Much research has been conducted to obtain detailed and accurate information of blood flow field for clarification and advanced diagnosis of circulatory diseases. In order to reproduce complicated real flow field, we have proposed the concept of integration of measurement and computation, and developed the ultrasonic-measurement-integrated (UMI) simulation [1, 2]. In this study, we propose MR-measurement-integrated (MR-MI) simulation by applying the methodology to the MR (Magnetic Resonance) measurement. In the MR-MI simulation, the difference between the computational result and the PC MRI (Phase-Contrast Magnetic Resonance Imaging) measurement data of the three-dimensional velocity vector field is fed back to the numerical simulation (see Fig. 1). The objective in this study is to investigate the characteristics of MR-MI simulation by numerical experiment dealing with a model steady blood flow field in a cerebral aneurysm.

Soft hydrogels serving as substrates for cell attachment are used to culture many types of cells. The mechanical properties of these gels influence the cell morphology, growth and differentiation. Young’s moduli of such hydrogels were calculated by our group using spherical beads of 400μm radius embedded in micro-liter samples [1]. This concept of inclusion-based measurement has also been utilized by other researchers to measure local viscoelastic responses [2, 3]. The present research involves scaling down the size of the inclusions to the level at which local inhomogeneties in the media become significant, thus allowing us to evaluate the local elastic behavior of the hydrogels.

From an averaged point of view, blood can often be treated computationally as a single-phase fluid of non-Newtonian character. Such a model may be appropriate if information regarding the bulk motion of the blood is all that is required. If, however, one seeks to describe the mechanisms leading to diseases such as thrombosis in the presence of foreign surfaces such as prosthesis, accurate predictions of platelet behavior in the dynamic environment of the blood are required. There are several effects that necessitate a careful treatment of platelet dynamics. For example, it is well known that the presence of red blood cells has a significant impact on radial distribution of platelets as well as the shear stress experienced by the platelets [1]. Therefore, the paths of and forces experienced by individual platelets are to be determined in order to predict the location and predilection for thrombus formation. However, since the length scales of the platelets are much smaller than the typical dimensions of the flow regions through which blood flows, it is not possible to capture platelet dynamics in a single-scale computation. Therefore, a multiscale technique for incorporating the dynamics of platelets and platelet-RBC interactions into large-scale flow simulations is required. We therefore examine a suspension of ellipsoidal and circular rigid particles that are representative of red blood cells and platelets carried in a Newtonian fluid to study the interaction or red blood cells and platelets.

The mechanical function of soft collagenous tissues is inherently multiscale, with the tissue dimension being in the centimeter length scale and the underlying collagen network being in the micrometer length scale. The strong sensitivity of soft tissues to network orientation and fiber-fiber interactions necessitates the incorporation of both scales in a multiscale methodology [1,2] in order to predict their mechanical response. The computational demands of a multiscale methodology, however, are strenuous and the use of parallel processing and adaptivity are required for succeeding acceptable solution times.

Primary among the mechanical factors linked with abdominal aortic aneurysm (AAA) rupture is peak wall stress, frequently quantified as either the maximum principal or Von Mises stress exerted along the diseased arterial wall. Intraluminal pressure, as an impinging normal force on the wall, has been hypothesized as the dominant influence on this stress and thus the majority of numerical modeling studies of AAA mechanics have focused on static computational solid stress (CSS) predictions [1,2]. Unfortunately, retrospective studies comparing the magnitude of wall stress with the failure strength of the aneurysmal wall have yet to consistently predict the outcome for patient-specific AAAs [3,4]. Previous studies have shown that hemodynamics also plays a significant role in both the biological and mechanical factors that exist within AAAs. In the present investigation, partially and fully coupled fluid-structure interaction (p-FSI and f-FSI, respectively) computations of patient-specific AAA models are presented and compared to identify the effect of fluid flow in the biomechanical environment of these aneurysms.

3D physiologic geometry of St. Jude Medical (SJM) valve after implantation was simulated with non-Newtonian two-phase blood model. The simulation used the unsteady Reynolds averaged Navier-Stokes (URANS) approach and the Wilcox k-ω turbulent model. Platelet stress accumulation and the resulting platelet damage were calculated from the results. Thrombogenic potential of two bileaflet MHV geometries was conducted using fluid-structure interaction (FSI) computation. Two commercially available valve geometries, SJM and ATS, which differ mostly in their hinge design, were simulated in a straight geometry with sudden expansion downstream of the valve. The thrombogenic potential of the two valves was based on computed wall shear stresses on the leaflets and cumulative shear stress on multiple particles released during forward and reverse flow phases. Platelet stress accumulation along pertinent trajectories from the FSI studies indicated that the SJM valve has a higher thrombogenic potential then the ATS valve. Flow patterns generated by the valve are conducive to platelet activation provide optimal conditions for activated platelets to interact with each other and form aggregates are hypothesized to be the source of thromboemboli formation, increasing the risk for cardioembolic stroke. The new damage model developed was utilized to estimate the effects of repeated passages and platelet senescence on this thrombogenic potential. Flow and pressure effects on a cell like a platelet can be well represented by a continuum mechanics model down to the order of the μm level. However, the molecular effects of adhesion/aggregation bonds are on the order of nm. Thus we also adopt a discrete particles dynamics (DPD) approach in which the macroscopic model provides information about the flow induced stresses that may activate blood cellular constituents. This multiscale modeling approach concentrates on flow regions in prosthetic devices like MHVs and cardiovascular pathologies that have a high propensity to activate platelets and form aggregates. Preliminary simulations of blood flow in simple geometries using this approach, which widely departs from the traditional continuum approach, is successful in generating viscous blood flow velocity distributions in these geometries.

High Intensity Focused Ultrasound (HIFU) has shown considerable promise as a minimally-invasive technique for various therapeutic applications such as tumor ablation and vessel cauterization. The efficacies of these HIFU procedures depend on various operational parameters such as total acoustic power, pulse duration and transducer dimensions. In this study, the effect of total acoustic power on the tissue temperature rise is studied both experimentally and numerically. Experimentally, HIFU ablations, at different acoustic powers, were carried out in a tissue mimicking material embedded with thermocouples. Temperature rise measured from the in-vitro experiments were then validated with the numerical computations. Results show that experimental and numerical temperature rise match accurately. Our numerical model was able to predict the peak temperature rise within ∼12% of the experimental results. Results also show that the tissue temperature rise is linearly proportional to the input acoustic power. For the acoustic power levels considered in this study, the results suggest that acoustic non-linearity does not play a major role on the tumor ablation procedure.

Numerical analysis of the cardiovascular system can provide medical researchers with information that cannot (easily) be measured in a clinical setting and may contribute to a better comprehension and insight into the pathophysiology of cardiovascular diseases. In addition, numerical models offer a computational environment in which both new and existing medical procedures and devices can be tested and optimized, which is both cost effective and patient friendly. The continuous improvement of computational methods, computational power and medical imaging techniques strengthens the general belief that computational models will eventually find their way into clinical practice, what can be seen in todays trend toward more realistic patient-specific models. This new trend brings along more complex geometries and larger computational domains for which the quality of the corresponding meshes can become a bottle neck for the image-based analysis, both in computation time as in accuracy of the solution. Therefore, the Treemesh method has been developed by De Santis et al. [1], an algorithm to generate high-quality hexahedral meshes in a multi-block structured way. However, when the geometry becomes too complicated, a high-quality Treemesh for the fluid domain can become hard to generate. In this work, the extended Treemesh (XTreemesh) method has been developed, an algorithm to auto-generate a high-quality unstructured hexahedral mesh for both the fluid domain and the structural domain of complex image-based vascular tree geometries.

Because the left ventricle (LV) is not completely emptied during systole, oxygenated blood from the left atrium interacts with residual blood from preceding cycles. It is hypothesized that LV flow is optimal for transporting blood under normal conditions; yet proving this remains a challenge. Furthermore, clinical evaluation of LV hemodynamics has tremendous diagnostic importance for patients with cardiomyopathy. We have performed Doppler-echocardiography on 6 patients with dilated cardiomyopathy and 6 healthy volunteers. Using novel processing of the color-Doppler data, bi-directional velocity field maps in the apical long axis plane were derived. Resulting flow data was used to perform Lagrangian coherent structure (LCS) computation, which enabled novel characterization of the transport topology in the LV during filling and ejection. This framework was used to quantify stasis in the LV, which can be used as a surrogate for diagnosing pumping deficiencies and thrombosis risk. This framework also enables characterization of LV vortices, which have previously received much attention using Eulerian characterizations. The framework presented here uncovers the well-defined boundaries to both E-wave and A-wave filling vortices, which has not been previously reported.

The mechanisms and the causes of the in-stent restenosis process in coronary arteries are not fully understood. One of the most relevant phenomena, which seems to be associated to this process, is an altered hemodynamics in the stented wall region [1]. In vivo local measurements of velocities and their gradients in human coronary arteries are very difficult and can hardly be applied to successfully investigate the fluid dynamic field [1]. Alternatively, virtual models of blood flow in patient-specific coronary arteries allow the study of local fluid dynamics and the computation of the wall shear stress (WSS) and other quantities which can be related to the risk of restenosis.

Medical image processing has emerged as a powerful way to simulate fluid flows through realistic models of complex patient-specific geometries without relying upon simplifying geometric approximations. However, image-based flow modeling processes traditionally involve several steps (e.g. image segmentation, surface mesh generation, volumetric flow mesh generation, and finally computational simulation) that must often be performed using separate pieces of software. This work presents an alternative methodology in which the entire image-based flow modeling process takes place on a Cartesian domain with the image embedded as an implicit surface, circumventing the need for complex surface meshes and body-fitted flow meshes. The complete framework is demonstrated with flow calculations performed in a computed tomography (CT) image reconstruction of an intracranial aneurysm (ICA). Flow calculations are compared against calculations performed following a standard multi-step route using the Vascular Modeling Toolkit (VMTK) [1, 2] and Fluent™ (Ansys, Inc., Lebanon, NH).

Direct measurement of cartilage contact is not feasible, necessitating the use of musculoskeletal models to estimate the internal soft tissue loading associated with movement 2 . However, current computational models used to simulate movement often utilize simplified joint kinematic constraints which can only provide an estimate of the net joint reaction force 1 . Detailed finite element models of the knee have been created 14 which can provide estimates of cartilage tissue stress, but are too computationally expensive to solve within the context of a whole body simulation of movement. Discrete element analysis (DEA) provides a viable alternative for rapidly computing contact stress patterns in movement 15 , However, even DEA models can be computationally expensive as high resolution polygonal meshes are needed to accurately represent complex cartilage geometries such as that seen on the femoral condyles and tibia plateau. In a DEA model, much of the computation time is spent querying for contact between triangles of two polygonal surfaces. The objective of this study was to investigate the potential for using graphic processor computation (GPU) to expedite contact detection 5 . To do this, we use a contact detection algorithm that pre-constructs hierarchical bounding-volumes (BVH) of the target body to increase the efficiency of contact detection. We then show that implementing a parallel computational version of this algorithm on the GPU greatly speed up performance and thus make it more viable to simulate cartilage contact within the context of human movement.

Due to the systemic nature of the response of the human body to vascular trauma, accurate simulation may require the consideration of larger portions of the vasculature than have previously been modeled. A patient-specific 3D model, composed of 78 arteries extending from the neck to the toes, was constructed to meaningfully represent many of the major arteries in the systemic circulation. Recently developed boundary conditions are applied at the 78 outlets to realistically represent the remaining vasculature not resolved by the 3D domain. The pulsatile blood flow simulation is started on an initial boundary layer mesh with a few million elements which is then adapted to a higher resolution by applying solution-based adaptivity. The effects of the mesh resolution on flow features are studied. Pressure, velocity and wall shear stress results will be presented.

We have designed, created, and evaluated a translating stage system (TSS) used for Micro Particle Image Velocimetry (μ-PIV) measurements surrounding the tip of a migrating semi-infinite bubble in a glass capillary tube. This study is relevant to acute respiratory distress syndrome (ARDS), which is characterized by pulmonary airway collapse and fluid occlusion. Subsequent airway reopening, driven by mechanical ventilation, may generate damaging mechanical stresses on the airway walls that can result in ventilator-induced lung injury (VILI). Our goal in this study is to characterize and demonstrate a system for mechanically canceling the mean forward velocity of a migrating bubble, allowing continuous measurement near the tip region of interest. This system will be used to validate our previous boundary element computations in a surfactant-free system [1] and further explore airway reopening phenomenon related to surfactant physiochemical interaction.

Computer simulations of vascular tissue adaptation under various physiological and pathological conditions have emerged as a new area of research and aided researchers in their understanding of stress-mediated growth and remodeling (G&R) in these structures. With advances in computational biomechanics and biomedical imaging techniques, combinations of these advanced methods will provide promising tools for medical diagnosis and surgical planning in the future (e.g., [1]). Recently Figueroa et al. [2] presented a new computational framework that brings advances in computational biosolid and biofluid mechanics together in order to exploit new information on the biology of vascular growth and remodeling (G&R). Although the framework presented in their paper was generalized for simplicity, they did illustrate the effectiveness of this framework by applying it to a fusiform aneurysm growth with idealized geometry. In the present work, we employ this framework and test it on an anatomically realistic model of abdominal aortic aneurysm (AAA) growth. Similarly to Figueroa et al., when the stress-mediated kinetics only depends on intramural stress, the shape of the aneurysm and the expansion rate are similar to the results from the computation without using an iterative loop. However, we expect that when the stress-mediated kinetics depends on either shear or other hemodynamic components, the evolution of an AAA can change significantly.

Generation of a complete map of arterial wall mechanical properties can improve treatment of cardiovascular diseases via contributions to design of patient specific vascular substitutes used to alleviate atherosclerosis and stenoses, which are predominant in arterial pathways (i.e., abdominal aorta, carotids, or femoral arteries). Clinically useful estimation of arterial properties from patient data requires both efficient algorithms and models that are both complex enough to capture clinically important properties and simple enough to allow rapid computation. In this study, we used mechanical models accounting for both elastic and viscoelastic wall deformation to analyze how vessel properties and associated model parameters vary with artery type. It is known that for the aorta wall, deformation is dominated by nonlinear elastic dynamics, while for the smaller vessels (e.g. the carotid artery) deformation is dominated by viscoelastic responses. The latter is correlated with composition of the vessels; the aorta contains significantly less smooth muscle cells (∼40%) than the carotid artery (∼60%), and has significantly more elastin (see Fig 1).

The computation of the blood flow in arterial models requires the solution of the incompressible Navier-Stokes equations [1,2]. The geometry of these models is characterized by complex branching tubular domains. The main challenge with this kind of geometries is that the convergence rate of the pressure Poisson solver is dominated by the graph depth of the computational grid. The purpose of this paper is to present a novel technique (deflated conjugate gradients algorithm) for accelerating the solution of the pressure Poisson equation that is especially suited for elongated domains.

In this paper, an integrated numerical approach is introduced to determine the human brain responses when the head is exposed to blast explosions. The procedure is based on a 3D non-linear finite element method (FEM) that implements a simultaneous conduction of explosive detonation, shock wave propagation, and blast-brain interaction of the confronting human head. Due to the fact that there is no reported experimental data on blast-head interactions, several important checkpoints should be made before trusting the brain responses resulting from the blast modeling. These checkpoints include; a) a validated human head FEM subjected to impact loading; b) a validated air-free blast propagation model; and c) the verified blast waves-solid interactions. The simulations presented in this paper satisfy the above-mentioned requirements and checkpoints. The head model employed here has been validated again impact loadings. In this respect, Chafi et al. [1] have examined the head model against the brain intracranial pressure, and brain’s strains under different impact loadings of cadaveric experimental tests of Hardy et al. [2]. In another report, Chafi et al. [3] has examined the air-blast and blast-object simulations using Arbitrary Lagrangian Eulerian (ALE) multi-material and Fluid-Solid Interaction (FSI) formulations. The predicted results of blast propagation matched very well with those of experimental data proving that this computational solid-fluid algorithm is able to accurately predict the blast wave propagation in the medium and the response of the structure to blast loading. Various aspects of blast wave propagations in air as well as when barriers such as solid walls are encountered have been studied. With the head model included, different scenarios have been assumed to capture an appropriate picture of the brain response at a constant stand-off distance of nearly 80cm (2.62 feet) from the explosion core. The impact of brain response due to severity of the blast under different amounts of the explosive material, TNT (0.0838, 0.205, and 0.5lb) is examined. The accuracy of the modeling can provide the information to design protection facilities for human head for the hostile environments.

Atherosclerosis is characterized by lipid accumulation in the arterial wall, followed by an inflammatory response. Plaque formation is generally observed near bifurcations in coronary arteries. The composition of atherosclerotic plaques depends on the location, and it was hypothesized that blood flow induced shear stress influences plaque composition 2 . To study the impact of shear stress on atherosclerotic disease in human coronary arteries, we developed a technique that enables us to generate 3D lumen reconstruction based on multislice computer tomography (MSCT) and intravascular ultrasound (IVUS).We describe two approaches to generate 3D reconstructions of human coronary artery bifurcations and apply them to coronary segments with bifurcations. We will evaluate the effect on shear stress distribution and its relationship to wall thickness.

Sudden rupture of vulnerable atherosclerotic plaques is a major contributing cause of acute myocardial infarctions and ischemic strokes [1]. A vulnerable plaque is defined as a plaque with a large necrotic core and a thin fibrous cap. In order to characterize, and further comprehend, plaque vulnerability from a mechanical point of view, the rupture process can be seen as a mechanical process which occurs when the tensile stress in the fibrous cap exceeds its material strength. Biomechanical stress modeling of plaques using the Finite Element Method (FEM) has been used as a tool to provide insight in the stress distribution in plaques and shows potential to facilitate identification of vulnerable plaques using novel biomechanics-based risk-stratification criteria [1, 2, 3]. Accurate stress prediction using computational modeling depends on a number of factors including material models, computational methods, initial conditions and accurate reconstruction of the plaque geometry from ex vivo or in vivo imaging. The latter received limited attention and lacks a critical evaluation. In case of 3D modeling, the arterial geometry is typically reconstructed by stacking MRI or histology slices in the axial direction and interpolating the geometry without consensus on minimal requirements of inter slice distance (axial sampling distance) [4]. Due to time constraints during the imaging procedure, especially MRI suffers from a limited axial resolution (typically an inter slice distance of 1–3 mm), which might compromise accurate geometry reconstruction which could in turn influence resulting stress computations.