Tim A.S. Kaufmann

Flow Analysis of Ventricular Assist Device Inflow and Outflow Cannula Positioning Using a Naturally Shaped Ventricle and Aortic Branch

Tip geometry and placement of rotary blood pump inflow and outflow cannulae influence the dynamics of flow within the ventricle and aortic branch. Cannulation, therefore, directly influences the potential for thrombus formation and end-organ perfusion during ventricular assist device (VAD) support or cardiopulmonary bypass (CPB). The purpose of this study was to investigate the effect of various inflow/outflow cannula tip geometries and positions on ventricular and greater vessel flow patterns to evaluate ventricular washout and impact on cerebral perfusion. Transparent models of a dilated cardiomyopathic ventricle and an aortic branch were reconstructed from magnetic resonance imaging data to allow flow measurements using particle image velocimetry (PIV). The contractile function of the failing ventricle was reproduced pneumatically, and supported with a rotary pump. Flow patterns were visualized around VAD inflow cannulae, with various tip geometries placed in three positions in the ventricle. The outflow cannula was placed in the subclavian artery and at several positions in the aorta. Flow patterns were measured using PIV and used to validate an aortic flow computational fluid dynamic study. The PIV technique indicated that locating the inflow tip in the left ventricular outflow tract improved complete ventricular washout while the tip geometry had a smaller influence. However, side holes in the inflow cannula improved washout in all cases. The PIV results confirmed that the positioning and orientation of the outflow cannula in the aortic branch had a high impact on the flow pattern in the vessels, with a negative blood flow in the right carotid artery observed in some cases. Cannula placement within the ventricle had a high influence on chamber washout. The positioning of the outflow cannula directly influences the flow through the greater vessels, and may be responsible for the occasional reduction in cerebral perfusion seen in clinical CPB.

Flow Distribution During Cardiopulmonary Bypass in Dependency on the Outflow Cannula Positioning

Oxygen deficiency in the right brain is a common problem during cardiopulmonary bypass (CPB). This is linked to an insufficient perfusion of the carotid and vertebral artery. The flow to these vessels is strongly influenced by the outflow cannula position, which is traditionally located in the ascending aorta. Another approach however is to return blood via the right subclavian artery. A computational fluid dynamics (CFD) study was performed for both methods and validated by particle image velocimetry (PIV). A 3-dimensional computer aided design model of the cardiovascular (CV) system was generated from realtime computed tomography and magnetic resonance imaging data. Mesh generation (CFD) and rapid prototyping (PIV) were used for the further model creation. The simulations were performed assuming usual CPB conditions, and the same boundary conditions were applied for the PIV validation. The flow distribution was analyzed for 55 cannula positions inside the aorta and in relation to the distance between the cannula tip and the vertebral artery branch for subclavian cannulation. The study reveals that the Venturi effect due to the cannula jet appears to be the main reason for the loss in cerebral perfusion seen clinically. It provides a PIV-validated CFD method of analyzing the flow distribution in the CV system and can be transferred to other applications.

The Impact of Aortic/Subclavian Outflow Cannulation for Cardiopulmonary Bypass and Cardiac Support: A Computational Fluid Dynamics Study

Approximately 100 000 cases of oxygen deficiency in the brain occur during cardiopulmonary bypass (CPB) procedures each year. In particular, perfusion of the carotid and vertebral arteries is affected. The position of the outflow cannula influences the blood flow to the cardiovascular system and thus end organ perfusion. Traditionally, the cannula returns blood into the ascending aorta. But some surgeons prefer cannulation to the right subclavian artery. A computational fluid dynamics study was initially undertaken for both approaches. The vessel model was created from real computed tomography/magnetic resonance imaging data of young healthy patients. The simulations were run with usual CPB conditions. The flow distribution for different cannula positions in the aorta was studied, as well as the impact of the cannula tip distance to vertebral artery for the subclavian position. The study presents a fast method of analyzing the flow distribution in the cardiovascular system, and can be adapted for other applications such as ventricular assist device support. It revealed that two effects cause the loss of perfusion seen clinically: a vortex under the brachiocephalic trunk and low pressure regions near the cannula jet. The results suggest that cannulation to the subclavian artery is preferred if the cannula tip is sufficiently far away from the branch of the vertebral artery. For the aortic positions, however, the cannula should be injected from the left body side.

Implementation of intrinsic lumped parameter modeling into computational fluid dynamics studies of cardiopulmonary bypass

Stroke and cerebral hypoxia are among the main complications during cardiopulmonary bypass (CPB). The two main reasons for these complications are the cannula jet, due to altered flow conditions and the sandblast effect, and impaired cerebral autoregulation which often occurs in the elderly. The effect of autoregulation has so far mainly been modeled using lumped parameter modeling, while Computational Fluid Dynamics (CFD) has been applied to analyze flow conditions during CPB. In this study, we combine both modeling techniques to analyze the effect of lumped parameter modeling on blood flow during CPB. Additionally, cerebral autoregulation is implemented using the Baroreflex, which adapts the cerebrovascular resistance and compliance based on the cerebral perfusion pressure. The results show that while a combination of CFD and lumped parameter modeling without autoregulation delivers feasible results for physiological flow conditions, it overestimates the loss of cerebral blood flow during CPB. This is counteracted by the Baroreflex, which restores the cerebral blood flow to native levels. However, the cerebral blood flow during CPB is typically reduced by 10-20% in the clinic. This indicates that either the Baroreflex is not fully functional during CPB, or that the target value for the Baroreflex is not a full native cerebral blood flow, but the plateau phase of cerebral autoregulation, which starts at approximately 80% of native flow.

Implementation of cerebral autoregulation into computational fluid dynamics studies of cardiopulmonary bypass.

Peri- or postoperative neurological complications are among the main risks for patients undergoing extracorporeal circulatory support (ECC). Two of the main reasons are an increased risk for strokes and altered flow conditions leading to cerebral hypoperfusion. This is strongly affected by cerebral autoregulation, which is the bodys intrinsic ability to provide sufficient cerebral blood flow (CBF) despite changes in cerebral perfusion pressure (CPP). This complex mechanism has been mainly neglected in numerical studies, which have often been applied for analysis of ECC. In this study, a mathematical model is presented to implement cerebral autoregulation into computational fluid dynamics (CFD) studies. CFD simulations of cardiopulmonary bypass (CPB) were performed in a 3D model of the cardiovascular system, with flow variations between 4.5-6 L/min. Cerebral outlets were modeled using an equation to calculate CBF based on CPP. Assuming full regulation, CBF was kept constant for CPP between 80 and 120 mm Hg. A deviation in CBF of 20% occurred for CPP between 55-80 mm Hg and 120-145 mm Hg, respectively. The level of regulation was varied to take possible impairment of cerebral autoregulation into account. Furthermore, chronic hypertension was modeled by increasing the baseline CPP. Results indicate that even for full autoregulation, CBF is decreased during CPB. It is even lower for impaired autoregulation and hypertensive patients, demonstrating the strong impact of autoregulation on CBF. It is therefore imperative to include this mechanism into CFD studies. The presented model can help to improve CPB support conditions based on patient-specific autoregulation parameters.

Hemodynamic analysis of outflow grafting positions of a ventricular assist device using closed-loop multiscale CFD simulations: Preliminary results

Subclavian arteries are a possible alternate location for left ventricular assist device (LVAD) outflow grafts due to easier surgical access and application in high risk patients. As vascular blood flow mechanics strongly influence the clinical outcome, insights into the hemodynamics during LVAD support can be used to evaluate different grafting locations. In this study, the feasibility of left and right subclavian artery (SA) grafting was investigated for the HeartWare HVAD with a numerical multiscale model. A 3-D CFD model of the aortic arch was coupled to a lumped parameter model of the cardiovascular system under LVAD support. Grafts in the left and right SA were placed at three different anastomoses angles (90°, 60° and 30°). Additionally, standard grafting of the ascending and descending aorta was modelled. Full support LVAD (5l/min) and partial support LVAD (3l/min) in co-pulsation and counter-pulsation mode were analysed. The grafting positions were investigated regarding coronary and cerebral perfusion. Furthermore, the influence of the anastomosis angle on wall shear stress (WSS) was evaluated. Grafting of left or right subclavian arteries has similar hemodynamic performance in comparison to standard cannula positions. Angularity change of the graft anastomosis from 90° to 30° slightly increases the coronary and cerebral blood flow by 6-9% while significantly reduces the WSS by 35%. Cannulation of the SA is a feasible anastomosis location for the HVAD in the investigated vessel geometry.

In vitro flow investigations in the aortic arch during cardiopulmonary bypass with stereo-PIV

The cardiopulmonary bypass is related to complications like stroke or hypoxia. The cannula jet is suspected to be one reason for these complications, due to the sandblast effect on the vessel wall. Several in silico and in vitro studies investigated the underlying mechanisms, but the applied experimental flow measurement techniques were not able to address the highly three-dimensional flow character with a satisfying resolution. In this work in vitro flow measurements in a cannulated and a non-cannulated aortic silicone model are presented. Stereo particle image velocimetry measurements in multiple planes were carried out. By assembling the data of the different measurement planes, quasi 3D velocity fields with a resolution of~1.5×1.5×2.5 mm(3) were obtained. The resulting velocity fields have been compared regarding magnitude, streamlines and vorticity. The presented method shows to be a suitable in vitro technique to measure and address the three-dimensional aortic CPB cannula flow with a high temporal and spatial resolution.

Mimicking of Cerebral Autoregulation by Flow-Dependent Cerebrovascular Resistance: A Feasibility Study

Understanding circulatory autoregulation is essential for improving physiological control of rotary blood pumps and support conditions during cardiopulmonary bypass (CPB). Cerebral autoregulation (CAR), arguably the most critical, is the bodys intrinsic ability to maintain sufficient cerebral blood flow (CBF) despite changes in aortic perfusion pressure. It is therefore imperative to include this mechanism into computational fluid dynamics (CFD), particle image velocimetry (PIV), or mock circulation loop (MCL) studies. Without such inclusions, potential losses of CBF are overestimated. In this study, a mathematical model to mimic CAR is implemented in a MCL- and PIV-validated CFD model. A three-dimensional model of the human vascular system was created from magnetic resonance imaging records. Numerical flow simulations were performed for physiological conditions and CPB. The inlet flow was varied between 4.5 and 6 L/min. Arterial outlets were modeled using vessel-specific, flow-dependent cerebrovascular resistances (CVRs), resulting in a variation of the pressure drop between 0 and 80mmHg. CBF is highly dependent on the level of CAR during CPB. By varying the CVR parameters up to the beginning of plateau phase, it can be regulated between 0 and 80% of physiological CBF. So while implementing autoregulation, CBF remains unchanged during a simulated native cardiac output of 5L/min or CPB support of 6L/min. Neglecting CAR, constant backflow from the brain occurs for some cannula positions. Using flow-dependent CVR, CBF returns to its baseline at a rate of recovery of 0.25s. Results demonstrate that modeling of CAR by flow-dependent CVR delivers feasible results. The presented method can be used to optimize physiological control of assist devices dependent upon different levels of CAR representing different patients.

Numerical washout study of a pulsatile total artificial heart

Purpose For blood pumps with long term indication, blood stagnation can result in excessive thromboembolic risks for patients. This study numerically investigates the washout performance of the left pump chamber of a pulsatile total artificial heart (TAH) as well as the sensitivity of the rotational orientation of the inlet bileaflet mechanical heart valve (MHV) on blood stagnation. Methods To quantitatively evaluate the washout efficiency, a fluid-structure interaction (FSI) simulation of the artificial heart pumping process was combined with a blood washout model. Four geometries with different orientations (0°, 45°, 90° and 135°) of the inlet valve were compared with respect to washout performance. Results The calculated flow field showed a high level of agreement with particle image velocimetry (PIV) measurements. Almost complete washout was achievable after three ejection phases. Remains of old blood in relation to the chamber volume was below 0.6% for all configurations and were mainly detected opposite to the inlet and outlet port at the square edge where the membrane and the pump chamber are connected. Only a small variation in the washout efficiency and the general flow field was observed. An orientation of 0° showed minor advantages with respect to blood stagnation and recirculation. Conclusions Bileaflet MHVs were demonstrated to be only slightly sensitive to rotation regarding the washout performance of the TAH. The proposed numerical washout model proved to be an adequate tool to quantitatively compare different configurations and designs of the artificial organ regarding the potential for blood stagnation where experimental measurements are limited.

Transient, three-dimensional flow field simulation through a mechanical, trileaflet heart valve prosthesis.

Thromboembolic complications are one of the major challenges faced by designers and researchers in development of artificial organs with blood-contacting devices such as heart valve prostheses, especially mechanical valves. Besides increasing the thrombogenic potential, these valves change the hydrodynamic performance of the heart. In this study, the flow through a trileaflet, mechanical heart valve prosthesis was modeled with transient computational fluid dynamics to analyze flow patterns causing thrombus formations on valves. The valve was simulated under conditions of a test rig (THIA II), which was specially designed to analyze different valves with respect to thrombosis. The main goal of this study was to mimic the exact conditions of the test rig to be able to compare numerical and experimental results. The boundary conditions were obtained from experimental data as leaflet kinematics and pressure profiles. One complete cycle of the valve was simulated. Numerical flow and pressure results were analyzed and compared with experimental results. Shear stress and shear rates were determined with respect to thrombogenic potential, especially in the pivot regions, which seem to be the main influence for activation and deposition of thrombocytes. Approximately 0.7% of the blood volume moving through the fluid domain of the valve was exposed to shear rates high enough to cause platelet activation. However, shear rates of up to 20,000 s−1 occurred in pivot regions. The pressure differences between the simulation and experimental data were approximately 2.5% during systole and increased up to 25% during diastole. The presented method, however, can be used to gain more information about the flow through different heart valve prostheses and, thus, improve the development process.