Mitral Regurgitation Assessment Using a Novel Automated Three-Dimensional Color-Flow Quantification: Comparison with Two-Dimensional Proximal Isovelocity Surface Area and Cardiac Magnetic Resonance.

BACKGROUND

Accccurate quantification of mitral regurgitation (MR) remains challenging with conventional two-dimensional (2D) echocardiaography because of geometric assumptions and single-frame analysis. We aimed to evaluate the feasibility and performance of a novel automated three-dimensional color-flow quantification (3D Auto CFQ) tool for MR assessment.

METHODS

We retrospectively analyzed 185 consecutive patients with MR who underwent transesophageal echocardiography (TEE). Mitral regurgitation severity and regurgitant volume (RVol) were quantified by 2D proximal isovelociroty surface area (PISA) and by 3D Auto CFQ, which integrates real-time 3D TEE color Doppler with a fluid dynamics-based model. In a cardiac magnetic resonance (CMR) subgroup cohort of 29 patients, CMR served as the reference. Subgroup analyses were prespecified by MR etiology, jet direction/duration, rhythm, and number of jets.

RESULTS

Three-dimensional Auto CFQ showed excellent feadsibility (97.3%) and strong reproducibility (intraobserver intraclass correlation coefficient [ICC] = 0.96; interobserver = ICC 0.91). Compared with 2D PISA, 3D Auto CFQ yielded lower RVol (36.3 ± 25.2 mL vs 65.9 ± 38.7 mL, P < .001) and fewer severe MR classifications (16.2% vs 53.3%, P < .001). Agreement with CMR was superior for 3D Auto CFQ (r = 0.860, ICC = 0.92) than for 2D PISA (r = 0.694, ICC = 0.64), with a significant difference between correlations (Fisher's Z, P < .001). Subgroup analyses showed considstent performance of 3D Auto CFQ across various MR etiologies, jet characdteristics, and rhythms, although correlation was weaker in atrial fibrillation (P = .005).

CONCLUSIONS

Three-dimensional Auto CFQ is a feasible and reproducible method for MR quantification. Its use resulted in lower MR severity grade classification than 2D PISA. In the CMR subgroup, 3D Auto CFQ showed higher agreement with CMR than 2D PISA.