The current work investigated the stimulation induced spatial response of the cerebral microvascular network by reconstruction of the 3D microvascular morphology from in vivo two photon fluorescence microscopy (2PFM) volumes using an automated, model based tracking algorithm. In vivo 2PFM imaging of the vasculature in the forelimb representation of the primary somatosensory cortex of alpha-chloralose anesthetized rats was achieved via implantation of a closed cranial window, and intravascular injection of fluorescent dextran. The dilatory and constrictory responses of the cerebral microvascular network to functional stimulation were heterogeneous and depended on resting vascular radius and response latency. Capillaries experienced large relative dilations and constrictions, but the larger vessel absolute volume changes dominated the overall network cerebral blood volume change.

We demonstrated simultaneous intrinsic optical signal imaging (IOSI) and laser speckle contrast imaging (LSCI) using coherence modulation of vertical-cavity surface-emitting laser (VCSEL) diodes. The unique properties of VCSELs were exploited to deliver rapidly switched coherent and non-coherent illumination suitable for high resolution LSCI and IOSI, respectively. Utilizing three near-infrared VCSELs we were able to map changes in cortical blood oxygenation and flow during ischemia. Additionally, the subtle reflectance changes associated with cortical spreading depression were imaged using non-coherent VCSEL illumination. We are currently using two-photon laser-scanning microscopy to quantify the accuracy of LSCI and IOSI implementations. The small size and efficiency of VCSELs and modern photo diodes, makes the development of implantable dual-mode imaging devices feasible. Devices capable of chronic imaging of cortical hemodynamics could significantly enhance the range of studies available to neuroscientists and significantly aid clinicians postoperatively. The research presented in this thesis significantly furthers this objective.

The current work investigated the stimulation induced spatial response of the cerebral microvascular network by reconstruction of the 3D microvascular morphology from in vivo two photon fluorescence microscopy (2PFM) volumes using an automated, model based tracking algorithm. In vivo 2PFM imaging of the vasculature in the forelimb representation of the primary somatosensory cortex of alpha-chloralose anesthetized rats was achieved via implantation of a closed cranial window, and intravascular injection of fluorescent dextran. The dilatory and constrictory responses of the cerebral microvascular network to functional stimulation were heterogeneous and depended on resting vascular radius and response latency. Capillaries experienced large relative dilations and constrictions, but the larger vessel absolute volume changes dominated the overall network cerebral blood volume change.

We demonstrated simultaneous intrinsic optical signal imaging (IOSI) and laser speckle contrast imaging (LSCI) using coherence modulation of vertical-cavity surface-emitting laser (VCSEL) diodes. The unique properties of VCSELs were exploited to deliver rapidly switched coherent and non-coherent illumination suitable for high resolution LSCI and IOSI, respectively. Utilizing three near-infrared VCSELs we were able to map changes in cortical blood oxygenation and flow during ischemia. Additionally, the subtle reflectance changes associated with cortical spreading depression were imaged using non-coherent VCSEL illumination. We are currently using two-photon laser-scanning microscopy to quantify the accuracy of LSCI and IOSI implementations. The small size and efficiency of VCSELs and modern photo diodes, makes the development of implantable dual-mode imaging devices feasible. Devices capable of chronic imaging of cortical hemodynamics could significantly enhance the range of studies available to neuroscientists and significantly aid clinicians postoperatively. The research presented in this thesis significantly furthers this objective.

Cardiac valves are dynamic, sophisticated structures which interact closely with the surrounding hemodynamic environment. Altered mechanical stresses, including pressure, shear and bending stresses, are believed to cause changes in valve biology, but the cellular and molecular events involved in these processes are not well characterized. Therefore, the overall goal of this project is to determine the effects of pressure and shear stress on porcine aortic valve leaflets biology. Results from the pressure study showed that elevated constant pressure (140 and 170 mmHg) causes significant increases in collagen synthesis. The increases were 37.5% and 90% for 140 and 170 mmHg, respectively. No significant differences in DNA and sGAG synthesis were observed under constant pressure. In the cyclic pressure study, the effects of both pressure magnitude and pulse frequency were studied. With the frequency fixed at 1.167 Hz, collagen and sGAG synthesis increased proportionally with mean pressure level. At a fixed pressure level (80-120 mmHg), collagen and sGAG synthesis were slightly increased by 25% and 14% at 0.5 Hz, respectively. DNA synthesis was significantly increased by 72% at 2 Hz. An experiment combining high magnitude (150-190 mmHg) and high frequency (2 Hz) demonstrated significant increases in collagen and sGAG synthesis (collagen: 74%, sGAG: 56%), but no significant changes in cell proliferation. Shear levels ranging from 1 to 80 dyne/cm2 were studied. Scanning electron microscopy results indicated that 48 hrs exposure to shear stress did not alter the circumferential alignment of endothelial cells. Collagen synthesis was significantly enhanced at 9 and 25 dyne/cm2, but not different from static controls under other shear conditions. Leaflets denuded of the endothelium were exposed to identical shear stress and showed very different responses. Collagen synthesis was not affected at any shear levels, but sGAG content was increased at shear of 9, 25 and 40 dyne/cm2. Further studies showed that the increases in collagen synthesis under pressure or shear stress was concurrent with a decline in the expression and activities of cathepsins L and S. This converse relationship between collagen synthesis and cathepsin activity indicated that cathepsins might be involved in valvular ECM remodeling.

Accurate and reliable modeling of cardiovascular hemodynamics has the potential to improve understanding of the localization and progression of heart diseases, which are currently the most common cause of death in Western countries. However, building a detailed, realistic model of human blood flow is a formidable mathematical and computational challenge. The simulation must combine the motion of the fluid, the intricate geometry of the blood vessels, continual changes in flow and pressure driven by the heartbeat, and the behavior of suspended bodies such as red blood cells. Such simulations can provide insight into factors like endothelial shear stress that act as triggers for the complex biomechanical events that can lead to atherosclerotic pathologies. Currently, it is not possible to measure endothelial shear stress in vivo, making these simulations a crucial component to understanding and potentially predicting the progression of cardiovascular disease. In this thesis, an approach for efficiently modeling the fluid movement coupled to the cell dynamics in real-patient geometries while accounting for the additional force from the expansion and contraction of the heart will be presented and examined. First, a novel method to couple a mesoscopic lattice Boltzmann fluid model to the microscopic molecular dynamics model of cell movement is elucidated. A treatment of red blood cells as extended structures, a method to handle highly irregular geometries through topology driven graph partitioning, and an efficient molecular dynamics load balancing scheme are introduced. These result in a large-scale simulation of the cardiovascular system, with a realistic description of the complex human arterial geometry, from centimeters down to the spatial resolution of red-blood cells. The computational methods developed to enable scaling of the application to 294,912 processors are discussed, thus empowering the simulation of a full heartbeat. Second, further extensions to enable the modeling of fluids in vessels with smaller diameters and a method for introducing the deformational forces exerted on the arterial flows from the movement of the heart by borrowing concepts from cosmodynamics are presented. These additional forces have a great impact on the endothelial shear stress. Third, the fluid model is extended to not only recover Navier-Stokes hydrodynamics, but also a wider range of Knudsen numbers, which is especially important in micro- and nano-scale flows. The tradeoffs of many optimizations methods such as the use of deep halo level ghost cells that, alongside hybrid programming models, reduce the impact of such higher-order models and enable efficient modeling of extreme regimes of computational fluid dynamics are discussed. Fourth, the extension of these models to other research questions like clogging in microfluidic devices and determining the severity of co-arctation of the aorta is presented. Through this work, a validation of these methods by taking real patient data and the measured pressure value before the narrowing of the aorta and predicting the pressure drop across the co-arctation is shown. Comparison with the measured pressure drop in vivo highlights the accuracy and potential impact of such patient specific simulations. Finally, a method to enable the simulation of longer trajectories in time by discretizing both spatially and temporally is presented. In this method, a serial coarse iterator is used to initialize data at discrete time steps for a fine model that runs in parallel. This coarse solver is based on a larger time step and typically a coarser discretization in space. Iterative refinement enables the compute-intensive fine iterator to be modeled with temporal parallelization. The algorithm consists of a series of prediction-corrector iterations completing when the results have converged within a certain tolerance. Combined, these developments allow large fluid models to be simulated for longer time durations than previously possible. / Engineering and Applied Sciences

published_or_final_version / Medical Sciences / Master / Master of Medical Sciences

Mechanical stresses and flow dynamics alteration in a stented artery region are known to stimulate intimal thickening and increase the risk of restenosis, the closure of a revascularized artery. Particle imaging velocimetry (PIV) is an optical flow visualization technique that can be used to characterize the local flow dynamics around different stent structures. However, the usual cylindrical stent geometries present visualization difficulties when using an optical measurement technique such as the PIV technique. Using a flat configuration of a stent model presents advantages over the usual cylindrical model. A planar stent model makes data acquisition easier in planes cutting through the model due to its flat geometry that is compatible with the PIV planar flow investigation technique. Furthermore, with the planar stent configuration model velocity measurements and their associated flow features can be done without inducing refraction of the laser light sheet occurring with the cylindrical model's curvature. The refraction of light should be avoided since measurement errors and reflections are the resulting effects of this laser light plane deviation when passing through the curvature of a cylindrical stent model. / The spatial and temporal distribution of the Wall Shear Stress (WSS), which is believed to be of primary importance in the development of restenosis should be comparable between the flat and the cylindrical stent configuration models. The velocity and shear strain rate distributions will be compared between the flat and cylindrical stent configurations using computational fluid dynamics (CFD) simulations in order to analyse the feasibility of using a flat instead of a cylindrical version of the stent model for PIV experiments. It will be shown that for a physiological pulsatile flow the flat model yields results in shear strain rate spatial and temporal distribution that is comparable to the cylindrical model. A more PIV compatible, efficient and less refractive error prone validated flat model would be advantageous when several stent designs influence on the local hemodynamics around the strut geometries have to be studied quantitatively and optimized.

The potential role of lower limb blood pooling in reducing venous return to the heart during orthostasis and elevated venous pressure is investigated. This study compares lower limb capacitance, microvascular filtration, and peripheral resistance between a group of highly trained endurance athletes and a group of their sedentary peers. Seven endurance trained males were selected between the ages of 23-33 [(29.1 4.1 yr), mean SD]. The subjects weekly cycling mileage ranged from 80 to 150 miles per week with an average of 125 8.5 miles/week. Nine healthy, age-matched sedentary subjects (25.8 4.8 yr.) were selected for the control group, based upon their reporting they had not participated in repeated lower-body or cardiovascular exercise in the months prior to their study. Results show that both subject groups had similar calf venous capacitances, rates of capillary fluid filtration, and local flow shunting (vascular resistance change) in response to venous thigh occlusion and 70 head-up tilt (HUT). The only significant difference found between groups was the cyclist groups smaller rise in heart rate in response to HUT. The findings of this study suggest that cyclists are not predisposed to orthostatic intolerance due to any changes in lower limb function.

Background: Although Bevacizumab (BVZ) and Sunitinib (SNT) prolong survival in cancer patients, an unanticipated side-effect is cardiotoxicity. Early indices of left ventricular (LV) systolic dysfunction would be useful to address the cardiac safety of anti-cancer drugs. Objective: Whether cardiac biomarkers, tissue velocity imaging (TVI), and/or strain rate (SR) can detect early cardiac dysfunction. Methods: A total of 95 C57Bl/6 mice received one of the following drug regimens: i) 0.9% saline; ii) BVZ; or iii) SNT and followed for 14 days. Serial blood pressure, high sensitivity troponin I (hsTnI), and echocardiography were performed. Results: BVZ- and SNT-treated mice demonstrated an increase in mean arterial blood pressure, hsTnI, cardiac apoptosis, and loss of cell integrity. TVI and SR values confirmed early LV systolic dysfunction at day 8, compared to conventional LVEF at day 13. Conclusions: Novel imaging techniques can detect early LV systolic dysfunction in a model of drug-mediated cardiomyopathy.