Standard surgical repair of intertrochanteric hip fractures requires accurate placement of a wire across the fracture using static fluoroscopic images. Few practice methods exist for perfecting this wire navigation skill outside the operating room. The objective of this research is to further understand skill development for orthopaedic drilling using a validated simulator, enabling more effective instruction and training. This includes the investigation of the relationship between practice and skill acquisition in conjunction with specific differences between experts and novices. This work details the creation and validation of an augmented reality wire navigation simulator for training orthopaedic drilling. This novel augmented reality simulator combines real-world objects, such as a surgical drill and synthetic bone, with virtually generated, radiation-free radiographic imaging. The central hypothesis is that an augmented reality wire navigation simulator will demonstrate construct validity and improve orthopaedic drilling skill through simulation training. This work identifies the differentiation of skill between experienced surgeons and novices completing the wire navigation task, demonstrating construct validity for the developed simulator. It also demonstrates that experienced surgeons are more accurate than novices in orthopaedic drilling (F(2, 39) = 3.721, p = 0.033). This provides evidence supporting the simulator's construct validity and value as a training and assessment tool in wire navigation of the proximal femur. Although the study was unsuccessful in providing sufficient evidence that training on the simulator directly transfers to more realistic drilling tasks, it revealed several discoveries about acquiring wire navigation skill. This work establishes a relationship between skill acquisition and practice for the wire navigation task. This learning curve shows that skill acquisition occurs much more slowly in wire navigation than previously assumed. The wire accuracy (i.e., tip-apex distance) is predicted to improve 0.1 to 0.3 millimeters with each successive practice repetition of the wire navigation task (95% CI, p < 0.001). In addition, the time to complete wire navigation improves between 0.4 and 2.0 seconds each subsequent practice trial (95% CI, p < 0.001). The developed simulator also identified several flaws in novice technique. First, novices do not account for the anteversion of the femoral neck indicating that the inclination angle is difficult for novices to understand and accurately drill from radiographic images. Another discovered flaw of novice orthopaedic residents is their lack of ability to accurately estimate distances in radiographic images. Novices were found to incorrectly estimate the wire accuracy by an average of 12.4 millimeters. Overall, this work establishes new findings, which can be used for future simulation development and coaching, enabling safer, more effective training methods for surgical residents.

With the advancement of technology, medical imaging has become a fast growing area of research. Some imaging questions require little physician analysis, such as diagnosing a broken bone, using a 2-D X-ray image. More complicated questions, using 3-D scans, such as computerized tomography (CT), can be much more difficult to answer. For example, estimating tumor growth to evaluate malignancy; which informs whether intervention is necessary. This requires careful delineation of different structures in the image. For example, what is the tumor versus what is normal tissue; this is referred to as segmentation. Currently, the gold standard of segmentation is for a radiologist to manually trace structure edges in the 3-D image, however, this can be extremely time consuming. Additionally, manual segmentation results can differ drastically between and even within radiologists. A more reproducible, less variable, and more time efficient segmentation approach would drastically improve medical treatment. This potential, as well as the continued increase in computing power, has led to computationally intensive semiautomated segmentation algorithms. Segmentation algorithms' widespread use is limited due to difficulty in validating their performance. Fusion models, such as STAPLE, have been proposed as a way to combine multiple segmentations into a consensus ground truth; this allows for evaluation of both manual and semiautomated segmentation in relation to the consensus ground truth. Once a consensus ground truth is obtained, a multitude of approaches have been proposed for evaluating different aspects of segmentation performance; segmentation accuracy, between- and within -reader variability. The focus of this dissertation is threefold. First, a simulation based tool is introduced to allow for the validation of fusion models. The simulation properties closely follow a real dataset, in order to ensure that they mimic reality. Second, a statistical hierarchical Bayesian fusion model is proposed, in order to estimate a consensus ground truth within a robust statistical framework. The model is validated using the simulation tool and compared to other fusion models, including STAPLE. Additionally, the model is applied to real datasets and the consensus ground truth estimates are compared across different fusion models. Third, a statistical hierarchical Bayesian performance model is proposed in order to estimate segmentation method specific accuracy, between- and within -reader variability. An extensive simulation study is performed to validate the model’s parameter estimation and coverage properties. Additionally, the model is fit to a real data source and performance estimates are summarized.

The aim of this study was to assess the validity of an objective method of speech intelligibility, the Rapid Speech Transmission Index (RASTI), in elementary school classrooms, The RASTI can be obtained more quickly than subjective measures of speech intelligibility and has been shown to be highly valid with adult listeners in auditoriums. In this study RASTI values were correlated with scores on a subjective test of speech intelligibility, the word intelligibility by Picture Identification (WIPI) test, for 45 students in two elementary classrooms (grades 1 through 3 and grade5). Results indicated that the RASTI value is a poor predictor of subjective speech intelligibility (WIPI) scores for these students. There was no significant difference between the correlations obtained in the two classrooms or between the correlations obtained with the average and largest of the three RASTI values and the WIPI scores. Further study needs to be done to determine the applicability of the RASTI to classroom environments.

The strand-Cartesian grid approach is a unique method of generating and computing fluid dynamic simulations. The strand-Cartesian approach provides highly desirable qual- ities of fully-automatic grid generation and high-order accuracy. This thesis focuses on the implementation of the Spalart-Allmaras turbulence model to the strand-Cartesian grid framework. Verification and validation is required to ensure correct implementation of the turbulence model.Mathematical code verification is used to ensure correct implementation of new algo- rithms within the code framework. The Spalart-Allmaras model is verified with the Method of Manufactured Solutions (MMS). MMS shows second-order convergence, which implies that the new algorithms are correctly implemented.Validation of the strand-Cartesian solver is completed by simulating certain cases for comparison against the results of two independent compressible codes; CFL3D and FUN3D. The NASA-Langley turbulence resource provided the inputs and conditions required to run the cases, as well as the case results for these two codes. The strand solver showed excellent agreement with both NASA resource codes for a zero-pressure gradient flat plate and bump- in-channel. The treatment of the sharp corner on a NACA 0012 airfoil is investigated, resulting in an optimal external sharp corner configuration of strand vector smoothing with a base Cartesian grid and telescoping Cartesian refinement around the trailing edge. Results from the case agree well with those from CFL3D and FUN3D. Additionally, a NACA 4412 airfoil case is examined, and shows good agreement with CFL3D and FUN3D, resulting in validation for this case.

Une assurance indicielle est proposée en réponse à l'augmentation des sécheresses impactant les prairies. Elle se base sur un indice de production fourragère (IPF) obtenu à partir d'images satellitaires de moyenne résolution spatiale pour estimer l'impact de l'aléa dans une zone géographique définie. Le principal enjeu lié à la mise en place d'une telle assurance réside dans la bonne estimation des pertes subies. Les travaux de thèse s’articulent autour de deux objectifs : la validation de l'IPF et la proposition d'amélioration de cet indice. Un protocole de validation est construit pour limiter les problèmes liés à l'utilisation de produit de moyenne résolution et au changement d’échelle. L'IPF, confronté à des données de référence de différentes natures, montre de bonnes performances : des mesures de production in situ (R² = 0,81; R² = 0,71), des images satellitaires haute résolution spatiale (R² = 0,78 - 0,84) et des données issues de modélisation (R² = 0,68). Les travaux permettent également d'identifier des pistes d'amélioration pour la chaîne de traitement de l'IPF. Un nouvel indice, basé sur une modélisation semiempirique combinant les données satellitaires avec des données exogènes relatives aux conditions climatiques et à la phénologie des prairies, permet d'améliorer la précision des estimations de production de 18,6 %. L’ensemble des résultats obtenus ouvrent de nombreuses perspectives de recherche sur le développement de l'IPF et ses potentiels d'application dans le domaine assurantiel.

As atividades a serem realizadas na área do empreendimento RMB (Reator Multipropósito Brasileiro), cujas instalações e infraestrutura agregadas serão localizadas no estado de São Paulo, no município de Iperó, podem envolver a liberação de efluentes para o meio ambiente, seja de forma controlada ou não, do mesmo modo que qualquer outra instalação. Os hidrocarbonetos policíclicos aromáticos (HPAs) estudados neste trabalho apresentam comportamento ubíquo, são considerados compostos químicos com alta toxicidade, persistência no meio ambiente, potencial bioacumulativo e forte tendência de se biomagnificar. O objetivo desse trabalho foi determinar 14 dos 16 HPAs majoritários, classificados pela USEPA, nas amostras de águas superficiais e subterrâneas da área de instalação do RMB e seu entorno. Foram avaliados quatro períodos distintos utilizando uma metodologia já desenvolvida e validada por Brito (2009). Foi verificado o cenário atual em relação às concentrações dos HPAs nas águas da região que será construído o empreendimento, RMB, fornecendo uma visão geral do grau de contaminação ou não dos corpos hídricos, o que contribuirá com um banco de dados para pesquisas futuras. Paralelamente, foi validado e aplicado um método analítico já existente para avaliação de HPAs nos sólidos suspensos da coluna d\'água das amostras coletadas. A avaliação preliminar da área estudada demonstrou que o local do projeto RMB é preservado, mas alguns corpos d\'água ao redor da área, como o rio Sorocaba, apresentam concentrações significativas de HPAs, provavelmente em razão das atividades antropogênicas no entorno, indicando que a área precisa de uma maior atenção. Utilizando a técnica analítica de HPLC com detecção UV/vis, foram quantificados compostos como Acenaftileno Fluoreno e Fenantreno acima do valor legislado nas amostras de água. Foram observadas baixas concentrações de praticamente todos os compostos estudados no material em suspensão e alguns compostos como Fluoreno, Acenafteno, Acenaftileno e Benzo[b]fluoranteno, em concentrações mais altas, é necessário um programa de monitoramento constante, para que a partir dos dados obtidos, possa-se tomar decisões concisas de controle. / The activities to be carried out in the area of the RMB (Brazilian Multipurpose Reactor) project, whose aggregate facilities and infrastructure will be located in the state of São Paulo, in the municipality of Iperó, may release effluents into the environment, either in a controlled or not in the same way as any other installation. The polycyclic aromatic hydrocarbons (PAHs) studied in this work are ubiquitous and considered chemical compounds with high toxicity, persistence in the environment, bio accumulative potential and strong tendency to bio magnify. The objective of this study was to determine 14 of the 16 major PAHs, classified by the USEPA, in the surface and groundwater samples of the RMB installation area and its surroundings. Four distinct periods were evaluated using a methodology already developed and validated. The current scenario was verified regarding the PAHs concentrations in the waters of the region that will be built the RMB, providing an overview of the degree of contamination of the water bodies, which will contribute to a database for future research. At the same time, an analytical method was validated and applied to evaluate PAHs in the suspended solids of the water column of the collected samples. The preliminary evaluation of the studied area demonstrated that the site of the RMB project is preserved, but some water bodies around the area, such as the Sorocaba River, present significant concentrations of PAHs, probably due to the anthropogenic activities in the surroundings, indicating that the area needs more attention. Using the analytical HPLC technique with UV/vis detection, compounds such as acenaphthylene, fluorene and phenanthrene were quantified above the legislated value in the water samples. Although low concentrations of practically all compounds studied in the suspended material have been observed and some compounds such as fluorene, acenaphthene, acenaphthylene, and benzo[b]fluoranthene at high concentrations, a constant monitoring program is required to that from the data obtained, one can make concise decisions of command and control.

Orientador: Guilherme de Paula Nogueira / Banca: Marcelo Vasconcelo Meireles / Banca: Daniel de Jesus Cardoso de Oliveira / Banca: Andréa Fontes Garcia / Resumo: Esse estudo teve como objetivo a padronização de um ensaio imunoenzimático (ELISA) para a determinação das concentrações plasmáticas de IGF-I total, utilizando o sistema de amplificação biotina-estreptavidina peroxidase em um ensaio competitivo. O IGF-I foi extraído da IGFBP, utilizando o tampão glicina acidificado seguido de neutralização do pH com hidróxido de sódio. As microplacas foram sensibilizadas com anti IgG de coelho, e as dosagens realizadas utilizando duas abordagens, um método sem competição (incubação prévia das amostras com o anticorpo anti-h-IGF-I) e outro com competição (adição simultânea de IGF-I biotilinado e amostra). Os melhores resultados foram obtidos utilizando o método competitivo, com a sensibilização da placa com 0,25 μg/poço de IgG anti-coelho, o anticorpo específico na diluição 1:250.000 e 0,06 ng/poço de IGF-I biotinilado. O ensaio in house apresentou, um limite inferior de detecção de 50 ng/mL, uma correlação de 0.945 entre doses quando comparado à uma metodologia comercial. Além disso, após 33 ensaios (1114 amostras) a metodologia apresentou uma boa precisão, com coeficientes de variação inter-ensaio de 12,94% (345,8 ng/mL) para os controles alto e 20,71% (131,6 ng/mL) para o baixo. Dessa forma, conclui-se que a metodologia imunoenzimática para quantificação de IGF-I total utilizando o sistema de amplificação biotina-estreptavidina peroxidase em um ensaio competitivo está estabelecida e apresenta-se como uma ferramenta útil para estudos que vis... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: This study aimed to standardize an enzyme-linked immunosorbent assay (ELISA) to determine plasma concentrations of total IGF-I using the amplification biotin-streptavidin peroxidase system in a competitive assay. The IGF-I was extracted from IGFBPs using the acidified glycine buffer followed by the pH neutralization with sodium hydroxide. The microplates were coated with anti-rabbit IgG, thereafter the measurements were carried out using two approaches, one without competition (prior incubation of samples with the anti-hIGF-I antibody) and another with competition (simultaneous addition of IGF-I and biotinylated sample). The best results were obtained using the competitive method, with the following combination of reagents: microplates were coated with 0.25 µg/well of anti-rabbit IgG, the specific antibody at a dilution of 1:250.000 and 0.06 ng/well of biotinylated IGF-I. The in house methodology showed sensitivity of detection limit of 50 ng/ml, a correlation between doses of 0.945 when compared to a commercial method. In addition, after 33 assays (quantification of 1114 samples) the proposed methodology presented a good precision, with interassay variation coefficients of 12.94% and 20.71% for the high and low controls, respectively. Finally, we concluded that ELISA method for the quantification of total IGF-I using the system biotin-streptavidin-peroxidase amplification in a competitive assay is established and is presented as a useful tool for studies aimed at monitoring ... (Complete abstract click electronic access below) / Doutor

Les maladies neuromusculaires sont un ensemble d'affections le plus souvent héréditaires d'une composante de l'unité motrice. L'expression principale de ces maladies est la perte de force musculaire responsable du handicap moteur.<br />Il existe un besoin de développement de moyens fiables et sensibles d'évaluation de la force musculaire appliqués aux maladies neuromusculaires.<br />Les mécanismes physiologiques de la production de force musculaire sont rappelés, les modalités de son évaluation et les applications aux maladies neuromusculaires sont décrites.<br />Les dynamomètres isocinétiques sont couramment utilisés comme outils d'évaluation de la force musculaire mais les protocoles proposés sont peu adaptés aux patients atteints de déficit d'origine neuromusculaire.<br />Nous avons développé un système d'exploitation du signal de mesure sur le dynamomètre cybex 6000® et un protocole de mesure adapté aux patients atteints de maladies neuromusculaires pour évaluer la force maximale volontaire isocinétique des fléchisseurs et extenseurs des coudes et des genoux. Une phase de développement électronique et informatique a été menée afin d'étudier les paramètres représentatifs de la force musculaire avec une meilleure précision que sur le logiciel commercialisé avec le dynamomètre. Un protocole d'évaluation a été appliqué, basé sur la comparaison de mesures effectuées en conditions identiques de mobilisation passive en situation de repos et situation d'effort maximal volontaire concentrique.<br />Une étude de faisabilité auprès de sujets atteints de polyradiculonevrite aigue a été effectuée. Une première étude de fiabilité a été menée auprès de 15 sujets atteints de maladies neuromusculaires au moyen d'un autre dynamomètre, le Biodex 3® selon un protocole équivalent. Cette étude a permis d'observer une bonne fiabilité (ICC<0.8) des mesures de moments maximal (en Nm) et de travail mécanique (en Joules) des fléchisseurs et extenseurs de genoux. La plus petite différence détectable statistiquement (SDD) était de 12Nm pour tous les sujets et de 4Nm pour les sujets ayant un testing manuel inférieur à 4. Une étude de validation contre critère a permis d'observer auprès de 38 sujets atteints de maladies neuromusculaires une bonne corrélation des paramètres isocinétiques avec les paramètres cliniques (Testing manuel standardisé et Mesure de fonction motrice (MFM)). Il existe un lien significatif entre la mesure isocinétique des extenseurs de genou et la dimension 1 (Station debout et transferts) de la MFM ainsi que le score global de la MFM (coefficient de corrélation de spearman >0,7). Une valeur de pic de couple inférieure à 40Nm pour les extenseurs de genou est prédictive d'un score fonctionnel de MFM inférieur à 75% pour la dimension D1 (station debout et transferts) et inférieur à 80% pour le score total MFM T.<br />Une étude de sensibilité au changement a été conduite auprès de 12 sujets atteints de maladies neuromusculaires lentement évolutive. Nous n'avons pas observé de changement clinique significatif au sein de cette population mais une analyse par sujet a montré une importante variabilité d'évolution et une bonne concordance d'évolution des scores fonctionnels (MFM) avec les mesures isocinétiques.

Dans le cadre de la quantification de la qualité d'un<br />modèle par rapport à une référence expérimentale, nous présentons<br />une nouvelle théorie basée sur le concept de méconnaissance qui<br />permet de manipuler un modèle qui est forcément une représentation<br />imparfaite de la réalité, et d'obtenir des résultats prédictifs<br />qui tiennent compte du manque de connaissance que l'on a vis-à-vis<br />de la structure réelle étudiée. Cette théorie, qui mêle théorie<br />des intervalles et théorie des probabilités, repose sur la<br />globalisation des diverses sources d'erreur à l'échelle des<br />sous-structures en utilisant une variable interne scalaire<br />contenue dans un intervalle dont les bornes suivent des lois de<br />probabilités.<br /><br />À partir de ces bornes, si l'on considère une quantité<br />d'intérêt définie sur l'ensemble de la structure (une fréquence<br />propre par exemple), il est possible à partir de la donnée du<br />modèle de méconnaissances de base de calculer un intervalle<br />d'appartenance de cette quantité d'intérêt, dont les bornes sont<br />probabilistes. On définit alors des intervalles de méconnaissances<br />dites effectives, qui ont une certaine probabilité de contenir la<br />quantité d'intérêt associée au modèle.<br /><br />La quantification de la qualité d'un modèle par rapport<br />à une référence expérimentale se traduit alors par la<br />détermination des méconnaissances de base qui sont les plus<br />représentatives des dispersions constatées. L'idée majeure de<br />cette détermination est de considérer que plus on a d'informations<br />expérimentales, plus on est susceptible de réduire les<br />méconnaissances de base. La démarche dite de réduction des<br />méconnaissances est appliquée avec succès sur des cas d'études<br />académiques ainsi que sur un exemple réel de structure<br />industrielle : le support de satellites SYLDA5 présent dans le<br />lanceur Ariane 5.

Le soudage par points est la principale technique d'assemblage des tôles de carrosserie automobile. Cependant, la variabilité de ce procédé, et en particulier de l'état des électrodes de soudage, pose deux problèmes non résolus à ce jour: le contrôle non destructif de la qualité des soudures et la commande des paramètres de soudage. Utilisé fréquemment dans le domaine des réseaux de neurones, et réputé pour donner de bons résultats avec peu d'exemples, le leave-one-out conduit en réalité souvent à la sélection de modèles surajustés. Or, par un développement de Taylor, nous pouvons calculer l'effet du retrait d'un exemple de la base d'apprentissage sur la sortie du modèle. Nous prouvons que l'estimation de l'erreur de généralisation obtenue à partir de ce calcul est plus fiable que celle obtenue classiquement par apprentissage. Par ailleurs, nous montrons le lien existant entre le leave-one-out calculé et les intervalles de confiance sur la sortie du modèle. Enfin, nous proposons une technique originale de sélection de modèles non linéaires qui évite le surajustement en limitant l'influence de chaque exemple sur l'estimation des paramètres du modèle. Nous présentons une méthode permettant d'obtenir un modèle de prévision du diamètre d'une soudure, valable dans des conditions préétablies. Plus particulièrement, nous montrons comment constituer une base de données initiale, sélectionner le modèle adéquat, puis compléter progressivement la base d'apprentissage. Appliquée à deux types de tôles, cette méthode permet d'atteindre une précision proche de l'erreur de mesure. En outre, utilisés au sein d'une loi de commande, ces modèles permettent d'optimiser l'usure des électrodes.