Non-invasive cardiovascular magnetic resonance assessment of pressure recovery distance after aortic valve stenosis

Figures

• 

  Fig. 1. Definition of the pressure concepts in the context of aortic stenosis. Blood flows from left ventricular outflow tract (LVOT) into the ascending aorta (AAo) across the aortic valve that defines the geometric orifice area (GOA). When crossing the valve, a portion of the potential energy of the fluid is converted into kinetic energy, which is maximal at the effective orifice area (EOA) defining the peak pressure drop (ΔPpeak) where the ejection jet is narrowest. Downstream along the AAo the kinetic energy is transformed back to potential energy until a fully developed laminar flow profile where is the correct point to measure the net pressure drop (ΔPnet). The pressure recovery (ΔPrec) is the difference between ΔPpeak and ΔPnet. The pressure recovery distance (PRecDist) is defined as the distance necessary for the blood flow to go from the ΔPpeak to ΔPnet

• 

  Fig. 2. Illustration of the definition of the pressure recovery distance based on momentum (PRecDist-M; accessed via simplified advective work–energy relative pressure formulation—SAW) or based on turbulent dissipation (PRecDist-T) relative to the effective orifice area (EOA), based on an exemplary experimental result in our phantom. The distance is defined from the point of the effective orifice area (when momentum is greatest), until the recovery of 95% either the momentum created (in PRecDist-M) or until the accumulation of 95% of turbulent dissipation (in PRecDist-T )

• 

  Fig. 3. Workbench 1—flow phantom with 8 pressure sensors (channel 1 to 8, CH1-CH8), with their location relative to the aortic valve (AV point at X location = 0), and the respective measurement of pressure recovery distance (PRecDist). A A slice of the magnitude of the CMR image, showing most of the locations of the eight pressure ports. B Measurement of the pressure recovery distance from the measured pressure acquired (ΔPmeasured) and interpolated from the eight sensors. C Advective pressure component (ΔPSAW) from the 4D flow CMR data and from simulation (computed fluid dynamics, CFD) along the centerline of the phantom with 100 ml/s constant flow. Note that sensor 8 is located at 500 mm after aortic valve (AV) and is not being presented in the diagram

• 

  Fig. 4. Workbench 2, a rigid pipe with constant flow and different 3D-printed heart valves to simulate aortic valve stenosis (AS). A The valve geometries considered—tricuspid aortic valve (TAV), two different configurations of bicuspid aortic valves (BAV1 and BAV2), two circular valves with different diameters (Circ 1 and Circ2), and two malfunctioning prosthetic heart valves (PHV1 and PHV2) [29]. B The respective cross-sectional blood flow profile at vena contracta (VC) cross-plane. C The respective streamlines distal from the valve

• 

  Fig. 5. Agreement of pressure recovery distance (PRecDist) estimated from momentum computation (PRecDist-M) and measured PRecDist in workbench 1 including 3 constant (blue) and 5 pulsatile flow regimes (red) (total n = 8). A Linear regression analysis; B Bland–Altman plot for the agreement

• 

  Fig. 6. Agreement of pressure recovery distance (PRecDist) estimated from momentum computation (PRecDist-M) and turbulence-based (PRecDist-T) in workbench 2 including 4 constant flow regimes across 7 different valves (n = 28). A Linear regression analysis; B Bland–Altman plot for the agreement

• 

  Fig. 7. Relationship between pressure recovery (i.e., reduction of advection) and energy loss (i.e., turbulence dissipation) across the 28 experimental conditions of workbench 2, colour coded accordingly to their stenotic level (Peak advective pressure drop; ΔPSAW). Each line is built by the amount of advection (by ΔPSAW, in mmHg, Y axis) and the dissipation accumulated (by turbulence, in mmHg, X axis) at each point along the centreline of the vessel phantom. The low stenotic lines (i.e. weaker momentum in the jet) correspond to the more irregular (i.e. noisy) relationships. On opposition, the larger the stenotic level, the better the agreement. Furthermore the agreement is better as more momentum has been recovered

• 

  Fig. 8. Relationship between peak and net pressure drop (ΔP), given by the comparison between the peak advective pressure drop (SAW) at the EOA and the total turbulence dissipation given by work–energy relative pressure including the turbulent dissipation component (WERP-T) for all the valves (left panel) and each valve undergoing four different flow conditions (right panels). The overall agreement (left panel) is good, but each valve (right panels) reports quite different linear regression coefficients between peak advection and net pressure drop caused by turbulent dissipation, indicating that each valve has its own peak versus net pressure drop signature. The axis of the individual valve plots has the same units as the aggregate plot

• 

  Fig. 9. Relationship between pressure recovery distance via momentum recovery (PRecDist-M) and studied factors in 32 bicuspid aortic valve subjects. A Linear relationship with the magnitude of aortic stenotic burden (assessed by the simplified advective work energy relative pressure, ΔPSAW), and identification of two subgroups that do not fit well the model. B Lack of relationship with remaining factors considered and that have been reported to influence net pressure drop: radius, eccentricity angle (Ecc Angle) and eccentricity displacement (Ecc disp.) analysed at the effective and geometric orifice area location (respectively, EOA and GOA locations). Note that the patients presenting a single plateau of pressure recovery are presented in blue and those with double pressure recovery plateau are presented in red

• 

  Fig. 10. Example of a double plateau case (A), as well as a mild coarctation (CoA) located at plane c (B). Both cases display the advective momentum (ΔPSAW) along the centerline, the velocity profiles in 3 cross-section planes: (a) effective orifice area (EOA), (b) distal ascending aorta and (c) aortic arch; and the 3D geometry with the streamlines seeded from the same 3 cross-section planes