2007 STUDENT EVENTS:

Jevin Scrivens, Ph.D. Defense, Thursday, June 28th, 2007, 1:30p.m., Molecular Sciences Building, Room 3201
Advisor: Stephen P. Deweerth, Ph.D.  (Georgia Institute of Technology)
Co-Advisor: Lena H. Ting, Ph.D. (Georgia Institute of Technology)
Committee: Wayne J. Book, Ph.D. (Georgia Institute of Technology), Young-Hui Chang, Ph.D. (Georgia Institute of Technology) and
T. Richard Nichols, Ph.D. (Georgia Institute of Technology

TITLE: "The Interactions of Stance Width and Feedback Control Gain: A Modeling Study of Bipedal Postural Control"

Compliant postural strategies employed by animal neuromuscular systems may provide robust solutions and inspiration for robotic and prosthetic design. In traditional robotic systems for manufacturing applications, control algorithms are designed to enable precise control over joint trajectories, overriding the natural dynamics of the mechanical system. However, in devices that both interact with and emulate humans, these algorithms produce very stiff systems that can create potentially dangerous levels of contact force. Thus, applying traditional control algorithms to an unstable, upright, bipedal configuration such as in the humanoid robot AsimoT (Honda Corp.) results in joint stiffness, and energetic expenditures that far exceed those found in humans and animals.  In contrast, animals display fluid movements with compliant behaviors that generate relatively low reaction forces when contacted or perturbed. The fluidity of these movements results from the relatively low actuator forces combined with the passive dynamics of the mechanical system. Taking motivation from biological systems the overall objective of this work is to improve robot control design by developing control methodology that can generate postural responses to perturbation in a stable and compliant manner. Towards this objective, we investigated compliant operation under changing geometry. More specifically, we investigated the effect of stance width on postural control under perturbation, and determined the changes required in a compliant controller as stance width changes. The results of this study may help provide insight to the methods of control for a task that bipeds and quadrupeds must accomplish on a frequent basis.

TITLE: " The Regulation of Conformation and Binding Kinetics of Integrin a L b 2"

The interaction mediated by integrin a L b 2 and its ligand plays major role in many immune responses by regulating leukocyte adhesion. This study investigated the conformational regulation of a L b 2 and the effects of conformational change on the ligand binding of a L b 2 . Micropipette adhesion frequency assay was used to measure the two-dimensional binding affinity and kinetics of a L b 2 on K562 cells and neutrophils. The conformations of a L b 2 were regulated by mutations, antibodies, small molecule antagonists, as well as divalent cations. Our results indicated that the change in binding affinity and off-rate was mostly due to the a L I domain conformational change. Without affecting the I domain conformation, the extension of a L b 2 only increases the on-rate for several fold by providing a better orientation and accessibility of the molecule on cell surface. The binding characteristics of divalent cations to I domain MIDAS and other metal ion binding sites in a L b 2 are determined by the nature of divalent cations, Mn 2+ has higher binding affinity to the metal ion binding sites than Mg 2+ . The conformation of I domain also affected the binding of divalent cations. Open and intermediate I domains have higher binding affinity for Mn 2+ and Mg 2+ than WT and closed I domains. Divalent cations dissociate from I domain MIDAS very slowly but from those metal ion binding sites that important for conformational change of a L b 2 rapidly. One of the most important biological processes mediated by a L b 2 and other b 2 integrins is the recruitment and migration of neutrophils during inflammation. The activation of b 2 integrins by E-selectin binding to neutrophils in this process was also investigated. The binding of E-selectin, but not P- or L-selectin, activates b 2 integrins in a timescale of ~ 5 seconds and the activation may require the crosslink of E-selectin ligands. These results provide insights into the relationship between the conformational change and the function of a L b 2 and most importantly would contribute to the understanding of integrin regulation mechanisms.

TITLE: " Developing a Minimally Invasive Sustained Release System for Glioma Therapy"

Malignant brain tumor is one of the most lethal forms of cancers. In the United States alone, approximately 20,500 new cases of primary malignant brain and central nervous system tumors are expected to be diagnosed in 2007 with 12,740 deaths estimated. Treatment of malignant brain tumor remains a major challenge despite recent advance in surgery and other adjuvant therapies, such as chemotherapy. The failure of potential effective chemotherapeutics for brain tumor treatment is usually not due to the lack of potency of the drug, but rather can be attributed to lack of therapeutic strategies capable of overcoming blood brain barrier for effective delivery of drug to the brain tumor. In this thesis, we developed a minimally invasive sustained release system for glioma therapy. The present study was initiated in an effort to incorporated Doxorubicin (DOX) loaded PLGA particle into an agarose gel, which can provide a continuous release of DOX locally to the tumor site. DOX, a toposiomearase II inhibitor, is not currently used clinically for brain tumor treatment because when delivered systemically it does not cross BBB. Our hydrogel particle system can overcome this shortcoming of DOX. The results from this study demonstrate that the DOX/PLGA particle gel system can maintain the bioactivity of DOX and sustained release DOX for at least 15 day in vitro. The result of in vivo study showed the DOX/PLGA particle gel treated group had significantly extend the medium survival of 9L glioma bearing rat from 21 days to 29 days. Therefore, the success experience of this local and sustained delivery device might benefit the development of future glioma therapy strategy.

Rhima Coleman, Ph.D. Defense, Monday, June 25th, 2007, 10:30a.m., IBB Building, Room 1128
Advisor: Dr. Robert Guldberg (Georgia Institute of Technology)
Committee Members: Dr. Regis O'Keefe (University of Rochester Medical Center), Dr. Barbara Boyan (Georgia Institute of Technology), Dr. Ray Vito (Georgia Institute of Technology) and Dr. Ravi Bellamkonda (Georgia Institute of Technology)

TITLE: " DEVELOPMENT OF A SMALL ANIMAL MODEL TO STUDY TISSUE ENGINEERING STRATEGIES FOR GROWTH PLATE DEFECTS"

The growth plate is a cartilaginous tissue responsible for the longitudinal growth of long bones. Given the recent rise in the number of growth plate injuries and the variability in success of current therapies, there is a significant need for a greater understanding of growth plate injury pathology and the development of improved treatment strategies. Prior to developing treatment strategies for growth plate injury repair, it is essential to first understand the interconnection between alterations in growth plate morphology and subsequent limb deformities. To that end, we have established a surgical defect model of growth plate injury in Sprague Dawley rats and developed a novel technique to quantitatively monitor growth plate morphology in health and disease using microcomputed tomography (micro-CT) imaging. In an effort to develop a tissue engineering treatment strategy for growth plate injury, the role of monolayer expansion, 3D scaffold, and growth factor regimen in the chondrogenic differentiation of rat BMSCs was also examined. This research study has demonstrated the utility of micro-CT as a non-invasive imaging modality for assessing growth plate injury and repair. This work has also provided an improved understanding of the interrelationship of monolayer expansion, 3D culture environment, and growth factor regimen in BMSC chondrogenic differentiation. Finally, this work suggests that an injectable in situ gelling hydrogel is a feasible method for decreasing limb length discrepancies. However, neither implantation of agarose alone into the defect nor the inclusion of BMSCs fully corrects growth disruption.

TITLE: "NONLINEAR AND NETWORK CHARACTERIZATION OF BRAIN FUNCTION USING FUNCTIONAL MRI DATA"

Functional magnetic resonance imaging (fMRI) has emerged as the method of choice to non-invasively investigate brain function in humans. Though brain is known to act as a nonlinear system, there has not been much effort to explore the applicability of nonlinear analysis techniques to fMRI data. Also, recent trends have suggested that functional localization as a model of brain function is incomplete and efforts are being made to develop models based on networks of regions to understand brain function. Therefore this thesis attempts to introduce the twin concepts of nonlinear dynamics and network analysis into a broad spectrum of fMRI data analysis techniques.   First, we characterized the nonlinear univariate dynamics of fMRI noise using the concept of embedding to explain the origin of tissue-specific differences of baseline activity in the brain. The embedding concept was extended to the multivariate case to study nonlinear functional connectivity in the distributed motor network during resting state and continuous motor task. The results showed that the nonlinear method may be more sensitive to the desired gray matter signal. Subsequently, the scope of connectivity was extended to include directional interactions using Granger causality. An integrated approach was developed to alleviate the confounding effect of the spatial variability of the hemodynamic response and graph theory was employed to characterize the network topology. This methodology proved effective in characterizing the dynamics of cortical networks during motor fatigue. The nonlinear extension of Granger causality showed that it was more robust in the presence of confounds such as baseline drifts. Finally, we utilized the integration of the spatial correlation function to study connectivity in local brain networks. We showed that our method is robust and can reveal interesting information including the default mode network during resting state. Application of this technique to anesthesia data showed dose dependent suppression of local connectivity in the default mode network, particularly in the frontal areas. Given the body of evidence emerging from our studies, nonlinear and network characterization of fMRI data seems to provide novel insights into brain function.

TITLE: "Multicontrast MRI of Atherosclerotic Plaques: Acquisition, Characterization and Reconstruction"

Atherosclerosis is a major contributor to cardiovascular disease, which is the leading cause of death in western countries. The characterization of vulnerable atherosclerotic plaque is crucial for the management and treatment of the clinical sequelae of atherosclerosis. MRI studies of atherosclerosis require the use of combined MR contrast mechanisms, known as multicontrast MRI, to achieve plaque characterization. However, there are some challenges yet to be addressed in order to apply this technique to clinical settings. Firstly, technical limitations still restrict the in-vivo usage of this technique, especially for coronary plaque imaging. Since most of the MR plaque imaging and analysis techniques were developed based on ex-vivo scans, one of the specific aims in the current research is to assess the effect of vessel preservation on multicontrast MRI. To achieve this goal, an MR compatible tissue culture chamber was fabricated to simulated in-vivo conditions for plaque tissues during the MR acquisition. Results from this study indicate preservation of plaque tissues does not affect the multicontrast MRI analysis, and thus techniques developed under ex-vivo conditions still apply in in-vivo studies. Secondly, efficient means to perform automatic plaque characterization based on multicontrast MRI data is still lacking. The current study, therefore, proposes an a priori information enhanced characterization (PIEC) technique to fill this gap. Specifically, PIEC divided characterization into two sub-steps: classification and labeling. In the classification step, we developed a fuzzy c- means based technique to segment multicontrast MR images in the presence of noise and field inhomogeneity. In the labeling step, a Bayesian approach is used to tag the segmentation result obtained from the first step using tissues’ quantitative MR properties. Thirdly, acquiring multicontrast MRI can be time- consuming, which precludes its clinical applications. To accelerate the acquisition, “shared k-space” reconstruction techniques were interrogated to reduce the acquisition time without sacrificing MR image resolution and SNR. Our study suggests that these techniques are promising in expediting the multicontrast MRI acquisition.

Diane de Julien de Zelicourt , Ph.D. Proposal, Monday, June 11th, 2007, 3:00p.m., U.A. Whitaker Bldg., Room 2110
Advisor: Ajit P. Yoganathan (Georgia Institute of Technology and Emory University)
Committee Members: Fotis Sotiropoulos, Ph.D., (University of Minnesota), Don P. Giddens, Ph.D., (Georgia Institute of Technology and Emory University), Pedro del Nido, M.D., (Boston Children's Hospital), Shiva Sharma, M.D., (Pediatric Cardiology Service) and Robert Taylor, Ph.D., (Georgia Institute of Technology and Emory University )

TITLE: " Development of an Unstructured Immersed-Boundary Method; Application to Pulsatile Simulations in Patient-Specific Fontan Anatomies"

Single ventricle congenital heart defects affect about 2 babies per 1000 live births. In these patients, the oxygenated blood coming from the lungs and the deoxygenated blood coming back from the rest of the body mix in the single ventricle leading to acute cyanosis. The concept of a total right ventricular bypass, first introduced by Fontan and Baudet in 1971 , is a palliative surgical procedure aimed at separating the systemic and pulmonary circulations thereby eliminating venous blood mixing. In its current form the Fontan procedure is performed in three stages, ultimately resulting in what is called a total cavopulmonary connection (TCPC) where the inferior (IVC) and superior (SVC) vena cavae are anastomosed directly onto the pulmonary arteries (PAs) thus bypassing the right side of the heart. Two options are currently used for the 3 rd and last stage of the procedure, which connects the IVC to the PAs: intra-atrial tunnels where the IVC is routed through the atrium and extra-cardiac conduits where the IVC goes around the heart. Whether to choose one option over the other is one of the most debated clinical questions in the Fontan arena. So far, clinical studies have been inconclusive leading to balanced pros and cons for both approaches. Bioengineering studies may bring in some additional critical information by bringing in a fundamental understanding of the hemodynamics associated with each option and by trying to optimize their design.   Over the past decades, computational fluid dynamics (CFD) has arisen as an attractive option to accurately model such complex biomedical phenomena. However, few CFD studies have been able to account for the geometrical complexity of patient-specific anatomies , and none has investigated the impact of in vivo pulsatile flows. A major difficulty for handling patient-specific anatomies, stems from their geometrical complexity. This thesis thus explores the applicability of state of the art CFD techniques towards modeling in vivo pulsatile fluid dynamics of patient-specific TCPC. This work will first involve developing a CFD solver that easily handles geometries of arbitrary complexity and is optimized so as to allow for pulsatile simulations to be carried in a practical time-frame. The tool will then be applied to Fontan patients of our database to investigate the impact of pulsatility in intra-atrial and extra-cardiac TCPCs.

Tracy Denison , Ph.D. Proposal, Friday, June 8th, 2007, 4:00p.m., Clinic B Building, Room B4300, Emory University
Advisor: Barbara D. Boyan, Ph.D. (Dept of Biomedical Engineering, Georgia Institute of Technology and Emory University)
Committee Members: Hanjoong Jo, Ph.D. (Dept of Biomedical Engineering, Georgia Institute of Technology and Emory University) , Athanassios Sambanis, Ph.D. (Dept of Chemical & Biomolecular Engineering, Georgia Institute of Technology) , Zvi Schwartz, D.M.D., Ph.D. (Dept of Biomedical Engineering, Georgia Institute of Technology and Emory University) and Lynda F. Bonewald, Ph.D. (Dept of Oral Biology, [School of Dentistry] University of Missouri-Kansas City)

TITLE: " The Effect of Fluid Shear Stress on Chondrocytes in Endochondral Ossification"

While many studies have considered the role of mechanical forces on articular and meniscal cartilage maintenance, less attention has been given to the growth plate. Cartilage tissue contains a hydrated matrix and mechanical loading can induce shear stress upon chondrocytes as water is expelled during loading. The overall goal of this study is to determine if fluid shear stress can influence the maturation of chondrocytes as they progress along uniquely defined stages of endochondral ossification. A better understanding in this area will contribute to the basic understanding of the growth plate, as well as potentially progress tissue engineering strategies that involve the growth plate, bone fracture callus healing, or osteochondral plugs. This project will focus on the general hypothesis that fluid shear stress modulates the differentiation or phenotypic expression of resting zone chondroctyes in the growth plate. This study will focus on three specific aims. Specific Aim 1 will use the chondrogenic ATDC5 cell line to develop a cell culture system with sensitivity to the Vitamin D metabolite 24,25(OH)2D3, indicating comparability with the resting zone chondrocyte. Specific Aim 2 will compare the responses of both the ATDC5 cultured cells and primary resting zone chondrocytes to fluid shear stress. Specific Aim 3 will use the ATDC5 model to investigate the role of Dentin Matrix Protein 1 (DMP1), a protein observed to be produced by hypertrophic growth plate chondrocytes and to be upregulated by mechanical stress in osteocytes, to determine if it may play a unique role in any mechanotransduction effects involved in endochondral ossification.

TITLE: "Delivery of Cdc42, Rac1, and BDNF to Promote Axonal Outgrowth after Spinal Cord Injury"

Physical injury to the spinal cord often leads to permanent functional loss. After injury, a series of events occur around the lesion site, including the deposition of growth cone inhibitory astroglial scar tissue containing chondroitin sulfate proteoglycans (CSPG)-rich regions. Therapeutic strategies are primarily directed at encouraging axons to extend through these inhibitory regions for regeneration to occur. The work presented in this dissertation investigates the effect of three proteins, constitutively active (CA)-Cdc42, CA-Rac1, and brain-derived neurotrophic factor (BDNF) on axonal outgrowth through CSPG-rich inhibitory regions after spinal cord injury. Cdc42 and Rac1 are members of the Rho GTPase family and BDNF is a member of the neurotrophin sub-family. These three proteins affect the actin cytoskeleton dynamics, positively, thereby, promoting axonal outgrowth. An in vitro modified stripe assay and an in vivo dorsal over-hemisection model, creating a ~2 mm defect, were used to investigate whether delivered Cdc42, Rac1, and BDNF, either individually or in combination, promoted axonal outgrowth through the CSPG-rich regions. Local sustained delivery of the protein(s) was achieved by using an in situ gelling hydrogel scaffold containing microtubules loaded with the proteins implanted into the spinal cord cavity. The results from this study demonstrate that the treated groups had significantly higher percentage of axons from the corticospinal tract (CST) that traversed the CSPG-inhibitory regions, as well as penetrate the glial scar compared to the controls. Treatment with BDNF , CA -Cdc42, and CA-Rac1 also reduced the inflammatory response that occurs after injury around the lesion site. Therefore, the local delivery of CA-Cdc2, CA-Rac1, and BDNF, individual and combination exhibits the ability of axons to extend through CSPG inhibitory regions, as well as reduce the glial scar components.

Megan Oest , Ph.D. Defense, Tuesday, June 5th, 2007, 10:00a.m., IBB Building, Room 1107
Advisor: Robert E. Guldberg, Ph.D. (Georgia Institute of Technology)
Committee Members: Andrés J. García, Ph.D., David J. Mooney, Ph.D., W. Robert Taylor, M.D., Ph.D. and Ravi Bellamkonda, Ph.D.

TITLE: " Dual Osteogenic and Angiogenic Growth Factor Delivery as a Treatment for Segmental Bone Defects"

Bone loss due to traumatic injury or disease is a significant clinical problem for orthopaedic surgeons to address. Current treatments for bone defects rely heavily on processed cadaveric allografts, although use of single growth factor delivery has recently come into the clinic. Allograft treatments are limited by their high failure rate, and the efficacy of single growth factor delivery in an absorbable collagen sponge for regeneration of large defects has not been established. Use of a single treatment that incorporates both structural and biological functions has not been widely investigated. The goals of this dissertation were to establish a robust, reproducible small animal bone defect model for evaluation of tissue-engineered treatments and to test the efficacy of low dose growth factor co-delivery to repair these defects. A critically-sized segmental femoral defect model was established. A novel micro-computed tomography (micro-CT)-based method for quantifying vascularity was adapted for use in the rat model, and was able to image vascularity within the defect site. Evaluation of co-delivered BMP-2 and TGF- b 3 revealed a positive dose-dependent response. Addition of low doses of VEGF, an angiogenic growth factor, did not enhance mineralization or vascularization within the implanted constructs over delivery of osteogenic growth factors alone. Modification of the construct geometry may have permitted more extensive revascularization of the defects. A significant positive correlation between bone volumes measured using in vivo micro-CT and torsional mechanical properties (stiffness and maximum torque) was established. This work establishes the efficacy of low-dose growth factor therapy for bone regeneration, provides a novel in vivo model for bone defect repair, and quantified a predictive relationship between bone repair parameters measured in vivo and mechanical properties measured post-mortem

TITLE: "Interaction of surface energy and microarchitecture in determining cell and tissue response to biomaterials"

Biomaterials are widely used in medical practice to help maintain, improve or restore the diseased tissues or organs. The successful integration of biomaterials with host tissue depends on substratum surface properties, as well as host tissue quality and its regulatory environment. The overall goal of this dissertation is to incorporate these three factors to achieve better biomaterial-host tissue interactions. Important surface properties include surface topography, surface energy, chemical composition and surface charge. We designed a new titanium (Ti) substratum with modified surface chemical composition by preventing the contamination when in contact with the atmosphere. The new Ti surface has lower carbon contamination and promotes osteoblast differentiation phenotype. The osteogenic effect is synergistic with micrometer and sub-micrometer scale surface structures. To further investigate the effects of bone quality on peri-implant bone formation, we developed a novel mouse femoral medullary bone formation model. This new model will facilitate research evaluating the effects of biomaterial surface treatments in host animals with deficient bone development, including genetically engineered mice. Finally, we studied sexual dimorphism in the response of osteoblasts to systemic regulatory hormones 1a,25-dihydroxyvitamin D 3 and 17ß-estradiol. The results showed intrinsic differences in male and female osteoblasts with respect to their differentiation and their responses to hormones, suggesting that host chromosomal sex should be considered in biomaterial research. Taken together, this research provides fundamental information on biomaterial surface properties and the regulation of host tissue response, which are important in guiding biomaterial design and evaluation.

TITLE: " Three-Dimensional Distribution of Stress in Models of Human Coronary Atherosclerotic Plaque Based on Acrylic Histologic Sections"

Each year in the United States over a million people experience a myocardial infarction. The majority of these attacks are caused by coronary artery plaque cap rupture with subsequent thrombus formation. Because rupture is a mechanical event and the tendency of a plaque to rupture is due in part to increases in the mechanical stresses in the fibrous cap, mechanical analyses are important to understanding plaque stability. Results from mechanical analyses are capable of providing clinically relevant information pertaining to plaque stability assessments. Histology is the only method capable of identifying plaque features that are associated with vulnerability. Therefore, minimally distorted histologic sections should serve as a basis for constructing the computation models used in mechanical analyses. Further, because substantial longitudinal variations in geometry and mechanical properties often exist in plaques, computational models should be three-dimensional (3-D). Finally, given the complex geometries of atherosclerotic plaques and the fact that they are composed of different materials, the finite element (FE) method should be used to determine the distribution of stress under physiological loading. Until now, a critical need has existed to determine the distribution of stress in 3-D FE models of human coronary atherosclerotic plaques based on minimally distorted histologic sections. In this research study, a method to measure and correct for distortions caused by acrylic histologic processing was first created. The devised strain-based method yields a limited set of parameters needed for a first order correction. Thus, corrections can be easily implemented using FE methods. Next, a methodology to create 3-D finite FE models of human coronary atherosclerotic plaques was developed. This methodology accounts for longitudinal variations in geometry and mechanical properties and uses stable acrylic histologic sections as a basis for model construction. Models of plaques, ranging in disease severity, were generated using the developed methodology. Lastly, the distributions of stress in these models were obtained and the effects of some plaque features on stresses were determined. Results from this study confirm that morphological description of a plaque is not sufficient to predict plaque rupture. The findings suggest that in many cases the 3-D stress field within a plaque must be known in order to assess plaque stability. Finally, the results show that patient specific models must be developed if the 3-D stress field within a plaque is to be determined.

TITLE: "R egulatory mechanisms in the chondrogenesis of mesenchymal progenitors: the roles of cyclic tensile loading and cell -matrix interactions"

Cartilage tissue engineering represents an exciting potential therapy for providing permanent and functional regeneration of healthy cartilage tissues, but these treatment options have yet to be successfully implemented in a clinical setting. One of the primary obstacles for cartilage engineering is obtaining a sufficient supply of cells capable of regenerating a functional cartilage matrix. Mesenchymal progenitors can easily be isolated from multiple tissues, expanded in vitro, and possess a chondrogenic potential, but it remains unclear what types or combinations of signals are required for lineage-specific differentiation and tissue maturation. The overall goal of this dissertation was to investigate how the coordination of biochemical stimuli with cues from mechanical forces and the extracellular matrix regulate the chondrogenesis of bone marrow stromal cells (BMSCs). These studies explored the potential for cyclic tensile loading and chondrogenic factors, TGF- b 1 and dexamethsone, to promote fibrochondrocyte-specific differentiation of BMSCs. The application of cyclic tensile displacements to cell-seeded fibrin constructs promoted fibrochondrocyte patterns of gene expression and the development of a fibrocartilage-like matrix. These responses were influenced by the specific loading conditions examined and the differentiation state of the BMSCs. Additionally, the roles of integrin adhesion and cytoskeletal organization in BMSC differentiation were examined within engineered hydrogels presenting controlled densities of biomimetic ligands. Adhesion to the arginine-glycine-aspartic acid (RGD) motif inhibited chondrogenesis in a density-dependent manner and was influenced by interactions with the f-actin cytoskeleton. Together, this research provided fundamental insights into the regulatory mechanisms involved in the chondrogenesis of mesenchymal progenitor cells.

TITLE: "Mechanics of the Mitral Valve after Surgical Repair: An In-vitro Study"

Mitral valve disease is a prevalent problem both as a primary and secondary pathology. It continues to be an important cause of global disease load, with an increased incidence of congenital heart defects in the developing world, and degenerative and ischemic heart disease in the western world. Valve replacement has markedly improved the natural history of the disease, but long-term survival rates are alarmingly low. Today, mitral valve repair/reconstruction by surgical intervention is the procedure of choice over replacement. Although repair is theoretically and intellectually more appealing, there is concern about its capacity to produce predictable, durable, superior results compared with replacement. A rational approach to this problem depends on the thorough understanding of the mitral valve function and mechanics after surgical reconstruction. The goal of this research is to develop pathological models of mitral valve disease and simulate surgical repair procedures on these models in an in-vitro left heart bioreactor using native porcine mitral valves. The hemodynamic function of the valve pre and post repair is quantified using measurable end-points such as valve leakage volume, effective orifice area, valve coaptation length and tenting area. The loading on the valve under physiological, pathological and post-repair conditions is quantified by measuring the leaflet strains, leaflet curvatures and the forces acting on the mitral valve chordae using novel pre-validated engineering devices and techniques.

TITLE: "Role of spontaneous bursts in functional plasticity and spatiotemporal dynamics of dissociated cortical cultures"

What changes in our brain when we learn? This is perhaps the most intriguing question of science in this century. In an attempt to learn more about the inner workings of neural circuitry, I studied cultured 2-dimensional networks of neurons on multi-electrode arrays (MEAs). MEAs are ideal tools for studying long-term neural ensemble activity because many individual cells can be studied continuously for months, through electrical stimulation and recording. One of the most prominent patterns of activity observed in these cultures is network-wide spontaneous bursting, during which most of the active electrodes in the culture show elevated firing rates. We view the persistence of spontaneous bursting /in vitro/ as a sign of arrested development due to deafferentation. Substituting distributed electrical stimulation for afferent input transformed the activity in dissociated cultures from bursting to more dispersed spiking, reminiscent of activity in the adult brain. Burst suppression reduced the variability in neural responses making it easier to induce and detect functional plasticity caused by tetanic stimulation. This suggests that spontaneous bursts interfere with the effects of external stimulation and that a burst-free environment leads to more stable connections and predictable effects of tetanization. Moreover, our culture models continuously receive input stimulation in the form of background electrical stimulation, and so better resemble the intact brain than isolated (non-continuously stimulated) cultures. The proportion of GABAergic neurons in the cultures was significantly increased (p<1e-2, paired t-test) after burst-quieting for 2 days, suggesting that burst suppression operated through the homeostatic control of inhibitory neurotransmitter levels. We also studied the role of spontaneous bursts as potential carriers of information in the network by clustering these spatiotemporally diverse bursts. Spontaneous burst clusters were stable over hours and tetanic stimulation significantly reorganized the distribution of the clusters. In summary, this body of work explores the rules of network-level functional plasticity and provides the input (electrical stimulation) - output (spatiotemporal patterns) mappings for behavioral studies in embodied hybrid systems. The results of this study may also have clinical implications in the development of sensory prostheses and treatment of diseases of aberrant network activity such as epilepsy.

TITLE: "Coated Microneedles and Microdermabrasion For Transdermal Delivery"

Delivery of drugs into the human body is necessary for the drugs to produce their therapeutic effect. When the common oral route is unsuitable, hypodermic injections are typically used. However, they cause pain, needle phobia, and pose safety risks by causing transmission of blood borne pathogens like HIV and hepatitis B viruses.    Transdermal patches provide an attractive and easily accessible alternative delivery method. However, the topmost layer of the skin, the stratum corneum, offers a formidable transport barrier. Different chemical and physical enhancement methods have been investigated to overcome this barrier but with limited success. Therefore, delivery of the new class of biopharmaceutics is still a challenge, and they continue to be injected into the human body. The skin is also an immunological organ and can orchestrate immune responses. In particular it has antigen presenting dendritic cells called the Langerhans cells that can initiate strong immunological responses. Transdermal immunization has been shown to produce robust immune responses against hepatitis B antigen, more importantly with a reduced dose.    Microneedles, sharp micron sized structures that can pierce the skin, provide a largely painless and versatile method for transdermal delivery. In addition, microdermabrasion, an FDA approved method used extensively for cosmetic applications, has potential to selectively remove the stratum corneum and enable topical drug delivery. The goals of this doctoral thesis were to develop and characterize (i) drug coated microneedles and (ii) microdermabrasion for transdermal drug delivery.    For developing coated microneedles, first, pain from solid stainless steel microneedles was characterized in human subjects to determine the effect of microneedle dimensions on pain. Next, a microneedle drug coating process was developed to study the versatility of the coatings and their in vitro performance. Then a rational basis was developed to design coating solution formulations to obtain uniform and thick coatings and to control the coating mass deposited on microneedles. Lastly, a strong T-cell response and tumor rejection was demonstrated in mice, after immunization with hepatits C DNA-vaccine coated onto microneedles. For microdermabrasion the effect of microdermabrasion operating conditions on skin layers was characterized in vivo on monkeys and humans, and delivery of sodium fluorescein and MVA (Modified Vaccinia Ankara) virus was demonstrated through microdermabraded skin in the monkeys.    Results from this research will promote understanding of coatings of micro-structures and the effects of microdermabrasion on skin. More broadly, this research will advance the field of transdermal drug delivery by enabling delivery of hydrophilic molecules and large molecular weight compounds including proteins, DNA and other biotherapeutics through the skin. The outcomes of this doctoral research also have potential to significantly impact mass immunizations by providing safer and less painful alternatives to hypodermic needle-based immunizations.

TITLE: "Near-Infrared Quantum Dots for In Vivo Imaging of Cancer"

Nanotechnology is a new multidisciplinary approach to science aiming to shrink the size of devices and tools to the nanometer dimension, and to assemble atoms and molecules into larger, complex structures with unique functions.  Among the many medically-relevant innovations already generated from nanotechnology, quantum dots (QDs) are unique in their broad applicability in the detection and study of many diseases, most notably cancer. QDs are fluorescent nanoparticles (typically 2-8 nm diameter) that can strongly emit light in the near-infrared spectrum, in which biological tissue is relatively transparent and emits very little background autofluorescence.  The focus of this proposal is the use of near-infrared QDs as novel probes for in vivo imaging of cancer.  Preliminary work has resulted in the preparation of brightly fluorescent QDs with emission wavelengths spanning the visible range and the near-infrared.  The surfaces of these nanoparticles have been tailored to be widely tunable in both size and charge, and these probes have been found to remain fluorescent after injection into living mice, visualized via whole-body near-infrared imaging.  The properties of these QDs, including emission wavelength, overall size, and bioaffinity ligand specificity, will be studied systematically to determine the efficacy of these probes as targeted contrast agents for the sensitive, high-resolution detection of cancer.  Additionally, proof-of-concept experiments will evaluate their potential use as histological stains for cancer tissue sections, and for use as intracellular probes for living cells.

TITLE: "Simplifying a Motoneuron Model: How Does Altering Mechanistic Complexity Affect the Model?"

Mathematical models of neuronal activity have been a hallmark of neuroscience for decades, as they provide an excellent platform for rigorous examination of hypotheses of neuronal function. However, the nonlinear dynamics of many of these models make their internal interactions difficult to understand. In addition, as our understanding of neuroscience has increased, the size and therefore the computational demand of neuron models has also increased. This has led many researchers to reduce these more complex models to simpler forms. However, although these simpler models may be easier to understand and more computationally efficient, what has been lost by reducing model complexity? T he goal of this project is to analyze how the replacement of complex mechanisms with simpler degenerate mechanisms affects the robustness, sensitivity robustness, flexibility, and output correlations of a motoneuron model. We propose that through the use of parameter searching, sensitivity analysis, and multivariate analysis, these properties can be measured. The approach relies upon using the previously mentioned techniques to compare three motoneuron models, each with differing levels of complexity. To do this, we will analyze the properties of a motoneuron model with three degenerate spiking mechanisms: Markov-based sodium channels, Hodgkin-Huxley sodium channels, and a simple threshold mechanism. The analysis will provide insight into how the model changes when mechanistic complexity is decreased and may lead to methods capable of helping modelers better manage  neuronal complexity.

TITLE: "Roles of Membrane Rafts in CD32A Mediated formation of Phagocytic Contact Area"

Membrane rafts are highly dynamic heterogeneous sterol- and sphingolipid-rich micro-domains on cell surfaces. They are generally believed to provide residency for cell surface molecules (e.g., adhesion and signaling molecules) and scaffolding to facilitate the functions of these molecules such as membrane trafficking, receptor transport, cell signaling, and endocytosis. Using immuno-fluorescent laser scanning confocal and Interference Refractive Microscopy (IRM), we studied the spatial and temporal distributions of membrane rafts and surface receptors, signaling molecules, and cell organelles during the formation of phagocytic contact areas. K562 cells, which naturally express CD32A, a cell surface receptor for the Fc portion of Immuno-globulin ???IgG), was chosen as a model for neutrophils. An opsonized target was modeled using a glass supported lipid bilayer reconstituted with IgG. CD32A was found to cluster and co-localize with detergent resistant membranes (DRMs). Placing the K562 cells on the lipid bilayer triggered a process of contact area formation that includes binding between receptors and ligands, their recruitment to the contact area, a concurrent DRM movement to and concentration in the contact area, and transport of CD32A, IgG, and DRMs to the Golgi complex. Experiment was repeated in the presence of agents known to disrupt detergent resistant membranes (DRMs), dissolve actin microfilaments, and inhibit myosin motor activity, which abolished the CD32A clusters and prevented the contact area formation. The procedures were repeated using micro-beads coated with a lipid bilayer reconstituted with IgG as the opsonized target instead of the glass supported lipid bilayer. Disruption of membrane rafts, salvation of the actin cytoskeleton, and inhibition of myosin II activity were found to inhibit phagocytosis. These data suggest membrane rafts play several important roles in CD32A mediated phagocytosis including pre-clustering CD32A, transport of CD32A to the phagocytic cup, and transport of the opsonized target towards the Golgi complex.

TITLE: "Three-Dimensional Extracellular Matrix Hydrogel Environments for Embryonic Stem Cell Growth"

Embryonic stem cells (ESCs) are pluripotent cells derived from the inner cell mass of the blastocyst that can give rise to cells of the ectoderm, endoderm and mesoderm lineages. Once isolated from the blastocyst, ESCs can be cultured indefinitely in vitro in an undifferentiated state or can be induced to differentiate. In the case of mouse ESCs (mESCs), the cytokine leukemia inhibitory factor (LIF) is added to culture media to maintain pluripotency and is removed to induce differentiation. Although it is known that extracellular matrix (ECM) components influence stem cell maintenance, proliferation and differentiation, the precise effects of ECM environments on embryonic stem cell behavior have not been systematically studied. The main purpose of this thesis project was to investigate the behavior of undifferentiated mESCs cultured in different 3-D hydrogel matrices and to determine whether viscoelastic and biochemical variations in the matrices differentially affect the ability of stem cells to self-renew; that is, retain their pluripotency or undifferentiated phenotype. Their behavior in 3D environments was compared to mESC behavior in traditional 2D culture. In addition, a new method of casting hydrogels in polydimethylsiloxane (PDMS) molds was developed in order to efficiently cast multiple hydrogels of varying sizes and shapes.    Results showed that mESC constructs cast in PDMS molds possessed similar characteristics to those cultured in control disposable vinyl molds throughout the duration of the studies, suggesting that the casting method developed had no adverse affect on 3-D mESC culture. Viability analysis indicated that both cells in 2D and cells in the 3D hydrogel environments remained viable throughout the duration of the experiment. The viscoelastic moduli of the collagen and fibrin hydrogels increased with increasing protein concentration. In both 2D and 3D environments mESCs grew as distinct independent clusters or colonies; however cell colonies in 2D were less uniform in size and shape and morphology. No significant differences in the nuclear density of the cell colonies were observed in the different matrices. In contrast, mESC colonies grown on 2D substrates were more spread and hence, the nuclear density was less. Colonies grown in 2D were also less homogeneous than those grown in 3D hydrogels, as indicated by broader distributions of shapes and sizes. The mESCs retained their undifferentiated morphology in 3D comparable to those in 2D and expressed the pluripotent marker Oct-4 in both fibrin and collagen hydrogels.    This thesis project suggests that 3-D hydrogel microenvironments are able to support the growth of mESCs in vitro and may serve as the basis of hydrogel carriers for stem cell transplantation therapies in vivo.

TITLE: "Biomaterials for Tissue Engineering for Rheumatoid Arthritis (controlling of Dendritic Cell Phenotype)"

The host response toward biomaterial component of tissue-engineered devices has been extensively investigated by exploring the potential inflammatory response of the host upon contact with biomaterials. Due to a biomaterial adjuvant effect, questions are raised to understand a role of biomaterials, which has been shown to modulate the host response. Specifically, it has been shown that the adjuvanticity of biomaterials affects the maturation of dendritic cells (DCs), professional antigen presenting cells (APCs) central to controlling immune response. Dendritic cells have been also recognized as the key regulator of the balance between tolerance and immunity determinant for pathogenesis of the autoimmune disease such as rheumatoid arthritis (RA).   The objective of this research is to understand the response of DCs to different biomaterials upon contact. We hypothesize that DCs respond with differential levels of maturation upon contact with different biomaterials on which iDCs were cultured, and the adjuvant effects of biomaterials is elucidated through implantation into induced RA models of mice and the integration of biomaterials is also elucidated in the RA joint of rabbits.Four natural biomaterials (alginate, hyaluronic acid, chitosan, and agarose) and one synthetic biomaterials (Poly(lactic-co-glycolic acid) (PLGA, 75:25)) are used as model biomaterials to observe the effects of differently inherent features of biomaterials on human DC maturation in vitro.   Further, to assess the biomaterial effects in a RA induction, two different biomaterials, which show different results in supporting maturation of DCs from in vitro study, will be implanted into mice to test the adjuvant effect of biomaterials and into rabbits to test the integration of biomaterials in the knee joints with the induced RA models. In the tissue engineering for RA treatment, the biomaterial component needs not only to provide a chondrogenically favorable environment for successful cartilage regeneration, but also to be tolerated by the host immune system associated with cells such as macrophages, dendritic cells, osteoclast, synoviocytes, and lymphocytes.   The goal of this study is to elucidate the host response to different biomaterials and to ultimately understand the adjuvant effects to suggest new selection and design criteria for biomaterials to be used in tissue engineering including treatments for RA patients.

TITLE: "Bayesian Based Risk Stratification of Atrial Fibrillation in Coronary Artery Bypass Graft Patients"

Roughly thirty percent of coronary artery bypass graft (CABG) patients develop atrial fibrillation (AF) in the five days following surgery, increasing the risk of stroke, prolonging hospital stay three to four days, and increasing the overall cost of the procedure. Current pharmacologic and nonpharmacologic means of AF prevention are suboptimal, and their side effects, expense, and inconvenience limit their widespread application. An accurate method for identifying patients at high risk for postoperative AF would allow these methods to be focused on the patients on which its utility would be highest. The main objective of this research was to develop a Bayesian network (BN) which could model/predict/assign risk of the occurrence of atrial fibrillation in CABG patients using retrospective data. A secondary objective was to develop an integrated framework for more advanced methods of feature selection and fusion for medical classification/prediction. We determined that the naïve Bayesian network classifier used with features selected by a genetic algorithm is a better classifier to use, given our cohort. The naïve BN allows for reasonable prediction despite being presented with patients with missing data points as might occur in the hospital. This classifier achieves a sensitivity of 0.63 and a specificity of 0.73 with an AUC of 0.74. Furthermore, this system is based on probabilities that are well understood and easily incorporated into a clinical environment. These probabilities can be altered based on the cardiologists' prior knowledge through Bayesian statistics, allowing for online sensitivity analysis by doctors, to perceive the best treatment options.

TITLE: "A Mathematical Model of Sleep-Wake Cycle"

The daily sleep-wake cycle usually consists of three distinct states: wakefulness, non-rapid-eye-movement (NREM) and rapid-eye-movement (REM). The process of switching between different states is complex, but a common assumption is that it is regulated primarily by two processes (circadian process and homeostatic process) via the reciprocal interactions of a bunch of downstream neuron groups, not only resulting in often rapid transitions from one state to another, but also allowing for a certain degree of bi-stability that locks the organism for some while in a given state before it switches back. In order to better understand how the behavioral states are regulated by different neuron groups, by using the method of S-system, we come up with a mathematical model consisting of two phases. The first phase is about the switch between wakefulness and sleep, which is controlled by the interactions between wake- and sleep-promoting neurons; the second phase is about the generation of NREM-REM alternation, which is believed to be regulated by REM-OFF and REM-ON neurons. We interpret the circadian rhythm as external input and homeostatic regulation as a "feedback controller". Both open-loop and closed-loop forms of the two-phase model are investigated and implemented. Discharging activities of the corresponding neuron groups and the switches of behavioral states are shown in the simulation results, from which we can easily identify the basic roles of wake- and sleep-promoting neurons, REM-OFF and REM-ON neurons. The special regulating function of neuropeptide-orexin is also tested by simulation.

TITLE: "Prosthetic Vein Valve: Delivery and In Vitro Evaluation"

Chronic Venous Insufficiency (CVI) is a painful and debilitating disease that affects the superficial and deep vein valves of the legs. These incompetent valves allow reflux and subsequent pooling of blood. Prosthetic venous valves were constructed from a novel hydrogel biomaterial patented at Georgia Tech. The valves had flexible cusps similar to normal, anatomic venous valves. The purpose of this work was to evaluate the thrombotic potential of the GT venous valve in an in vitro study and to design a percutaneous delivery system. The flow system was modified from a one-pass, flow-through thrombosis assay using whole blood. Whole blood was perfused through the valves to mimic the pulsatile physiologic conditions. Cessation of flow indicated thrombotic obstruction. A group of valves were lined with Dacron to serve as a positive control. Histological analysis was performed using H&E staining and Carstair's stain (specific for platelets).   Whole blood perfused through GT prosthetic valves exhibited no thrombosis or platelet adherence. All GT valves were patent and competent after blood perfusion. In contrast, all the valves lined with Dacron occluded. H&E staining revealed no thrombus deposition on the GT vein valves. However, the Dacron valves were occluded by thrombus connecting the polymer fibers with adherent platelets, identified by Carstair's staining.  A percutaneous delivery system was designed after evaluating the GT valves for their compressibility and plastic deformation over time. Appropriate stents, catheters and sheaths were then selected.   The novel vein valve demonstrates excellent patency, low thrombogenicity, and long-term competency with pre-clinical bench testing.

TITLE: "Molecular Dynamics of RNA hairpins motif"

The prediction and design of three-dimensional structures of large RNAs are best approached using small structural motif . RNA hairpin loops are very common and important secondary structures. Tetraloops, one type of the hairpin loops, are the simplest, smallest and most frequent RNA motif. This project helps to address the folding of the standard tetraloop. The occurrence of the d2-tetraloop, a specific tetraloop with one residue less in the loop area, will be analyzed. Thermal unfolding will be used to help constitute the folding pathway of the tetraloop by reversing the unfolding trajectory in hopes of finding the d2-tetraloop like structure to prove that the d2-tetraloop is an intermediate during the standard tetraloop folding. Other members of hairpin loops other than tetraloops are known to be more flexible than tetraloops so that they are involved with certain tertiary interactions critical to biofunctions such as riboswitch. Conformational dynamics of these hairpin loops, such as pentaloops or heptaloops, will be studied for the purpose of finding the polymorphism of the RNA hairpin loop structure where some of the metastable states can easily be involved in tertiary interactions.

TITLE: "Bioengineered Scaffolds for Peripheral Nerve Regeneration"

Nerve autografts are widely used clinically to repair nerve grafts. However, nerve grafts have many limitations, such as, availability of donor nerve grafts, and loss of function at donor site. To overcome these problems, we have used tissue engineering approach to design three-dimensional (3D) scaffolds containing laminin-1 (LN-1) and nerve growth factor (NGF). After nerve injury and during embryonic development, several extra-cellular matrix molecules and neurotrophic factors contribute to nerve regeneration and repair. Our 3D agarose hydrogel scaffolds with gradients of LN-1 and NGF are designed to mimic these in vivo conditions to promote nerve regeneration in rats. To determine the effect of LN-1 gradients on neurite extension in vitro , dorsal root ganglia (DRG) from chick embryo were cultured in 3D hydrogels. A gradient of LN-1 molecules in agarose gels was made by diffusion. LN-1 was then immobilized to the agarose hydrogels using a photo-crosslinker, Sulfo-SANPAH. Anisotropic scaffolds with three different slopes of LN-1 gradients were used. Isotropic scaffolds with uniform concentrations of LN-1, at various levels, were used as positive control. DRG cultured in anisotropic scaffolds with optimal slope of LN-1 gradient extended neurites twice as faster as DRG in optimal concentration in isotropic scaffold. Also, in the anisotropic scaffolds the faster growing neurites were aligned along the direction of LN-1 gradient. To promote nerve regeneration in vivo , tubular polysulfone guidance channels containing agarose hydrogels with gradients of LN-1 and NGF (anisotropic scaffolds) were used to bridge 20-mm nerve gaps in rats. Nerve autografts were used as positive control and isotropic scaffolds, with uniform concentration of LN-1 and NGF, were used as negative control. After 4-months, the rats were sacrificed and nerve histology was done to test for nerve regeneration. Only anisotropic scaffolds and nerve autografts showed axonal regeneration. Both groups had similar number of myelinated axons and similar axonal-diameter distribution. However, nerve graft group performed better in functional outcome as measured by relative gastrocnemius muscle weight (RGMW) and electrophysiology. Optimization of performance of anisotropic scaffolds by varying the LN-1 and NGF concentration gradients might lead to development of scaffolds that can perform as well as nerve auotgrafts for nerve regeneration over long nerve gaps.

TITLE: "Investigation of Physical Processes in Digital X-Ray Tomosynthesis Imaging of the Breast"

Early detection is one of the most important factors in the survival of patients diagnosed with breast cancer. For this reason the development of improved screening mammography methods is one of primary importance. One problem that is present in standard planar mammography, which is not solved with the introduction of digital mammography, is the possible masking of lesions by normal breast tissue because of the inherent collapse of three-dimensional anatomy into a two-dimensional image. Digital tomosynthesis imaging has the potential to avoid this effect by incorporating into the acquired image information on the vertical position of the features present in the breast. Previous studies have shown that at an approximately equivalent dose, the contrast-detail trends of several tomosynthesis methods are better than those of planar mammography. By optimizing the image acquisition parameters and the tomosynthesis reconstruction algorithm, it is believed that a tomosynthesis imaging system can be developed that provides more information on the presence of lesions while maintaining or reducing the dose to the patient. Before this imaging methodology can be translated to routine clinical use, a series of issues and concerns related to tomosynthesis imaging must be addressed. This work investigates the relevant physical processes to improve our understanding and enable the introduction of this tomographic imaging method to the realm of clinical breast imaging. The processes investigated in this work included the dosimetry involved in tomosynthesis imaging, x-ray scatter in the projection images, imaging system performance, and acquisition geometry. A comprehensive understanding of the glandular dose to the breast during tomosynthesis imaging, as well as the dose distribution to most of the radiosensitive tissues in the body from planar mammography, tomosynthesis and dedicated breast computed tomography was gained. The analysis of the behavior of x-ray scatter in tomosynthesis yielded an in-depth characterization of the variation of this effect in the projection images. Finally, the theoretical modeling of a tomosynthesis imaging system, combined with the other results of this work was used to find the geometrical parameters that maximize the quality of the tomosynthesis reconstruction.

TITLE: " Intracellular Delivery of Proteins and Nucleic Acids for Vaccines"

Viral infections are a major cause of death and disability worldwide, and new vaccine strategies are needed to combat viruses such as HIV and influenza. An alternative to traditional vaccines, which are based on attenuated or inactivated viruses, is the subunit vaccine, comprising isolated viral protein and adjuvants. One type of adjuvant, or immunostimulatory molecule, is the double-stranded RNA analog poly(I:C), which induces an antiviral immune response through stimulation of Toll-like receptor 3 (TLR3). In the proposed work, two new vaccine delivery vehicles will be developed to simultaneously deliver the model antigen ovalbumin and poly(I:C) to phagocytic immune cells. The first is a self-assembling block copolymer micelle that efficiently encapsulates proteins and nucleic acids through electrostatic interactions and disulfide cross-linking. The second is a solid nanoparticle composed of a pH- sensitive polyketal that degrades into biocompatible, non-acidic molecules. Loading efficiencies and release kinetics of ovalbumin and poly(I:C) will be measured for the vaccine delivery vehicles. In vitro experiments will be conducted to measure activation of antigen-presenting cells and cross-priming of cytotoxic T lymphocytes, followed by testing of the vaccine's ability to generate an adaptive immune response in mice.

TITLE: "Image Segmentation and Shape Analysis of Blood Vessels with Applications to Coronary Atherosclerosis"

Atherosclerosis is a systemic disease of the vessel wall that occurs in the aorta, carotid, coronary and peripheral arteries. Atherosclerotic plaques in coronary arteries may cause the narrowing (stenosis) or complete occlusion of the arteries and lead to serious results such as heart attacks and strokes. Medical imaging techniques such as X-ray angiography and computed tomography angiography (CTA) have greatly assisted the diagnosis of atherosclerosis in living patients. Analyzing and quantifying vessels in these images, however, is an extremely laborious and time consuming task if done manually. A novel image segmentation approach and a quantitative shape analysis approach are proposed to automatically isolate the coronary arteries and measure important parameters along the vessels. The segmentation method is based on the active contour model using the level set formulation. Regional statistical information is incorporated in the framework through Bayesian pixel classification. A new conformal factor and an adaptive speed term are proposed to counter the problems of contour leakage and narrowed vessels resulting from the conventional geometric active contours. The proposed segmentation framework is tested and evaluated on a large amount of 2D and 3D, including synthetic and real 2D vessels, 2D non-vessel objects, and eighteen 3D clinical CTA datasets of coronary arteries. The centerlines of the vessels are proposed to be extracted using harmonic skeletonization technique based on the level contour sets of the harmonic function, which is the solution of the Laplace equation on the triangulated surface of the segmented vessels. The cross-sectional areas along the vessels can be measured while the centerline is being extracted. Local cross-sectional areas can be used as a direct indicator of stenosis for diagnosis. A comprehensive validation is performed by using digital phantoms and real CTA datasets. This study provides the possibility of fully automatic analysis of coronary atherosclerosis from CTA images, and has the potential to be used in a real clinical setting along with a friendly user interface. Comparing to the manual segmentation which takes approximately an hour for a single dataset, the automatic approach on average takes less than five minutes to complete, and gives more consistent results across datasets.

TITLE: "Design and Implementation of an Acoustic Resonator Based Biosensor System for Early Detection of Prostate Cancer"

The greatest determinant of cancer survival is early detection. Current genomic, proteomic and metabolomic research has demonstrated great strides in developing an understanding of various early-stage cancer mechanisms and subsequent proliferation. To take advantage of these advances and significantly reduce cancer mortality an appropriate sensing system must be developed for detecting the earliest signs of oncogenesis. Such a system will include a design that accommodates mass production at a reasonable cost and require minimal technical expertise to operate. To this end, I present a simple acoustic resonator based biosensor array design, fabricated using common mass production techniques from the microelectronics industry. The proposed sensor design involves the fabrication of an array of zinc oxide thin-film bulk acoustic resonators, deposited onto a tungsten/quartz acoustic mirror. Electrodes are fabricated to launch and support a thickness shear mode acoustic wave in the film by lateral field excitation. Devices of the array are then prepared for chemical-specificity by means of silanization of the surface and subsequent covalent binding of an antibody species to the surface. Incorporation of the sensor array into an appropriate flow cell completes the sensor system package design. To test the viability of the design and implementation, five markers, proven to be indicators of prostate cancer proliferation, will be screened in clinical samples provided by the Winship Cancer Institute.

TITLE: "Nonlinear Dynamics of the Autogenic Stretch Reflex during Postural Control and Rhythmic Movement"

The animal neuromuscular system has the remarkable ability to accomplish a variety of motor tasks in highly variable environmental conditions. Investigating the organizational principles of the neuromuscular system may allow us to design more robust and efficient controllers for robotic and prosthetic systems. The neuromuscular system can be organized in a hierarchy consisting, from top to bottom, of higher-brain centers, spinal reflexes, and the musculoskeletal plant. Feedback and feedforward pathways to the higher-brain centers are subject to long and potentially destabilizing time delays. Therefore, the peripheral layers of the neuromuscular system may play an important role in the maintenance of joint stability. Our overall hypothesis is that the autogenic stretch reflex modulates the mechanical properties of muscle in a nonlinear manner to minimize the effects of mechanical perturbations. To test our hypothesis, we will investigate the interactions between the frog (Rana pipiens) autogenic stretch reflex and the mechanical properties of muscle (plantaris longus) during physiologically relevant tasks. We will first develop a closed-loop hybrid biomechanical system that allows an in vitro frog muscle, still innervated by its intact spinal cord, to actuate a mechanical joint. We will then q uantify how the autogenic stretch reflex changes the stiffness and damping of muscle to stabilize joint perturbations during postural control. We hypothesize that the autogenic stretch reflex will reduce joint settling time and overshoot by varying muscle stiffness and damping as a nonlinear function of prior movement and muscle state. Finally, we will determine the perturbation rejection capabilities of the autogenic stretch reflex during rhythmic movement. We hypothesize that the autogenic stretch reflex will stabilize perturbations during rhythmic movement in a phase-dependent manner. This work will help quantify some of the nonlinear properties of the autogenic stretch reflex. Understanding the strategies used by spinal reflexes will be beneficial in the design of robust controllers for prosthetic and robotic devices

TITLE: "Quantum Dot-Fluorescent Protein FRET Probes for Monitoring Protease Activity"

Aberrant protease activity is implicated in a number of disease states and clinical conditions, including cancer, infectious disease, and cardiovascular and neurodegenerative disorders. While in vitro assays have been developed to measure the enzyme activity of many targets, few methodologies exist to assess relatively low enzyme levels in single live cells. Recent advances in nanobiotechnology provide the opportunity to overcome current limitations in order to facilitate the measurement of proteolytic activity in single live cells and in small volumes of bodily fluids, approaches that could both improve our understanding of various diseases through basic science studies and significantly advance clinical diagnostics.    The long-term goal of our research is to develop new tools that can be used to assess molecular interactions in live cells and bodily fluids for applications in basic biomedical research as well as potential uses in clinical diagnostics. The objective of this project is to develop a novel quantum dot-based probe that can be used to detect protease activity with very high sensitivity and specificity. We aim to achieve this through the development of a switchable QD probe that will enable the detection of protease activity with unprecedented sensitivity, allowing for the detection of low levels of enzymatic activity in single live cells via fluorescence microscopy and flow cytometry. The project entails developing fluorescence resonance energy transfer (FRET) probes consisting of QD donors and fluorescent protein (FP) acceptors that bind the QD surface via poly-histidine adhesion peptides. By introducing an enzyme cleavage site between the His-tag and the FP, a switchable QD probe will be created, whereby the QD will fluoresce with its characteristic brightness upon cleavage of the enzyme substrate and subsequent separation of the FP from the surface of the QD. The proposed research employs novel and more sensitive QD-based FRET pairs to vastly improve on existing enzyme activity assay technologies, providing a powerful tool for both in vitro and live, single cell measurements.     The goals of the project will be attained through work on three specific aims: (1) development of optimal QD/His-tag/FP combinations with the most efficient quenching of QD fluorescence; (2) quantitative evaluation of the sensitivity and dynamic range of the FRET probe in detecting caspase activity using control samples and cell lysates; (3) demonstration of the sensitive detection of cytosolic protease activity using both fluorescence microscopy and flow cytometry. In addition to immediate applicability in basic biomedical research, this probe has potential for clinical diagnostics using whole blood or other bodily fluids.