Numerical investigation of the performance of three hinge designs of bileaflet mechanical heart valves

Abstract

Thromboembolic complications (TECs) of bileaflet mechanical heart valves (BMHVs) are believed to be due to the nonphysiologic mechanical stresses imposed on blood elements by the hinge flows. Relating hinge flow features to design features is, therefore, essential to ultimately design BMHVs with lower TEC rates. This study aims at simulating the pulsatile three-dimensional hinge flows of three BMHVs and estimating the TEC potential associated with each hinge design. Hinge geometries are constructed from micro-computed tomography scans of BMHVs. Simulations are conducted using a Cartesian sharp-interface immersed-boundary methodology combined with a second-order accurate fractional-step method. Leaflet motion and flow boundary conditions are extracted from fluid-structure-interaction simulations of BMHV bulk flow. The numerical results are analyzed using a particle-tracking approach coupled with existing blood damage models. The gap width and, more importantly, the shape of the recess and leaflet are found to impact the flow distribution and TEC potential. Smooth, streamlined surfaces appear to be more favorable than sharp corners or sudden shape transitions. The developed framework will enable pragmatic and cost-efficient preclinical evaluation of BMHV prototypes prior to valve manufacturing. Application to a wide range of hinges with varying design parameters will eventually help in determining the optimal hinge design.

FIGURE 1

Hinge numerical model. In the large-scale numerical model, a BMHV model is inserted into a simplified aorta consisting of a straight tube with an axisymmetric expansion representing the sinus region. As shown in this figure, the hinge model corresponds to a section of this large-scale model. The butterfly recess characteristic of the hinge region is clearly visible in the zoom-in panel. The positions of the boundary planes defining the Cartesian grid are also included

FIGURE 2

Top and side views of the investigated hinge designs. The hinges of the two most commonly implanted bileaflet mechanical heart valves are studied: the St Jude Medical (SJM) valve and the CarboMedics (CM) valve. In order to assess the effect of the hinge gap width on the flow, two configurations of the SJM hinge are investigated: a SJM hinge with a regular hinge gap width (SJM regular hinge) and a SJM hinge with a larger than regular hinge gap width (SJM large hinge, not shown). The terminology used to describe the hinge recess is included. Note that the images are to scale and that the leaflet is in its fully closed position

FIGURE 3

Three-dimensional instantaneous streamtraces at peak systole and mid-diastole for all three hinge configurations. At peak systole, the streamtraces entering the hinge first through the ventricular side of the hinge are shown in blue, those entering the hinge through the aortic side of the hinge in red. At mid-diastole, the streamtraces are colored to help the identification of the three main leakage jets

FIGURE 4

Two-dimensional in-plane velocity vectors superimposed on the out-of-plane velocity (V ^o-p) contours at four instances of the cardiac cycle for the SJM regular hinge. The flow fields are shown at the flat level (level flush with the valve housing) and at the 195 μm level (level located 195 μm within the hinge recess, away from the flat level)

FIGURE 5

Two-dimensional in-plane velocity vectors superimposed on the out-of-plane velocity (V ^o-p) contours at four instances of the cardiac cycle for the SJM large hinge. The flow fields are shown at the flat level (level flush with the valve housing) and at the 195 μm level (level located 195 μm within the hinge recess, away from the flat level)

FIGURE 6

Two-dimensional in-plane velocity vectors superimposed on the out-of-plane velocity (V ^o-p) contours at four instances of the cardiac cycle for the CM hinge. The flow fields are shown at the flat level (level flush with the valve housing) and at the 195 μm level (level located 195 μm within the hinge recess, away from the flat level)

FIGURE 7

Iso-surfaces of shear stress levels at peak systole for all three hinge designs. Iso-surfaces of three shear stress levels (100, 500, and 1,000 dyn/cm²) are displayed in the near-hinge region. Top- and side-view schematics of the hinge are provided on the right side to illustrate the location of the shear stress iso-surfaces with respect to the hinge geometry

FIGURE 8

Iso-surfaces of shear stress levels at mid-diastole for all three hinge designs. Iso-surfaces of three shear stress levels (100, 500, and 1,000 dyn/cm²) are displayed in the near-hinge region. Top- and side-view schematics of the hinge are provided on the right side to illustrate the location of the shear stress iso-surfaces with respect to the hinge geometry

FIGURE 9

Characteristic flow features observed at peak systole and mid-diastole in all three hinge designs. Labels are used to identify each flow structure and are explained in the text in the discussion section

FIGURE 10

Cumulative distribution of the maximum shear stress experienced along the particle trajectories for the SJM regular, the SJM large, and the CM hinges

FIGURE 11

Cumulative distribution of the blood damage indices for hemolysis (left) and platelet activation (right) as a function of the particle percentage for all three hinge designs