Blood speckle imaging compared with conventional Doppler ultrasound for transvalvular pressure drop estimation in an aortic flow phantom

Figures

• 

  Fig. 1. The Ecoflex 0030 silicone aortic valve used for the experiments. Pictured from the front, outside of tube (left) and from the back, in situ (right)

• 

  Fig. 2. The phantom used for the velocity-based pressure drop estimation experiments

• 

  Fig. 3. A photograph taken during the experiments, illustrating probe orientation and tilt angle measurement

• 

  Fig. 4. Example plots of the pressure drop measured by the pressure sensors at 250 mL/s A pulsatile flow B constant flow rates. Blue points and error bars represent mean ± standard deviation from the 8 pressure sensors. The red point represents the maximum pressure drop and its location, estimated using a modified Akima piecewise cubic Hermite interpolation between the valve and sensor 4

• 

  Fig. 5. A Example continuous wave Doppler acquisition at 250 mL/s flow rate with manual measurement shown. B Example BSI velocity vector interface, showing the velocity of flow through the region of interest

• 

  Fig. 6. Angle-corrected pressure drop (simplified Bernoulli equation; mmHg) from BSI (orange dotted bars) and Doppler (pale bars), compared to ground truth acquired by sensors (dark bars). Data are presented as the mean estimated drop under each flow condition, grouping constant and pulsatile flows. Error bars represent ± standard deviation.* denotes statistical significance (p < 0.05)

• 

  Fig. 7. Absolute error of pressure drop estimations (mmHg) from BSI (orange bars) and Doppler (pale bars), compared to ground truth acquired by sensors. Data are presented as the mean value (cross), the median value (line), the upper and lower quartiles (box range) and the minimum and maximum values (whiskers) under each flow condition, grouping constant and pulsatile flows. * denotes statistical significance compared to the 100 mL/s flow rate ( p < 0.05)

• 

  Fig. 8. Bland–Altman plots illustrating the agreement between A Doppler vs. ground-truth and B BSI vs. ground-truth pressure drop estimations (n = 16). The solid line represents the bias between the 2 methods, while the 95% limits of agreement (± 1.96 standard deviation) are represented by the dashed lines

• 

  Fig. 9. Bland–Altman plots illustrating the agreement between A ground-truth B Doppler and C Blood Speckle Imaging estimations of pressure drop across two days of experiments (n = 16). The solid line represents the bias within the methods, while the 95% limits of agreement (± 1.96 standard deviation) are represented by the dashed lines

• 

  Fig. 10. A diagram illustrating velocity profiles in regions of A high flow and B low flow