Design Method Using Statistical Models for Miniature Left Ventricular Assist Device Hydraulics.

Left ventricular assist devices (LVADs) are increasingly used to treat heart failure patients. These devices' impeller blades and diffuser vanes must be designed for hydraulic performance and hemocompatibility. The traditional design method, applying mean-line theory, is not applicable to the design of small-scale pumps such as miniature LVADs. Furthermore, iterative experimental testing to determine how each geometric variable affects hydraulic performance is time and labor intensive. In this study, we tested a design method wherein empirical hydraulic results are used to establish a statistical model to predict pump hydraulic performance. This method was used to design an intra-atrial blood pump. Five geometric variables were chosen, and each was assigned two values to define the variable space. The experimental results were then analyzed with both correlation analysis and linear regression modeling. To validate the linear regression models, 2 test pumps were designed: mean value of each geometric variable within the boundaries, and random value of each geometric variable within the boundaries. The statistical model accurately predicted the hydraulic performance of both pump designs within the boundary space. This method could be expanded to include more geometric variables and broader boundary conditions, thus accelerating the design process for miniature LVADs.

Hemodynamic Evaluation of an Intra-Atrial Blood Pump on a Pulsatile Mock Circulatory Loop.

An intra-atrial pump (IAP) was proposed that would be affixed to the atrial septum to support the compromised left ventricle (LV) without harming the ventricular tissue in patients with early-stage heart failure. The IAP is designed to operate in parallel with the LV, drawing blood from the left atrium and unloading the LV. In previous hydraulic studies, different blade geometries were tested for the IAP; however, it is important to know how the blade geometry affects the IAP's hemodynamic performance in the human cardiovascular system. In this study, a mock circulatory loop (MCL) with physiological response was used to evaluate the hemodynamic effects of IAP blade geometry and connection configuration in the human cardiovascular system. In a $2 \times 2$ study, two different blade geometries (with steep vs flat pressure/flow curves) were tested in two different connection configurations: the proposed configuration (left atrium to aorta) and the conventional configuration for LVADs (LV to aorta). We found that atrial cannulation is feasible and creates a beneficial hemodynamic environment, although it is inferior to the one created by ventricular cannulation. The steepgradient pump performed better than the flat-gradient pump in atrial insertion.

Preliminary design of the internal geometry in a minimally invasive left ventricular assist device under pulsatile-flow conditions.

PURPOSE: A minimally invasive, partial-assist, intra-atrial blood pump has been proposed, which would unload the left ventricle with a flow path from the left atrium to the arterial system. Flow modulation is a common strategy for ensuring washout in the pump, but it can increase power consumption because it is typically achieved through motor-speed variation. However, if a pump's performance curve had the proper gradient, flow modulation could be realized passively. To achieve this goal, we propose a pump performance operating curve as an alternative to the more standard operating point. METHODS AND RESULTS: Mean-line theory was employed to generate an initial set of geometries that were then tested on a hydraulic test rig at ~20,000 r/min. Experimental results show that the intra-atrial blood pump performed below the operating region; however, it was determined that smaller hub diameter and longer chord length bring the performance of the intra-atrial blood pump device closer to the operating curve. CONCLUSION: We found that it is possible to shape the pump performance curve for specifically targeted gradients over the operating region through geometric variations inside the pump.

Spatial resolution limits of multislice computed tomography (MS-CT), C-arm-CT, and flat panel-CT (FP-CT) compared to MicroCT for visualization of a small metallic stent.

RATIONALE AND OBJECTIVES: Small metallic stents are increasingly used in the treatment of cerebral aneurysms and for revascularization in ischemic strokes. Realistic three-dimensional datasets of a stent were obtained by using three x-ray-based imaging methods in current clinical use. Multislice-CT (MS-CT), C-arm flat detector-CT (C-arm CT, ACT), and flat panel-CT (FP-CT) were compared with high-resolution laboratory MicroCT scans that served as a reference standard. The purpose was to assess and compare the quality and accuracy of current clinical three-dimensional reconstructions of a vascular stents. MATERIAL & METHODS: A 3 × 20 mm Cypher stent was deployed in a straight polytetrafluoroethylene tube and filled with nondiluted iodine contrast and BaSO(4). MS-CT images of the static tube phantom and stent were acquired using GE LightSpeed VCT Series, C-arm CT images were obtained using Artis (DynaCT, Siemens), FP-CT were obtained using a preclinical research CT (GE), and MicroCT images were obtained using eXplore Locus SP (GE). DICOM datasets were analyzed using Amira and Matlab. RESULTS: Because of blooming effects, the maximum intensity projections (MIPs) and volume renderings generated from MS-CT showed significantly increased strut dimensions with no distinction between the regular struts and connector struts while the lumen diameter is artificially reduced. The shape of the reconstructed stent surface differed remarkably from the real stent. C-arm CT and FP-CT volume renderings more accurately represented the struts. Consistently capturing the structure of the connectors and the strut shape definition was highly threshold dependent. The stent lumen was about 30% underestimated by MS-CT when compared to MicroCT. CONCLUSION: The spatial resolution of current clinical CT for imaging of small metallic stents is insufficient to visualize fine geometrical details. Further improvement in the spatial resolution of clinical imaging technologies combined with better software and hardware for image postprocessing will be necessary for detailed structural analysis, evaluation of the stent lumen in vivo, and to permit accurate assessment of stent patency and early detection potential in-stent stenosis.

Sensitivity of digital thermal monitoring parameters to reactive hyperemia.

Both structural and functional evaluations of the endothelium exist in order to diagnose cardiovascular disease (CVD) in its asymptomatic stages. Vascular reactivity, a functional evaluation of the endothelium in response to factors such as occlusion, cold, and stress, in addition to plasma markers, is the most widely accepted test and has been found to be a better predictor of the health of the endothelium than structural assessment tools such as coronary calcium scores or carotid intima-media thickness. Among the vascular reactivity assessment techniques available, digital thermal monitoring (DTM) is a noninvasive technique that measures the recovery of fingertip temperature after 2-5 min of brachial occlusion. On release of occlusion, the finger temperature responds to the amount of blood flow rate overshoot referred to as reactive hyperemia (RH), which has been shown to correlate with vascular health. Recent clinical trials have confirmed the potential importance of DTM as an early stage predictor of CVD. Numerical simulations of a finger were carried out to establish the relationship between DTM and RH. The model finger consisted of essential components including bone, tissue, major blood vessels (macrovasculature), skin, and microvasculature. The macrovasculature was represented by a pair of arteries and veins, while the microvasculature was represented by a porous medium. The time-dependent Navier-Stokes and energy equations were numerically solved to describe the temperature distribution in and around the finger. The blood flow waveform postocclusion, an input to the numerical model, was modeled as an instantaneous overshoot in flow rate (RH) followed by an exponential decay back to baseline flow rate. Simulation results were similar to clinically measured fingertip temperature profiles in terms of basic shape, temperature variations, and time delays at time scales associated with both heat conduction and blood perfusion. The DTM parameters currently in clinical use were evaluated and their sensitivity to RH was established. Among the parameters presented, temperature rebound (TR) was shown to have the best correlation with the level of RH with good sensitivity for the range of flow rates studied. It was shown that both TR and the equilibrium start temperature (representing the baseline flow rate) are necessary to identify the amount of RH and, thus, to establish criteria for predicting the state of specific patient's cardiovascular health.

Digital thermal monitoring (DTM) of vascular reactivity closely correlates with Doppler flow velocity.

The noninvasive measurement of peripheral vascular reactivity, as an indicator of vascular function, provides a valuable tool for cardiovascular screening of at-risk populations. Practical and economical considerations demand that such a test be low-cost and simple to use. To this end, it is advantageous to substitute digital thermal monitoring (DTM) for the more costly and complex Doppler system commonly used for this measurement. A signal processing model was developed to establish the basis for the relationship between finger temperature reactivity and blood flow reactivity following a transient brachial artery occlusion and reperfusion protocol (reactive hyperemia). Flow velocity signals were acquired from the radial artery of human subjects via an 8 MHz Doppler probe while simultaneous DTM signals were acquired from a distal fingertip via DTM sensors. The model transforms the DTM temperature signals into normalized flow signals via a deconvolution method which employs an exponential impulse function. The DTM normalized flow signals were compared to simultaneous, low-frequency, normalized flow signals computed from Doppler sensors. The normalized flow signals, derived from DTM and Doppler sensors, were found to yield similar reactivity responses during reperfusion. The reactivity areas derived from DTM and Doppler sensors, indicative of hyperemic volumes, were found to be within +/- 15%. In conclusion, this signal processing model provides a means to measure vascular reactivity using DTM sensors, that is equivalent to that obtained by more complex Doppler systems.

Flow visualization techniques in a mock ventricle supported by a nonpulsatile left ventricular assist device.

Little is known about flow patterns in ventricles supported by continuous flow left ventricular assist devices (LVADs), and valuable information can be obtained with simple flow visualization experiments. We describe the application of several experimental techniques for the in vitro study of ventricular flow patterns (e.g., unsteadiness, vortical motions, stagnation regions) in the presence of a continuous flow LVAD. We used dye streaks, particle paths, and hydrogen bubble techniques to visualize fluid flow in an idealized, static, transparent mock ventricle attached to a Jarvik 2000 continuous flow LVAD. We recorded ventricular flow behavior at various pump speeds while independently adjusting pump flow (by varying the afterload) to emulate in vivo pump flow at various phases of the cardiac cycle. Changes in ventricular flow behavior at different pump flow rates may be of clinical relevance, because continuous flow pumps are extremely sensitive to inflow and outflow pressures and instantaneous pump flow varies significantly at different points throughout the cardiac cycle. Further work is needed to quantitatively compare the flow behavior of different continuous flow devices in a pulsatile ventricular model.

Stability of pulsatile blood flow at the ostium of cerebral aneurysms.

The strength and direction of blood flow into and within a cerebral aneurysm are important issues in developing effective interventional strategies to stabilize the aneurysm. We tested the hypothesis that there are significant major hemodynamic features that are common to many aneurysm flows of the type studied here. This was investigated by performing computational fluid dynamic simulations of flow near 7 cerebral aneurysms using geometrical data obtained from clinical CT scans. Our numerical simulations of flow across the ostium plane of an aneurysm show that in many cases there is relatively stable flow structure that is maintained over the phase of the pulsatile flow cycle. The two main features of this flow are (1) quasi-permanent regions of flow influx and efflux across the ostium plane exist, separated by a "virtual boundary", and (2) a helical vortex flow pattern within the aneurismal sac with swirl in two orthogonal cross-sectional planes. These numerical observations are consistent with in vitro experimental data from ultrasound color-Doppler velocimetry and other numerical and experimental studies. The observed flow patterns are found to occur in different types of aneurysms (bifurcation and sidewall), and can persist even after flow parameters are perturbed beyond the normal range of physiological flow conditions. These results suggest that in many cases, major aspects of the behavior of aneurismal hemodynamics for important classes of aneurysms can be learned from an analysis of steady, non-pulsatile flow, which is simpler and faster to simulate than time-dependent, pulsatile flow. An understanding of this fluid dynamical behavior may also prove useful in the design of stents, coils, and various other endovascular flow diverting devices.

Preload sensitivity of the Jarvik 2000 and HeartMate II left ventricular assist devices.

The in vitro sensitivity of continuous flow pumps to preload and afterload pressure has been well characterized. We compared flow in the Jarvik 2000 and HeartMate II continuous flow left ventricular assist devices (LVADs) at different inflow and outflow pressures and different pump speeds. This allowed us to measure the impact of a changing inflow pressure on the pump flow rate at different speeds but against a constant afterload. The resulting preload sensitivity curves showed that, overall, both LVADs have a mean preload sensitivity of 0.07 L/min/mm Hg in the physiologic ranges of pressures and flows encountered during normal operation. The HeartMate II pump had an increased preload sensitivity (up to approximately 0.1 L/min/mm Hg) as the preload was increased. The preload sensitivity of the Jarvik 2000 LVAD was more variable, having several peaks and troughs as the preload was increased. In future LVADs, improved preload sensitivity may allow passive regulation of pump output, optimize ventricular unloading, and decrease the risk of ventricular suction by the pump.

Continuous flow total artificial heart: modeling and feedback control in a mock circulatory system.

We developed a mock circulatory loop and used mathematical modeling to test the in vitro performance of a physiologic flow control system for a total artificial heart (TAH). The TAH was constructed from two continuous flow pumps. The objective of the control system was to maintain loop flow constant in response to changes in outflow resistance of either pump. Baseline outflow resistances of the right (pulmonary vascular resistance) and the left (systemic vascular resistance) pumps were set at 2 and 18 Wood units, respectively. The corresponding circuit flow was 4 L/min. The control system consisted of two digital integral controllers, each regulating the voltage, hence, the rotational speed of one of the pumps. The in vitro performance of the flow control system was validated by increasing systemic and pulmonary vascular resistances in the mock loop by 4 and 8 Wood units (simulating systemic and pulmonary hypertension conditions), respectively. For these simulated hypertensive states, the flow controllers regulated circuit flow back to 4 L/min within seconds by automatically adjusting the rotational speed of either or both pumps. We conclude that this multivariable feedback mechanism may constitute an adequate supplement to the inherent pressure sensitivity of rotary blood pumps for the automatic flow control and left-right flow balance of a dual continuous flow pump TAH system.