Purpose: Given the right ventricular(RV)complex structure,three-dimensional(3D)methods would be more suitable for assessing RV volumes and function than twodimensional methods.Recently,3D speckle tracking echocardiography(3D-STE)has been increasingly used to quantify RV function and strain.However,direct comparisons of 3D-STE and cardiac magnetic resonance(CMR)imaging for evaluation of RV function and strain are limited.The aim of this study was to test the feasibility and accuracy of 3D-STE using comparison with CMR imaging.Methods: We enrolled 142 patients who agreed to undergo both CMR and 3D-STE on the same day.RV end-diastolic volume(RVEDV),RV end-systolic volume(RVESV),ejection fraction(EF)and longitudinal strain of RV free wall were obtained from 3DSTE and CMR.In addition,longitudinal strain of RV free wall was also obtained from 2D-STE.CMR imaging was the reference standard.The Pearson correlation coefficient and Bland-Altman analysis were used to assess inter-technique agreement.The patients were divided two groups: normal RV function or mild RV dysfunction group(CMRderived RVEF ≥30%,n=70)and severe RV dysfunction group(CMR-derived RVEF <30%,n=69).There are two groups divided by the main clinical diagnosis of patients: 51 patients with dilated cardiomyopathy,21 patients with heart transplantation,20 patients with coronary artery disease,18 patients with hypertrophic disease(10 patients with hypertrophic cardiomyopathy and 8 patients with hypertension),10 patients with valvular disease.All the above parameters were compared between 3D-STE and CMR,3D-STE technique was used to evaluate the accuracy in different degrees of right ventricular dysfunction and different diseases.Results: 3D-STE was feasible in 139 patients(98%).3D-STE-determined RV volumes,EF and longitudinal strain,and 2D-STE-derived RV longitudinal strain correlated strongly with CMR values(RVEDV,r=0.94;RVESV,r=0.95;RVEF,r=0.91;RV longitudinal strain,r=0.80,r=0.61;P <.001 for all).Compared with CMR reference,3D-STE-derived RVEDV,RVESV,RV longitudinal strain and 2D-STE-derived RV longitudinal strain were underestimated by 6.9±29.3ml,13.7±29.4ml,2.3±4.8%,2.6±6.6%,respectively.3D-STE-derived RVEF was overestimated by 4.2±6.3%.Compared with patients with RVEF≥30%,3D-STE-determined RV volumes,EF and longitudinal strain had better correlations with CMR values in patients with RVEF<30%(RVEDV,r=0.88 vs 0.92;RVESV,r=0.85 vs 0.93;RVEF,r=0.63 vs 0.76;RV longitudinal strain,r=0.52 vs 0.74;P <.001 for all).Compared with other patients,3DSTE-determined RV volumes,EF had better correlations with CMR values in dilated patients(RVEDV,r=0.94;RVESV,r=0.94;RVEF,r=0.90;P <.001 for all),3D-STEdetermined RV longitudinal strain had best correlation with CMR values in valvular patients(r=0.88,P <.001).3D-STE to measure RV volumes,EF,longitudinal strain were highly reproducible.Conclusions: 3D-STE is highly feasible and reproducible,and it correlates highly with the CMR method.3D-STE can accurately evaluate the RV volumes,function and strain in patients with different etiologies,especially in patients with severe RV dysfunction or dilated patients. |