David H. Frakes

Physics-Driven CFD Modeling of Complex Anatomical Cardiovascular Flows—A TCPC Case Study

Recent developments in medical image acquisition combined with the latest advancements in numerical methods for solving the Navier-Stokes equations have created unprecedented opportunities for developing simple and reliable computational fluid dynamics (CFD) tools for meeting patient-specific surgical planning objectives. However, for CFD to reach its full potential and gain the trust and confidence of medical practitioners, physics-driven numerical modeling is required. This study reports on the experience gained from an ongoing integrated CFD modeling effort aimed at developing an advanced numerical simulation tool capable of accurately predicting flow characteristics in an anatomically correct total cavopulmonary connection (TCPC). An anatomical intra-atrial TCPC model is reconstructed from a stack of magnetic resonance (MR) images acquired in vivo. An exact replica of the computational geometry was built using transparent rapid prototyping. Following the same approach as in earlier studies on idealized models, flow structures, pressure drops, and energy losses were assessed both numerically and experimentally, then compared. Numerical studies were performed with both a first-order accurate commercial software and a recently developed, second-order accurate, in-house flow solver. The commercial CFD model could, with reasonable accuracy, capture global flow quantities of interest such as control volume power losses and pressure drops and time-averaged flow patterns. However, for steady inflow conditions, both flow visualization experiments and particle image velocimetry (PIV) measurements revealed unsteady, complex, and highly 3D flow structures, which could not be captured by this numerical model with the available computational resources and additional modeling efforts that are described. Preliminary time-accurate computations with the in-house flow solver were shown to capture for the first time these complex flow features and yielded solutions in good agreement with the experimental observations. Flow fields obtained were similar for the studied total cardiac output range (1–3 l/min); however hydrodynamic power loss increased dramatically with increasing cardiac output, suggesting significant energy demand at exercise conditions. The simulation of cardiovascular flows poses a formidable challenge to even the most advanced CFD tools currently available. A successful prediction requires a two-pronged, physics-based approach, which integrates high-resolution CFD tools and high-resolution laboratory measurements.

Human Operator Performance Testing Using an Input-Shaped Bridge Crane

The payload oscillation inherent to all cranes makes it challenging for human operators to manipulate payloads quickly, accurately, and safely. An input-shaping controller was implemented on a large bridge crane at the Georgia Institute of Technology to reduce crane payload oscillation. The crane was used to study the performance of human operators as they drove the crane through obstacle courses. An image processing system was implemented to track the movement of the crane payload. Data from these experiments show that operators performed manipulation tasks faster, safer, and more effectively when input shaping was utilized to reduce payload sway.

Patient-specific surgical planning and hemodynamic computational fluid dynamics optimization through free-form haptic anatomy editing tool (SURGEM)

The first version of an anatomy editing/surgical planning tool (SURGEM) targeting anatomical complexity and patient-specific computational fluid dynamics (CFD) analysis is presented. Novel three-dimensional (3D) shape editing concepts and human–shape interaction technologies have been integrated to facilitate interactive surgical morphology alterations, grid generation and CFD analysis. In order to implement “manual hemodynamic optimization” at the surgery planning phase for patients with congenital heart defects, these tools are applied to design and evaluate possible modifications of patient-specific anatomies. In this context, anatomies involve complex geometric topologies and tortuous 3D blood flow pathways with multiple inlets and outlets. These tools make it possible to freely deform the lumen surface and to bend and position baffles through real-time, direct manipulation of the 3D models with both hands, thus eliminating the tedious and time-consuming phase of entering the desired geometry using traditional computer-aided design (CAD) systems. The 3D models of the modified anatomies are seamlessly exported and meshed for patient-specific CFD analysis. Free-formed anatomical modifications are quantified using an in-house skeletization based cross-sectional geometry analysis tool. Hemodynamic performance of the systematically modified anatomies is compared with the original anatomy using CFD. CFD results showed the relative importance of the various surgically created features such as pouch size, vena cave to pulmonary artery (PA) flare and PA stenosis. An interactive surgical-patch size estimator is also introduced. The combined design/analysis cycle time is used for comparing and optimizing surgical plans and improvements are tabulated. The reduced cost of patient-specific shape design and analysis process, made it possible to envision large clinical studies to assess the validity of predictive patient-specific CFD simulations. In this paper, model anatomical design studies are performed on a total of eight different complex patient specific anatomies. Using SURGEM, more than 30 new anatomical designs (or candidate configurations) are created, and the corresponding user times presented. CFD performances for eight of these candidate configurations are also presented.

Coupling Pediatric Ventricle Assist Devices to the Fontan Circulation: Simulations with a Lumped-Parameter Model

In pediatric ventricular assist device (VAD) design, the process of matching device characteristics and dimensions to the relevant disease conditions poses a formidable challenge because the disease spectrum is more highly varied than for adult applications. One example arises with single-ventricle congenital defects, which demand palliative surgeries that create elevated systemic venous pressure and altered pulmonary hemodynamics. Substituting a mechanical pump as a right ventricle has long been proposed to eliminate the associated early and postoperative anomalies. A pulsatile lumped-parameter model of the single-ventricle circulation was developed to guide the preliminary design studies. Two special modules, the pump characteristics and the total cavopulmonary connection (TCPC) module, are introduced. The TCPC module incorporates the results of three-dimensional patient-specific computational fluid dynamics calculations, where the pressure drop in the TCPC anastomosis is calculated at the equal vascular lung resistance operating point for different cardiac outputs at a steady 60/40 inferior vena cava/superior vena cava flow split. Preliminary results obtained with the adult parameters are presented with no ventricle remodeling or combined larger-size single ventricle. A detailed literature review of single-ventricle function is provided. Coupling a continuous pump to the single-ventricle circulation brought both the pulmonary and systemic venous pressures back to manageable levels. Selected VADs provided an acceptable cardiac output trace of the single left ventricle, after initial transients. Remodeling of the systemic venous compliance plays a critical role in performance and is included in this study. Pulsatile operation mode with rotational speed regulation highlighted the importance of TCPC and pulmonary artery compliances. Four different pumps and three patient-specific anatomical TCPC pathologies were studied. Magnitudes of the equivalent TCPC resistances were found to be comparable to the vascular resistances of the normal baseline circulation, significantly affecting both the VAD design and hemodynamics.

A New Method for Registration-Based Medical Image Interpolation

A new technique is presented for interpolating between grey-scale images in a medical data set. Registration between neighboring slices is achieved with a modified control grid interpolation algorithm that selectively accepts displacement field updates in a manner optimized for performance. A cubic interpolator is then applied to pixel intensities correlated by the displacement fields. Special considerations are made for efficiency, interpolation quality, and compression in the implementation of the algorithm. Experimental results show that the new method achieves good quality, while offering dramatic improvement in efficiency relative to the best competing method.

Application of an adaptive control grid interpolation technique to morphological vascular reconstruction

The problem of interslice magnetic resonance (MR) image reconstruction arises in a broad range of medical applications. In such cases, there is a need to approximate information present in the original subject that is not reflected in contiguously acquired MR images because of hardware sampling limitations. In the context of vascular morphology reconstruction, this information is required in order for subsequent visualization and computational analysis of blood vessels to be most effective. Toward that end we have developed a method of vascular morphology reconstruction based on adaptive control grid interpolation (ACGI) to function as a precursor to visualization and computational analysis. ACGI has previously been implemented in addressing various problems including video coding and tracking. This paper focuses on the novel application of the technique to medical image processing. ACGI combines features of optical flow-based and block-based motion estimation algorithms to enhance insufficiently dense MR data sets accurately with a minimal degree of computational complexity. The resulting enhanced data sets describe vascular geometries. These reconstructions can then be used as visualization tools and in conjunction with computational fluid dynamics (CFD) simulations to offer the pressure and velocity information necessary to quantify power loss. The proposed ACGI methodology is envisioned ultimately to play a role in surgical planning aimed at producing optimal vascular configurations for successful surgical outcomes.

Measuring glomerular number and size in perfused kidneys using MRI

The goal of this work was to nondestructively measure glomerular (and thereby nephron) number in the whole kidney. Variations in the number and size of glomeruli have been linked to many renal and systemic diseases. Here, we develop a robust magnetic resonance imaging (MRI) technique based on injection of cationic ferritin (CF) to produce an accurate measurement of number and size of individual glomeruli. High-field (19 Tesla) gradient-echo MR images of perfused rat kidneys after in vivo intravenous injection of CF showed specific labeling of individual glomeruli with CF throughout the kidney. We developed a three-dimensional image-processing algorithm to count every labeled glomerulus. MRI-based counts yielded 33,786 ± 3,753 labeled glomeruli (n = 5 kidneys). Acid maceration counting of contralateral kidneys yielded an estimate of 30,585 ± 2,053 glomeruli (n = 6 kidneys). Disector/fractionator stereology counting yielded an estimate of 34,963 glomeruli (n = 2). MRI-based measurement of apparent glomerular volume of labeled glomeruli was 4.89 × 10(-4) mm(3) (n = 5) compared with the average stereological measurement of 4.99 × 10(-4) mm(3) (n = 2). The MRI-based technique also yielded the intrarenal distribution of apparent glomerular volume, a measurement previously unobtainable in histology. This work makes it possible to nondestructively measure whole-kidney glomerular number and apparent glomerular volumes to study susceptibility to renal diseases and opens the door to similar in vivo measurements in animals and humans.

Total Cavopulmonary Connection Flow With Functional Left Pulmonary Artery Stenosis Angioplasty and Fenestration In Vitro

Background— In our multicenter study of the total cavopulmonary connection (TCPC), a cohort of patients with long-segment left pulmonary artery (LPA) stenosis was observed (35%). The clinically recognized detrimental effects of LPA stenosis motivated a computational fluid dynamic simulation study within 3-dimensional patient-specific and idealized TCPC pathways. The goal of this study was to quantify and evaluate the hemodynamic impact of LPA stenosis and to judge interventional strategies aimed at treating it. Methods and Results— Simulations were conducted at equal vascular lung resistance, modeling both discrete stenosis (DS) and diffuse long-segment hypoplasia with varying degrees of obstruction (0% to 80%). Models having fenestrations of 2 to 6 mm and atrium pressures of 4 to 14 mm Hg were explored. A patient-specific, extracardiac TCPC with 85% DS was studied in its original configuration and after virtual surgery that dilated the LPA to 0% stenosis in the computer medium. Performance indices improved exponentially (R2>0.99) with decreasing obstruction. Diffuse long-segment hypoplasia was ≈50% more severe with regard to lung perfusion and cardiac energy loss than DS. Virtual angioplasty performed on the 3-dimensional Fontan anatomy exhibiting an 85% DS stenosis produced a 61% increase in left lung perfusion and a 50% decrease in cardiac energy dissipation. After 4-mm fenestration, TCPC baffle pressure dropped by ≈10% and left lung perfusion decreased by ≈8% compared with the 80% DS case. Conclusions— DS <60% and diffuse long-segment hypoplasia <40% could be considered tolerable because both resulted in only a 12% decrease in left lung perfusion. In contrast to angioplasty, a fenestration (right-to-left shunt) reduced TCPC pressure at the cost of decreased left and right lung perfusion. These results suggest that pre-Fontan computational fluid dynamic simulation may be valuable for determining both the hemodynamic significance of LPA stenosis and the potential benefits of intervention.

Influence of stent configuration on cerebral aneurysm fluid dynamics

Embolic coiling is the most popular endovascular treatment available for cerebral aneurysms. Nevertheless, the embolic coiling of wide-neck aneurysms is challenging and, in many cases, ineffective. Use of highly porous stents to support coiling of wide-neck aneurysms has become a common procedure in recent years. Several studies have also demonstrated that high porosity stents alone can significantly alter aneurysmal hemodynamics, but differences among different stent configurations have not been fully characterized. As a result, it is usually unclear which stent configuration is optimal for treatment. In this paper, we present a flow study that elucidates the influence of stent configuration on cerebral aneurysm fluid dynamics in an idealized wide-neck basilar tip aneurysm model. Aneurysmal fluid dynamics for three different stent configurations (half-Y, Y and, cross-bar) were first quantified using particle image velocimetry and then compared. Computational fluid dynamics (CFD) simulations were also conducted for selected stent configurations to facilitate validation and provide more detailed characterizations of the fluid dynamics promoted by different stent configurations. In vitro results showed that the Y stent configuration reduced cross-neck flow most significantly, while the cross-bar configuration reduced velocity magnitudes within the aneurysmal sac most significantly. The half-Y configuration led to increased velocity magnitudes within the aneurysmal sac at high parent-vessel flow rates. Experimental results were in strong agreement with CFD simulations. Simulated results indicated that differences in fluid dynamic performance among the different stent configurations can be attributed primarily to protruding struts within the bifurcation region.

New techniques for the reconstruction of complex vascular anatomies from MRI images

The accurate representation of two-dimensional images in three dimensions has become important for many medical imaging applications and for cardiac magnetic resonance imaging (MRI) in particular. Reconstruction methods applied after data acquisition can produce three-dimensional information from two-dimensional data and make applications such as surgical planning more effective. Current reconstruction techniques usually demand contrast agents, and can suffer due to poor segmentation and sampling constraints that cause surface irregularities and distort dimensions. The novel technique presented here for anatomical modeling uses adaptive control grid interpolation (ACGI) to approximate data not captured by scanning, and a progressive shape-element segmentation technique to complete reconstruction. Quantitative validations conducted on models of pediatric cardiac malformations have confirmed the theoretical advantages of this technique, and that higher quality is achieved than with competing methods based on geometric parameters. Vascular diameters from reconstructions showed errors of less than 1% for a known geometry as compared to over 9% for competing methods. Qualitatively, models produced with the new methodology displayed substantial improvement over alternatives. Approximately 50 rare cardiac structures, including surgically altered Fontan and atypical aortic anatomies, have been reconstructed. All data used to create these reconstructions were acquired using standard pulse sequences and without contrast agents. Benefits of the new technique are particularly evident when complex vascular configurations complicate reconstruction. The proposed methodology enables a powerful tool allowing physicians to analyze and manipulate highly accurate and clearly presented vascular structures in an interactive medium.