Hemodynamics in the Pivoting Area of the Triflo Mechanical Heart Valve
Bernhard Vennemann1,2, David Hasler2, Silje E. Jahren2, Thomas Rösgen1, Thierry P. Carrel3, Dominik Obrist2
1: Institute of Fluid Dynamics, ETH Zurich, Switzerland
2: ARTORG Center, University of Bern, Switzerland
3: Department of Cardiovascular Surgery, Bern University Hospital, Switzerland
Mechanical heart valve replacements offer a long lifetime. However, they often show unphysiological hemodynamical behavior that severely increases the risk of shear induced blood damage, such as hemolysis and platelet activation. Tissue valves yield a more natural blood flow, but are limited in their life time to 10-15 years or even less in young patients. A Triflo FURTIVA mechanical heart valve, which was investigated in this project, has been developed to combine the favorable hemodynamics of tissue valves with the durability of mechanical heart valves (figure 1).
The pivoting area of mechanical heart valves is of special importance for flow diagnostics because it may be a region of high shear stresses combined with regions of low flow velocities, so called hot spots, where activated platelets might aggregate and form a thrombus that can ultimately lead to leaflet blockage. It is the aim of this study to assess the hemodynamics in the pivoting area of the Triflo valve.
Micro particle image velocimetry (µPIV) offers a scientific tool to investigate the flow on a small scale in an in vitro setup. Fluorescently labeled and neutrally buoyant particles are seeded in the flow which then closely follow the streamlines of the velocity field. The particles are illuminated with a series of short light pulses and captured with a camera. The velocity field is then determined from the relative shift between the particle patterns in two subsequent pictures (figure 2). Figure 4 illustrates the experimental setup that was built to investigate the hinge flow under physiological conditions. A novel 3D-printing approach was used to manufacture a transparent test section and valve housing for optical access to the hinge (figure 5). This technique has several advantages over traditional silicone casting as it allows to resolve even very small structures and complex geometries of the model while maintaining its functional integrity. The test section models the aortic root according to Swanson  as current research indicates the strong effect of the sinus portion on the hemodynamics of heart valves. The scene was captured with a high speed camera at 10.000 FPS. However, pulsed illumination at a frequency of 10 kHz is not possible with most currently available lasers. Therefore, we made use of a recently developed technique in which high power LED pulses that are well above the LED’s continuous current rating are used for localized volume illumination (figure 3). This enabled us to obtain velocity fields at a high rate to resolve the fast valve and flow dynamics. A mixture of glycerol, water, and sodium iodide served as a blood-mimicking fluid that matches both the kinematic viscosity of blood of 3.5 cSt as well as the refractive index (RI) of the transparent housing material. Matching the RIs of the different materials eliminates any distortions at the interfaces of the different media due to their differences in optical properties and ensures reliable particle identification for PIV.
Two flow scenarios with cardiac outputs (CO) of 3.0 L/min and 4.5 L/min respectively at a heart rate of 60 BPM were evaluated and the corresponding flow and pressure profiles are depicted in figure 6. Figure 7 shows the vertical velocity component of the resulting hinge flow, plotted over time. Here, one can distinguish between various flow phases during the cardiac cycle. The first phase (I) represents the acceleration of the fluid and the opening of the valve. The flow accelerates at a higher rate in the high CO case which also reflects on the hinge flow. This is followed by a phase of constant maximum velocity through the hinge in the fully-open leaflet position (II). The measured velocities in the gap did not exceed 1 m/s for 4.5 L/min CO and 0.8 m/s for 3.0 L/min CO. At a certain point the flow starts to decelerate and enables valve closure once the closing forces on the leaflets (due to vortical flow structures in the sinus portion) exceed the opposing forces of the core flow (III). The three individual leaflets closed in a very synchronous manner with only small variations. The average closing time was 55 ms and appeared to be independent of the cardiac output. However, the onset of closing was found to be at an earlier stage in the low CO case. The measured valve closing is considerably slower than for classical mechanical heart valves and is expected to have benefits in terms of hemodynamics . The fourth phase shows a small leakage flow through the gaps of the closed valve (IV). This leakage flow persists throughout the rest of the cycle and yields a wash-out of the hinge area with no region of flow stagnation.
Discussion and concluding remarks
The combination of several novel experimental technologies made it possible to perform transient flow measurements on a small scale and at a high frequency which allowed to resolve and study the fast valve dynamics during the cardiac cycle which often remain inaccessible to traditional techniques. In this work we proposed a cost-effective and efficient tool for the investigation of microfluidic phenomena.
The Triflo mechanical heart valve showed favorable hemodynamics due to synchronous and slow closing with no apparent regions of flow stagnation. The measured velocities did not exceed 0.8 m/s or 1.0 m/s for the two cases and a persistent leakage flow during the fully-closed phase provided a wash-out of the pivoting area. These aspects render the Triflovalve a potentially better alternative to current mechanical heart valve replacements.