In vitro study of trileaflet polytetrafluoroethylene conduit and its valve-in-valve transformation.

OBJECTIVES: Handmade trileaflet expanded polytetrafluoroethylene valved conduit developed using the flip-over method has been tailored for pulmonary valve reconstruction with satisfactory outcomes. We investigated the in vitro performance of the valve design in a mock circulatory system with various conduit sizes. In our study, the design was transformed into a transcatheter stent graft system which could fit in original valved conduits in a valve-in-valve fashion. METHODS: Five different sizes of valved polytetrafluoroethylene vascular grafts (16, 18, 20, 22 and 24 mm) were mounted onto a mock circulatory system with a prism window for direct leaflets motion observation. Transvalvular pressure gradients were recorded using pressure transducers. Mean and instant flows were determined via a rotameter and a flowmeter. Similar flip-over trileaflet valve design was then carried out in 3 available stent graft sizes (23, 26 and 28.5 mm, Gore aortic extender), which were deployed inside the valved conduits. RESULTS: Peak pressure gradient across 5 different sized graft valves, in their appropriate flow setting (2.0, 2.5 and 5.0 l/min), ranged from 4.7 to 13.2 mmHg. No significant valve regurgitation was noted (regurgitant fraction: 1.6-4.9%) in all valve sizes and combinations. Three sizes of the trileaflet-valved stent grafts were implanted in the 4 sizes of valved conduits except for the 16-mm conduit. Peak pressure gradient increase after valved-stent graft-in-valved-conduit setting was <10 mmHg in all 4 conduits. CONCLUSIONS: The study showed excellent in vitro performance of trileaflet polytetrafluoroethylene valved conduits. Its valved stent graft transformation provided data which may serve as a reference for transcatheter valve-in-valve research in the future.

The effect of turbulent viscous shear stress on red blood cell hemolysis.

Non-physiologic turbulent flow occurs in medical cardiovascular devices resulting in hemodynamic stresses that may damage red blood cells (RBC) and cause hemolysis. Hemolysis was previously thought to result from Reynolds shear stress (RSS) in turbulent flows. A more recent hypothesis suggests that turbulent viscous shear stresses (TVSS) at spatial scales similar in size to RBCs are related to their damage. We applied two-dimensional digital particle image velocimetry to measure the flow field of a free-submerged axisymmetric jet that was utilized to hemolyze porcine RBCs in selected locations. Assuming a dynamic equilibrium for the sub-grid scale (SGS) energy flux between the resolved and the sub-grid scales, the SGS energy flux was calculated from the strain rate tensor computed from the resolved velocity fields. The SGS stress was determined by the Smagorinsky model, from which the turbulence dissipation rate and then TVSS were estimated. Our results showed the hemolytic threshold of the Reynolds stresses was up to 517 Pa, and the TVSSs were at least an order of magnitude less than the RSS. The results provide further insight into the relationship between turbulence and RBC damage.

Numerical comparison of the closing dynamics of a new trileaflet and a bileaflet mechanical aortic heart valve.

The closing velocity of the leaflets of mechanical heart valves is excessively rapid and can cause the cavitation phenomenon. Cavitation bubbles collapse and produce high pressure which then damages red blood cells and platelets. The closure mechanism of the trileaflet valve uses the vortices in the aortic sinus to help close the leaflets, which differs from that of the monoleaflet or bileaflet mechanical heart valves which mainly depends on the reverse flow. We used the commercial software program Fluent to run numerical simulations of the St. Jude Medical bileaflet valve and a new trileaflet mechanical heart valve. The results of these numerical simulations were validated with flow field experiments. The closing velocity of the trileaflet valve was clearly slower than that of the St. Jude Medical bileaflet valve, which would effectively reduce the occurrence of cavitation. The findings of this study are expected to advance the development of the trileaflet valve.

Role of vortices in cavitation formation in the flow at the closure of a bileaflet mitral mechanical heart valve.

Bubble cavitation occurs in the flow field when local pressure drops below vapor pressure. One hypothesis states that low-pressure regions in vortices created by instantaneous valve closure and occluder rebound promote bubble formation. To quantitatively analyze the role of vortices in cavitation, we applied particle image velocimetry (PIV) to reduce the instantaneous fields into plane flow that contains information about vortex core radius, maximum tangential velocity, circulation strength, and pressure drop. Assuming symmetrical flow along the center of the St. Jude Medical 25-mm valve, flow fields downstream of the closing valve were measured using PIV in the mitral position of a circulatory mock loop. Flow measurements were made during successive time phases immediately following the impact of the occluder with the housing (O/H impact) at valve closing. The velocity profile near the vortex core clearly shows a typical Rankine vortex. The vortex strength reaches maximum immediately after closure and rapidly decreases at about 10 ms, indicating viscous dissipation; vortex strength also intensifies with rising pulse rate. The maximum pressure drop at the vortex center is approximately 20 mmHg, an insignificant drop relative to atmospheric vapor pressures, which implies vortices play a minor role in cavitation formation.

Turbulence characteristics downstream of a new trileaflet mechanical heart valve.

Design limitations of current mechanical heart valves cause blood flow to separate at the leaflet edges and annular valve base, forming downstream vortex mixing and high turbulent shear stresses. The closing behavior of a bileaflet valve is associated with reverse flow and may lead to cavitation phenomenon. The new trileaflet (TRI) design opens similar to a physiologic valve with central flow and closes primarily due to the vortices in the aortic sinus. In this study, we measured the St. Jude Medical 27 mm and the TRI 27 mm valves in the aortic position of a pulsatile circulatory mock loop under physiologic conditions with digital particle image velocimetry (DPIV). Our results showed the major principal Reynolds shear stresses were <100 N/m2 for both valves, and turbulent viscous shear stresses were smaller than 15 N/m2. The TRI valve closed more slowly than the St. Jude Medical valve. As the magnitudes of the shear stresses were similar, the closing velocity of the valves should be considered as an important factor and might reduce the risks of thrombosis and thromboembolism.

Cavitation phenomena in mechanical heart valves: studied by using a physical impinging rod system.

When studying mechanical heart valve cavitation, a physical model allows direct flow field and pressure measurements that are difficult to perform with actual valves, as well as separate testing of water hammer and squeeze flow effects. Movable rods of 5 and 10 mm diameter impinged same-sized stationary rods to simulate squeeze flow. A 24 mm piston within a tube simulated water hammer. Adding a 5 mm stationary rod within the tube generated both effects simultaneously. Charged-coupled device (CCD) laser displacement sensors, strobe lighting technique, laser Doppler velocimetry (LDV), particle image velocimetry (PIV) and high fidelity piezoelectric pressure transducers measured impact velocities, cavitation images, squeeze flow velocities, vortices, and pressure changes at impact, respectively. The movable rods created cavitation at critical impact velocities of 1.6 and 1.2 m/s; squeeze flow velocities were 2.8 and 4.64 m/s. The isolated water hammer created cavitation at 1.3 m/s piston speed. The combined piston and stationary rod created cavitation at an impact speed of 0.9 m/s and squeeze flow of 3.2 m/s. These results show squeeze flow alone caused cavitation, notably at lower impact velocity as contact area increased. Water hammer alone also caused cavitation with faster displacement. Both effects together were additive. The pressure change at the vortex center was only 150 mmHg, which cannot generate the magnitude of pressure drop required for cavitation bubble formation. Cavitation occurred at 3-5 m/s squeeze flow, significantly different from the 14 m/s derived by Bernoulli's equation; the temporal acceleration of unsteady flow requires further study.

Estimation of viscous dissipative stresses induced by a mechanical heart valve using PIV data.

Among the clinical complications of mechanical heart valves (MHVs), hemolysis was previously thought to result from Reynolds stresses in turbulent flows. A more recent hypothesis suggests viscous dissipative stresses at spatial scales similar in size to red blood cells may be related to hemolysis in MHVs, but the resolution of current instrumentation is insufficient to measure the smallest eddy sizes. We studied the St. Jude Medical (SJM) 27 mm valve in the aortic position of a pulsatile circulatory mock loop under physiologic conditions with particle image velocimetry (PIV). Assuming a dynamic equilibrium assumption between the resolved and sub-grid-scale (SGS) energy flux, the SGS energy flux was calculated from the strain rate tensor computed from the resolved velocity fields and the SGS stress was determined by the Smagorinsky model, from which the turbulence dissipation rate and then the viscous dissipative stresses were estimated. Our results showed Reynolds stresses up to 80 N/m2 throughout the cardiac cycle, and viscous dissipative stresses below 12 N/m2. The viscous dissipative stresses remain far below the threshold of red blood cell hemolysis, but could potentially damage platelets, implying the need for further study in the phenomenon of MHV hemolytic complications.

Role of vortices in cavitation formation in the flow across a mechanical heart valve.

BACKGROUND AND AIM OF THE STUDY: Cavitation occurs during mechanical heart valve closure when the local pressure drops below vapor pressure. The formation of stable gas bubbles may result in gaseous emboli, and secondarily cause transient ischemic attacks or strokes. It is noted that instantaneous valve closure, occluder rebound and high-speed leakage flow generate vortices that promote low-pressure regions in favor of stable bubble formation; however, to date no studies have been conducted for the quantitative measurement and analysis of these vortices. METHODS: A Björk-Shiley Monostrut (BSM) monoleaflet valve was placed in the mitral position of a pulsatile mock circulatory loop. Particle image velocimetry (PIV) and pico coulomb (PCB) pressure measurements were applied. Flow field measurements were carried out at t = -5, -3, -1, -0.5, 0 (valve closure), 0.3, 0.5, 0.75, 1.19, 1.44, 1.69, 1.94, 2, 2.19, 2.54, 2.79, 3.04, 3.29, 3.54, 5 and 10 ms. The vortices were quantitatively analyzed using the Rankine vortex model. RESULTS: A single counter-clockwise vortex was The instantaneous formation of cavitation bubbles at mechanical heart valve (MHV) closure, which subsequently damage blood cells and valve integrity, is a well-known and widely studied phenomenon (1-4). Contributing factors seem to include the water-hammer, squeeze flow and Venturi effects, all of which are short-lived. Both, Dauzat et al. (5) and Sliwka et al. (6) have detected high-intensity transient signals (HITS) with transcranial Doppler ultrasound in the carotid and cerebral arteries of MHV recipients, while Deklunder (7) observed clinical occurrences of cerebral gas emboli that were not seen with bioprosthetic valves. These detected over the major orifice, while a pair of counter-rotating vortices was found over the minor orifice. Velocity profiles were consistent with Rankine vortices. The vortex strength and magnitude of the pressure drop peaked shortly after initial occluder-housing impact and rapidly decreased after 0.5 ms, indicating viscous dissipation, with a less significant contribution from the occluder rebound effect. The maximum pressure drop was on the order of magnitude of 40 mmHg. CONCLUSION: Detailed PIV measurements and quantitative analysis of the BSM mechanical heart valve revealed large-scale vortex formation immediately after valve closure. Of note, the vortices were typical of a Rankine vortex and the maximum pressure change at the vortex center was only 40 mmHg. These data support the conclusion that vortex formation alone cannot generate the magnitude of pressure drop required for cavitation bubble formation.

Cavitation behavior observed in three monoleaflet mechanical heart valves under accelerated testing conditions.

Accelerated testing provides a substantial amount of data on mechanical heart valve durability in a short period of time, but such conditions may not accurately reflect in vivo performance. Cavitation, which occurs during mechanical heart valve closure when local flow field pressure decreases below vapor pressure, is thought to play a role in valve damage under accelerated conditions. The underlying flow dynamics and mechanisms behind cavitation bubble formation are poorly understood. Under physiologic conditions, random perivalvular cavitation is difficult to capture. We applied accelerated testing at a pulse rate of 600 bpm and transvalvular pressure of 120 mm Hg, with synchronized videographs and high-frequency pressure measurements, to study cavitation of the Medtronic Hall Standard (MHS), Medtronic Hall D-16 (MHD), and Omni Carbon (OC) valves. Results showed cavitation bubbles between 340 and 360 micros after leaflet/housing impact of the MHS, MHD, and OC valves, intensified by significant leaflet rebound. Squeeze flow, Venturi, and water hammer effects each contributed to cavitation, depending on valve design.

Squeeze flow measurements in mechanical heart valves.

High-speed squeeze flow during mechanical valve closure is often thought to cause cavitation, either between the leaflet tip and flat contact area in the valve housing, seating lip, or strut flat seat stop, depending on design. These sites have been difficult to measure within the housing, limiting earlier research to study of squeeze flow outside the housing or with computational fluid dynamics. We directly measured squeeze flow velocity with laser Doppler velocimetry at its site of occurrence within the St. Jude Medical (SJM), Omnicarbon (OC), and Medtronic Hall Standard (MHS) 29 mm valves in a mock circulation loop. Quartz glass provided an observation window to facilitate laser penetration. Our results showed increasing squeeze flow velocity at higher heart rates: 2.39-3.44 m/s for SJM, 3.07-4.33 m/s for OC, and 3.87-5.33 m/s for MHS. Strobe lighting technique captured the images of cavitation formation. Because these results were obtained in a mock circulation loop, one can assume this may occur in vivo resulting in valve damage, hemolysis, and thromboembolism. However, velocities of this magnitude alone cannot produce the pressure drop required for cavitation, and the applicability of the Bernoulli equation under these circumstances requires further investigation.

Computational hemodynamics of an implanted coronary stent based on three-dimensional cine angiography reconstruction.

Coronary stents are supportive wire meshes that keep narrow coronary arteries patent, reducing the risk of restenosis. Despite the common use of coronary stents, approximately 20-35% of them fail due to restenosis. Flow phenomena adjacent to the stent may contribute to restenosis. Three-dimensional computational fluid dynamics (CFD) and reconstruction based on biplane cine angiography were used to assess coronary geometry and volumetric blood flows. A patient-specific left anterior descending (LAD) artery was reconstructed from single-plane x-ray imaging. With corresponding electrocardiographic signals, images from the same time phase were selected from the angiograms for dynamic three-dimensional reconstruction. The resultant three-dimensional LAD artery at end-diastole was adopted for detailed analysis. Both the geometries and flow fields, based on a computational model from CAE software (ANSYS and CATIA) and full three-dimensional Navier-Stroke equations in the CFD-ACE+ software, respectively, changed dramatically after stent placement. Flow fields showed a complex three-dimensional spiral motion due to arterial tortuosity. The corresponding wall shear stresses, pressure gradient, and flow field all varied significantly after stent placement. Combined angiography and CFD techniques allow more detailed investigation of flow patterns in various segments. The implanted stent(s) may be quantitatively studied from the proposed hemodynamic modeling approach.

An experimental study of steady flow patterns of a new trileaflet mechanical aortic valve.

Hemodynamic research shows that thrombosis formation is closely tied to flow field turbulent stress. Design limitations cause flow separation at leaflet edges and the annular valve base, vortex mixing downstream, and high turbulent shear stress. The trileaflet design opens like a physiologic valve with central flow. Leaflet curvature approximates a completely circular orifice, maximizing effective flow area of the open valve. Semicircular aortic sinuses downstream of the valve allow vortex formation to help leaflet closure. The new trileaflet design was hemodynamically evaluated via digital particle image velocimetry and laser-Doppler anemometry. Measurements were made during peak flow of the fully open valve, immediately downstream of the valve, and compared with the 27-mm St. Jude Medical (SJM) bileaflet valve. The trileaflet valve central flow produces sufficient pressure to inhibit separation shear layers. Absence of downstream turbulent wake eddies indicates smooth, physiologic blood flow. In contrast, SJM produces strong turbulence because of unsteady separated shear layers where the jet flow meets the aortic sinus wall, resulting in higher turbulent shear stresses detrimental to blood cells. The trileaflet valve simulates the physiologic valve better than previous designs, produces smoother flow, and allows large scale recirculation in the aortic sinuses to help valve closure.

The closing behavior of mechanical aortic heart valve prostheses.

Mechanical artificial heart valves rely on reverse flow to close their leaflets. This mechanism creates regurgitation and water hammer effects that may form cavitations, damage blood cells, and cause thromboembolism. This study analyzes closing mechanisms of monoleaflet (Medtronic Hall 27), bileaflet (Carbo-Medics 27; St. Jude Medical 27; Duromedics 29), and trileaflet valves in a circulatory mock loop, including an aortic root with three sinuses. Downstream flow field velocity was measured via digital particle image velocimetry (DPIV). A high speed camera (PIVCAM 10-30 CCD video camera) tracked leaflet movement at 1000 frames/s. All valves open in 40-50 msec, but monoleaflet and bileaflet valves close in much less time (< 35 msec) than the trileaflet valve (>75 msec). During acceleration phase of systole, the monoleaflet forms a major and minor flow, the bileaflet has three jet flows, and the trileaflet produces a single central flow like physiologic valves. In deceleration phase, the aortic sinus vortices hinder monoleaflet and bileaflet valve closure until reverse flows and high negative transvalvular pressure push the leaflets rapidly for a hard closure. Conversely, the vortices help close the trileaflet valve more softly, probably causing less damage, lessening back flow, and providing a washing effect that may prevent thrombosis formation.