Three-dimensional proximal flow convergence automatic calculation for determining mitral valve area in rheumatic mitral stenosis.

Abstract

Management of patients with mitral stenosis (MS) depends heavily on the accurate quantification of mitral valve area (MVA) using echocardiography. All currently used two-dimensional (2D) methods have limitations. Estimation of MVA using the proximal isovelocity surface area (PISA) method with real time three-dimensional (3D) echocardiography may circumvent those limitations. We aimed to evaluate the accuracy of 3D direct measurement of PISA in the estimation of MVA. Twenty-seven consecutive patients (median age of 63years; 77.8% females) with rheumatic MS were prospectively studied. Transthoracic and transesophageal echocardiography with 2D and 3D acquisitions were performed on the same day. The reference method for MVA quantification was valve planimetry after 3D-volume multiplanar reconstruction. A semi-automated software was used to calculate the 3D flow convergence volume. Compared to MVA estimation using 3D planimetry, 3D PISA showed the best correlation (rho=0.78, P<.0001), followed by pressure half-time (PHT: rho=0.66, P<.001), continuity equation (CE: rho=0.61, P=.003), and 2D PISA (rho=0.26, P=.203). Bland-Altman analysis revealed a good agreement for MVA estimation with 3D PISA (mean difference -0.03cm2 ; limits of agreement (LOA) -0.40-0.35), in contrast to wider LOA for 2D methods: CE (mean difference 0.02cm2 , LOA -0.56-0.60); PHT (mean difference 0.31cm2 , LOA -0.32-0.95); 2D PISA (mean difference -0.03cm2 , LOA -0.92-0.86). MVA estimation using 3D PISA was feasible and more accurate than 2D methods. Its introduction in daily clinical practice seems possible and may overcome technical limitations of 2D methods.