Direct quantification of mitral regurgitant flow volume by real-time three-dimensional echocardiography using dealiasing of color Doppler flow at the vena contracta.

BACKGROUND: Real-time 3-dimensional color Doppler echocardiographic (RT3DE) imaging has recently been demonstrated to provide accurate direct measurement of vena contracta area (VCA). The quantification of mitral regurgitant (MR) flow directly at the lesion using color Doppler echocardiography, however, has been prevented because of multiple aliasing from high flow velocities. Recent studies, however, have demonstrated that flow at the vena contracta is laminar, with a narrow velocity spectrum that should allow the dealiasing of color Doppler flow velocities for the accurate measurement of MR flow. This hypothesis was tested in an in vitro flow model and initial patient application, with magnetic resonance imaging (MRI) used as a reference.

METHODS: In an in vitro flow model, MR jets of flow rates from 5 to 60 mL/s were produced through asymmetric orifices of 0.2 to 0.6 cm(2). From RT3DE data sets, MR flow was calculated by the automated integration of the nonaliased color Doppler velocities over the VCA, with aliasing avoided by maximum baseline shift. Aliased flow was calculated as VCA times the Nyquist velocity times the number of aliasing transitions derived from the maximum continuous-wave Doppler velocity. Total MR flow was calculated as the sum of nonaliased and aliased flow. This approach was also clinically evaluated in 23 patients for the measurement of MR stroke volume against MRI and the hemispheric and hemielliptic proximal isovelocity surface area methods.

RESULTS: In vitro RT3DE imaging of VCA was feasible in all flow stages without color Doppler aliasing. Flow rates calculated from RT3DE data sets showed excellent correlation with actual flow rates (r = 0.99), with a mean difference of -0.05 +/- 0.5 mL/s (not significant by t test). In vivo, good correlation and agreement were found between MR stroke volume by dealiasing and MRI (r = 0.91, -1.8 +/- 7.1 mL; not significant by t test), with better correlation and agreement compared with hemispheric proximal isovelocity surface area (r = 0.81, -17.4 +/- 9.4 mL, P < .05) and hemielliptic proximal isovelocity surface area (r = 0.89, -11.7 +/- 7.4 mL, P < .05).

CONCLUSIONS: Dealiasing of color Doppler flow at the vena contracta is feasible and appears promising for measuring MR severity quantitatively. This novel approach can be readily implemented in current systems to provide rapid semiautomated MR flow volume and MR fraction.