Evaluation of mitral regurgitation using ultrafast ultrasound: in-vitro validation in a pulsatile flow phantom

Introduction: Mitral regurgitation (MR) severity is assessed by measuring the regurgitant volume (RVol) with echocardiography using the proximal iso-velocity surface area (PISA) method. The low temporal resolution of focused ultrasound, the assumption of perfect hemispheric iso-velocity contours, and the use of a single point to assess the velocity are the pitfalls inherent to this method. This study aimed to evaluate a method using ultrafast ultrasound with higher temporal resolution, and multiple point for velocity assessment to provide a more accurate volume measurement.Methods: A custom left heart mock circulatory system, comprised of a left atrium and a left ventricle connected by an orifice simulating MR, was filled with demineralized water mixed with cellulose. Pulsatile flow was generated using predefined RVols of 20mL, 30mL and 40 mL. Data were acquired at 6000 frames/s by a conventional cardiac phased array probe (central frequency at 2.75MHz) emitting ultrafast diverging waves, connected to an ultrafast scanner. RVols were computed as the surface integral of the velocity through a planar disk at the input of the orifice and with the PISA method at this same point. The root mean squared error (RMSE), mean absolute percentage error (MAPE) and accuracy of the two methods was compared with a t-test to evaluate the performances of PISA and the planar disk method.Results: RVol estimation using the planar disk method showed a much lower absolute error values than the clinical standard PISA method (15±9% vs 81±5%, p<0.001).Conclusion: This study demonstrates that ultrafast ultrasound can provide accurate estimate of RVol in a model of MR by leveraging the complete axial velocity mapping at high temporal resolution. Its complementary role to PISA method should be evaluated in future clinical studies.