No description available.

No description available.

The overall objective of this research consists of two main parts: (1) provide evidence that autocrine production of IGF-I modulates tight junction permeability and (2) demonstrate the ability of IGFBPs to regulate IGF-I delivery across cell layers. To meet the first objective, parental and IGF-I secreting bovine mammary epithelial cells were tested for cell layer permeability, tight and adherens junction proteins, IGF-IR, and a downstream signaling components of IGF-IR. In comparison with parental cells, IGF-I secreting cells had high levels of IGF-IRs, but low levels of the junction components E-cadherin, b-catenin, and occludin. The differences in parental and IGF-I secreting cells was not due to extracellular stimuli since inclusion of IGF-I, IGFBP-3, or co-culture with SV40-IGF-I cells did not alter the barrier properties of parental cells, suggesting that intracrine signaling may alter cell connectivity. The second objective focused on exogenous rather than endogenous IGF-I and the role of IGFBPs and IGF-IRs in ligand transcytosis. Bovine aortic endothelial cells (BAECs) cultured on surfaces optimized to minimize paracellular transport were utilized to investigate the kinetics involved in the transport of insulin-like growth factor-I from the apical side of confluent monolayers to the basolateral side. Binding competitors were used to determine the role of the cell surface insulin-like growth factor-I receptor (IGF-IR) and cell surface insulin-like growth factor binding proteins (IGFBPs) in this transport process. Although IGFBPs initially retard delivery of IGF-I, using a computation model, this report shows that pulse durations of less than 6 hrs resulted in enhanced delivery of IGF-I in the presence of IGFBPs, above that for delivery in the absence of IGFBPs. In addition, the model was utilized to identify key parameters to target when developing engineered growth factors for the treatment of diseases. It is shown that the sorting factions and internalization rates are reasonable targets for the design of engineered growth factors. Since the sorting fractions are dictated by binding affinities in the acidic environment of the endosomes, it may be beneficial to design and analog of IGF-I that is more resistant to changes in pH, similar to those develop from epidermal growth factor. / Ph. D.

Understanding the complex growth process of dendritic arbors is essential for the medical field and disciplines like Biology and Neurosciences. The establishment of the dendritic patterns has received increasing attention from experimental researchers that seek to determine the cellular mechanisms that play a role in the growth of neural trees. Our goal in this thesis was to prove the recurrence formula for the probability distribution of all possible neural trees, as well as the formulas of the expected number of active branches and their variances. We also derived formulas for the spatial locations of the optimal targeting region for a tree with branching probability. These formulas were necessary for the simplified stochastic computational model that Osan et al have developed in order to examine how changes in branching probability influence the success of targeting neurons located at different distances away from a starting point.

This thesis aims to investigate how well associative learning can account for human sequence learning under incidental conditions. It seems that we can learn complex sequential information about events in our environment, for example language or music, incidentally, without being aware of it. Awareness is, however, a complex issue with arguments for (Dienes, 2012) and against (Shanks, 2005) the existence of implicit learning processes. A dual process account proposes that there exist two different learning systems, one based on conscious, controlled reasoning and rules, and the other based on automatic association formation, which can take place outside of awareness (McLaren, Green, & Mackintosh, 1994). This thesis attempts to use the predictions of an associative account in conjunction with a suitable method for investigating implicit learning: sequence learning (Destrebecqz & Cleeremans, 2003). The research involves a collection of serial reaction time (SRT) tasks whereby participants respond to on-screen stimuli that follow a sequence that they were (intentional learning) or were not (incidental learning) informed of. Following on from the experimental design of Jones and McLaren (2009) this thesis provides evidence that humans differ in their ability to learn different sequential contingencies. After training sequences of trials where the current trial location was twice as likely to be either: the same as (Same rule); or different to (Different rule) the location two trials before this, participants were far better at learning the latter rule. I found that this result was not adequately simulated by the benchmark associative model of sequence learning, the Augmented SRN (Cleeremans & McClelland, 1991), and present a revised model. This model, amongst other attributes, represents all the stimuli experienced by participants and can therefore learn stimulus-response contingencies. These seem to block learning (to some extent) about the Same rule thus providing an associative explanation of the advantage for acquisition of the Different rule. Further predictions regarding the role of additional stimuli alongside sequence learning were then derived from this associative account and tested on human participants. The first of these was that additional stimuli within the task will interact with sequence learning. I found that human participants show increased Same rule learning when additional, concurrently presented stimuli follow the previous element in the sequence. I demonstrate that when participants perform an SRT task where responses are predicted by the colour of a cue, they are able to learn about this relationship in the absence of awareness. Using this cue-response learning I further investigate cue-competition between sequences and colours under incidental conditions and find evidence that suggests between cue associations may alter the influence of cue competition. These results altogether suggest that stimuli – both simple and sequential – can be learned under incidental conditions. This thesis further proposes that learning about simple and more complex relationships between stimuli interacts according to the predictions of an associative account and provides evidence that contributes to a dual process understanding of human learning.

This PhD work constitutes a series of inter-disciplinary studies that use biologically plausible computational techniques and experiments with human subjects in analyzing facial expressions. The performance of the computational models and human subjects in terms of accuracy and response time are analyzed. The computational models process images in three stages. This includes: Preprocessing, dimensionality reduction and Classification. The pre-processing of face expression images includes feature extraction and dimensionality reduction. Gabor filters are used for feature extraction as they are closest biologically plausible computational method. Various dimensionality reduction methods: Principal Component Analysis (PCA), Curvilinear Component Analysis (CCA) and Fisher Linear Discriminant (FLD) are used followed by the classification by Support Vector Machines (SVM) and Linear Discriminant Analysis (LDA). Six basic prototypical facial expressions that are universally accepted are used for the analysis. They are: angry, happy, fear, sad, surprise and disgust. The performance of the computational models in classifying each expression category is compared with that of the human subjects. The Effect size and Encoding face enable the discrimination of the areas of the face specific for a particular expression. The Effect size in particular emphasizes the areas of the face that are involved during the production of an expression. This concept of using Effect size on faces has not been reported previously in the literature and has shown very interesting results. The detailed PCA analysis showed the significant PCA components specific for each of the six basic prototypical expressions. An important observation from this analysis was that with Gabor filtering followed by non linear CCA for dimensionality reduction, the dataset vector size may be reduced to a very small number, in most cases it was just 5 components. The hypothesis that the average response time (RT) for the human subjects in classifying the different expressions is analogous to the distance measure of the data points from the classification hyper-plane was verified. This means the harder a facial expression is to classify by human subjects, the closer to the classifying hyper-plane of the classifier it is. A bi-variate correlation analysis of the distance measure and the average RT suggested a significant anti-correlation. The signal detection theory (SDT) or the d-prime determined how well the model or the human subjects were in making the classification of an expressive face from a neutral one. On comparison, human subjects are better in classifying surprise, disgust, fear, and sad expressions. The RAW computational model is better able to distinguish angry and happy expressions. To summarize, there seems to some similarities between the computational models and human subjects in the classification process.

A computational model for mitochondrial function has been developed from oxygen concentration data measured in the Oroboros Oxygraph-2k and oxygen consumption rates measured in the Seahorse XF24 Analyzer. Measurements were acquired using embryonic-cultured cortical neurons and isolated mitochondria from CD1 mice. Based on the biological mechanism of mitochondrial activity, a computational model was developed using biochemical kinetic modelling. To modulate mitochondrial activity, dysfunctions were introduced by injecting the inhibiting reagents oligomycin, rotenone, and antimycin A, and the uncoupling reagent carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP) during measurements. To incorporate these changes, model equations were adapted and globally calibrated to experimental data using the genetic algorithm developed by Jason Fiege of the University of Manitoba by fitting oxygen concentration data. The model was coded in MATLAB R2014a along with the development of a graphical user interface for simulating mitochondrial bioenergetics in silico. / October 2016

Mechanical ventilation (MV) is a system that partially or fully assists patients whose respiratory system fails to achieve a gas exchange function. However, MV can cause a ventilator-associated lung injury (VALI) or even contribute to a multiple organ dysfunction syndrome (MODS) in acute respiratory distress syndrome (ARDS) patients. Despite advances in today technologies, mortality rates for ARDS patient are still high. A better understanding of the interactions between airflow from mechanical ventilator and the airway could provide useful information used to develop a better strategy to ventilate patients. The mechanisms, which mechanical ventilation induces airway inflammation, are complex processes and cover a wide range of spatial scales. The multiscale model of the airway have been developed combining the computational models at organ, tissue, and cellular levels. A model at the organ level was used to study behaviors of the airway during mechanical ventilation. Strain distributions in each layer of the airway were investigated using a model at the tissue level. The cellular inflammatory responses during mechanical ventilation were investigated through the cellular automata (CA) model incorporating all biophysical processes during inflammatory responses. The multiscale modeling framework started by obtaining airway displacements from the organ-level model. They were then transferred to the tissue-level model for determining the strain distributions in each airway layer. The strain levels in each layer were then transferred to the cellular-level model for inflammatory responses due to strain levels. The ratio of the number of damage cells to healthy cells was obtained through the cellular-level model. This ratio, in turn, modulated changes in the Young’s modulus of elasticity at the tissue and organ levels. The simulation results showed that high tidal volume (1400 cc) during mechanical ventilation can cause tissue injury due to high concentration of activated immune cells and low tidal volume during mechanical ventilation (700 cc) can prevent tissue injury during mechanical ventilation and can mitigate tissue injury from the high tidal volume ventilation. The multiscale model developed in this research could provide useful information about how mechanical ventilation contributes to airway inflammation so that a better strategy to ventilate patients can be developed.

Computational joint models provide insight into the biomechanical function of human joints. Through both deformable and rigid body modeling, the structure-function relationship governing joint behavior is better understood, and subsequently, knowledge regarding normal, diseased, and/or injured function is garnered. Given the utility of these computational models, it is imperative to supply them with appropriate inputs such that model function is representative of true joint function. In these models, Magnetic Resonance Imaging (MRI) or Computerized Tomography (CT) scans and literature inform the bony anatomy and mechanical properties of muscle and ligamentous tissues, respectively. In the case of the latter, literature reports a wide range of values or average values with large standard deviations due to the inability to measure the mechanical properties of soft tissues in vivo. This makes it difficult to determine which values within the published literature to assign to computational models, especially patient-specific models. Therefore, while the use of published literature serves as a reasonable first approach to set up a computational model, a means of improving the supplied input data was sought. This work details the application of artificial neural networks (ANNs), specifically feedforward and radial basis function networks, to the optimization of ligament stiffnesses for the improved performance of pre- and post-operative, patient-specific foot/ankle computational models. ANNs are mathematical models that utilize learning rules to determine relationships between known sets of inputs and outputs. Using knowledge gained from these training data, the ANN may then predict outputs for similar, never‑before-seen inputs. Here, an optimal network of each ANN type was found, per mean square error and correlation data, and then both networks were used to predict optimal ligament stiffnesses corresponding to a single patient’s radiographic measurements. Both sets of predictions were ultimately supplied to the patient-specific computational models, and the resulting kinematics illustrated an improvement over the existing models that utilized literature-assigned stiffnesses. This research demonstrated that neural networks are a viable means to hone in on ligament stiffnesses for the overall objective of improving the predictive ability of a patient-specific computational model.

Este trabalho está inserido em um projeto que procura estudar a viabilidade técnica da aplicação de um relativamente novo conceito de tratamento térmico, chamado de Têmpera e Partição (T&P), como alternativa para o processamento de ferros fundidos nodulares com alta resistência mecânica. O processo T&P tem por objetivo a obtenção de microestruturas multifásicas constituídas de martensita e austenita retida, estabilizada em carbono. A martensita confere elevada resistência mecânica, enquanto a austenita confere ductilidade. No processo T&P, após a austenitização total ou parcial da liga, o material é temperado até uma temperatura de têmpera TT entre as temperaturas Ms e Mf para produzir uma mistura controlada de martensita e austenita. Em seguida, na etapa de partição, o material é mantido isotermicamente em uma temperatura igual ou mais elevada (denominada temperatura de partição TP) para permitir a partição de carbono da martensita para a austenita. O carbono em solução sólida diminui a temperatura Ms da austenita, estabilizando-a à temperatura ambiente. O presente trabalho procurou estudar aspectos de transformações de fases -- com ênfase na evolução microestrutural e cinética das reações -- do tratamento térmico de Têmpera e Partição (T&P) aplicado a uma liga de ferro fundido nodular (Fe-3,47%C-2,47%Si-0,2%Mn). Tratamentos térmicos consistiram de austenitização a 880 oC por 30 min, seguido de têmpera a 140, 170 e 200 oC e partição a 300, 375 e 450 oC por até 2 h. A caracterização microestrutural foi feita por microscopia óptica (MO), eletrônica de varredura (MEV), difração de elétrons retroespalhados (EBSD) e análise de microssonda eletrônica (EPMA). A análise cinética foi feita por meio de ensaios de dilatometria de alta resolução e difração de raios X in situ usando radiação síncrotron. Resultados mostram que a ocorrência de reações competitivas -- reação bainítica e precipitação de carbonetos na martensita -- é inevitável durante a aplicação do tratamento T&P à presente liga de ferro fundido nodular. A cinética da reação bainítica é acelerada pela presença da martensita formada na etapa de têmpera. A reação bainítica acontece, a baixas temperaturas, desacompanhada da precipitação de carbonetos e contribui para o enriquecimento em carbono, e consequente estabilização, da austenita. Devido à precipitação de carbonetos na martensita, a formação de ferrita bainítica é o principal mecanismo de enriquecimento em carbono da austenita. A microssegregação proveniente da etapa de solidificação permanece no material tratado termicamente e afeta a distribuição da martensita formada na etapa de têmpera e a cinética da reação bainítica. Em regiões correspondentes a contornos de célula eutética são observadas menores quantidades de martensita e a reação bainítica é mais lenta. A microestrutura final produzida pelo tratamento T&P aplicado ao ferro fundido consiste de martensita revenida com carbonetos, ferrita banítica e austenita enriquecida estabilizada pelo carbono. Adicionalmente, foi desenvolvido um modelo computacional que calcula a redistribuição local de carbono durante a etapa de partição do tratamento T&P, assumindo os efeitos da precipitação de do crescimento de placas de ferrita bainítica a partir da austenita. O modelo mostrou que a cinética de partição de carbono da martensita para a austenita é mais lenta quando os carbonetos precipitados são mais estáveis e que, quando a energia livre dos carbonetos é suficientemente baixa, o fluxo de carbono acontece da austenita para a martensita. A aplicação do modelo não se limita às condições estudadas neste trabalho e pode ser aplicada para o planejamento de tratamentos T&P para aços. / The present work belongs to a bigger project whose main goal is to study the technical feasibility of the application of a relatively new heat treating concept, called Quenching and Partitioning (Q&P), as an alternative to the processing of high strength ductile cast irons. The aim of the Q&P process is to obtain multiphase microstructures consisting of martensite and carbon enriched retained austenite. Martensite confers high strength, whereas austenite confers ductility. In the Q&P process, after total or partial austenitization of the alloy, the material is quenched in a quenching temperature TQ between the Ms and Mf temperatures to produce a controlled mixture of martensite and austenite. Next, at the partitioning step, the material is isothermally held at a either equal or higher temperature (so called partitioning temperature TP) in order to promote the carbon diffusion (partitioning) from martensite to austenite. The present work focus on the study of phase transformations aspects -- with emphasis on the microstructural evolution and kinetics of the reactions -- of the Q&P process applied to a ductile cast iron alloy (Fe-3,47%C-2,47%Si-0,2%Mn). Heat treatments consisted of austenitization at 880 oC for 30 min, followed by quenching at 140, 170, and 200 oC and partitioning at 300, 375 e 450 oC up to 2 h. The microstructural characterization was carried out by optical microscopy (OM), scanning electron microscopy (SEM), backscattered diffraction (EBSD), and electron probe microanalysis (EPMA). The kinetic analysis was studied by high resolution dilatometry tests and in situ X-ray diffraction using a synchrotron light source. Results showed that competitive reactions -- bainite reaction and carbides precipitation in martensite -- is unavoidable during the Q&P process. The bainite reaction kinetics is accelerated by the presence of martensite formed in the quenching step. The bainite reaction occurs at low temperatures without carbides precipitation and contributes to the carbon enrichment of austenite and its stabilization. Due to carbides precipitation in martensite, growth of bainitic ferrite is the main mechanism of carbon enrichment of austenite. Microsegregation inherited from the casting process is present in the heat treated material and affects the martensite distribution and the kinetics of the bainite reaction. In regions corresponding to eutectic cell boundaries less martensite is observed and the kinetics of bainite reaction is slower. The final microestructure produced by the Q&P process applied to the ductile cast iron consists of tempered martensite with carbides, bainitic ferrite, and carbon enriched austenite. Additionally, a computational model was developed to calculate the local kinetics of carbon redistribution during the partitioning step, considering the effects of carbides precipitation and bainite reaction. The model showed that the kinetics of carbon partitioning from martensite to austenite is slower when the tempering carbides are more stable and that, when the carbides free energy is sufficiently low, the carbon diffuses from austenite to martensite. The model is not limited to the studied conditions and can be applied to the development of Q&P heat treatments to steels.