TTE distance measurements that do not need the RV free wall to be fully visualized. Methods: The RV volume was approximated by an ellipsoid composed of three distances. Distance measurements, modeled RV stroke volumes (RVSV), and RV ejection fraction (RVEF) were compared to reference values obtained from cardiac magnetic resonance (CMR) imaging for 12 healthy volunteers. Results: Inter-modality comparisons showed that distance measurements were significantly underestimated in TTE compared to CMR. The modeled RV volumes using TTE distance measurements were underestimated compared to reference CMR volumes. There was, however, for TTE an agreement between modeled RVSV and left ventricular stroke volumes determined by biplane Simpson's rule. Similar agreement was shown between modeled RVSV based on CMR distance measurements and the CMR reference. Regarding RVEF, further studies including patients with a wider range of RVEF are needed to evaluate the method. Conclusion: In conclusion, the ellipsoid model of the RV provides good estimates of RVSVs, but volumes based on distance measurements from different modalities cannot be used interchangeably. (Echocardiography 2016;33:844-853)

BackgroundThere is today no established approach to estimate right ventricular ejection fraction (RVEF) using 2D transthoracic echocardiography (TTE). The aim of this study was to evaluate a new method for RVEF calculations using 2D TTE and compare the results with cardiac magnetic resonance (CMR) imaging and tricuspid annular plane systolic excursion (TAPSE).MethodsA total of 37 subjects, 25 retrospectively included patients and twelve healthy volunteers, were included to give a wide range of RVEF. The right ventricle (RV) was modeled as a part of an ellipsoid enabling calculation of the RV volume by combining three distance measurements. RVEF calculated according to the model, RVEFTTE, were compared with reference CMR-derived RVEF, RVEFCMR. Further, TAPSE was measured in the TTE images and the correlations were calculated between RVEFTTE, TAPSE and RVEFCMR.ResultsThe mean values were RVEFCMR = 43 ± 12% (range 20–66%) and RVEFTTE = 50 ± 9% (range 34–65%). There was a high correlation (r = 0.80, p < 0.001) between RVEFTTE and RVEFCMR. Bland-Altman analysis showed a mean difference between RVEFCMR and RVEFTTE of 6 percentage points (ppt) with limits of agreement from −11 to 23 ppt. The mean value for TAPSE was 19 ± 5 mm and the correlation between TAPSE and RVEFCMR was moderate (r = 0.54, p < 0.001). The correlation between RVEFTTE and RVEFCMR was significantly higher (p < 0.05) than the correlation between TAPSE and RVEFCMR.ConclusionsThe ellipsoid model shows promise for RVEF calculations using 2D TTE for a wide range of RVEF, providing RVEF estimates that were significantly better correlated to RVEF obtained from CMR compared to TAPSE.Electronic supplementary materialThe online version of this article (doi:10.1186/s12947-017-0096-5) contains supplementary material, which is available to authorized users.

Short-axis (SA) magnetic resonance (MR) images are commonly planned parallel to the left atrioventricular valve. This orientation leads to oblique slices of the right ventricle (RV) with subsequent difficulties in separating the RV from the right atrium in the SA images. The insertion points of the tricuspid valve (TV) in the myocardium can be clearly identified in the right ventricle long axis (RVLA) and four-chamber (4CH) views. The purpose of this study was to develop a method that transfers the position of the tricuspid plane, as seen in the RVLA and 4CH views, to the SA images to facilitate the separation of the RV from the atrium. This methodology, termed Dissociating the Right Atrium from the Ventricle Volume (DRAW), was applied in 20 patients for calculations of right ventricular stroke volume (RVSV). The RVSV using DRAW (RVSV(DRAW)) was compared to left ventricular stroke volumes (LVSV) obtained from flow measurements in the ascending aorta. The RVSV was also determined using the conventional method (RVSV(CONV)) where the stack of images from the SA views are summarized, and a visual decision is made of the most basal slice to be included in the RV. The mean difference between RVSV(DRAW) and LVSV was 0·1 ± 12·7 ml, while the mean difference between RVSV(CONV) and LVSV was 0·33 ± 14·3 ml. Both the intra- and interobserver variability were small using the DRAW methodology, 0·6 ± 3·5 and 1·7 ± 2·7 ml, respectively. In conclusion, the DRAW method can be used to facilitate the separation of the RV and the atrium.

Right ventricle stroke volume is determined by different individual volume changes as follows: RVSV(TPD) together with RVSVOther contributes to almost the entire RVSV in nearly equal proportions, while RVSV(PPD) contributes only a small amount and is approximately 30% of either RVSV(TPD) or RVSV(Other) .