Much health related research depends heavily on the analysis of a rapidly expanding universe of observational data. A challenge in analysis of such data is the lack of sound statistical methods and tools that can address multiple facets of estimating treatment or exposure effects in observational studies with a large number of covariates. We sought to advance methods to improve analysis of large observational datasets with an end goal of understanding the effect of treatments or exposures on health. First we compared existing methods for propensity score (PS) adjustment, specifically Bayesian propensity scores. This concept had previously been introduced (McCandless et al., 2009) but no rigorous evaluation had been done to evaluate the impact of feedback when fitting the joint likelihood for both the PS and outcome models. We determined that unless specific steps were taken to mitigate the impact of feedback, it has the potential to distort estimates of the treatment effect. Next, we developed a method for accounting for uncertainty in confounding adjustment in the context of multiple exposures. Our method allows us to select confounders based on their association with the joint exposure and the outcome while also accounting for the uncertainty in the confounding adjustment. Finally, we developed two methods to combine het- erogenous sources of data for effect estimation, specifically information coming from a primary data source that provides information for treatments, outcomes, and a limited set of measured confounders on a large number of people and smaller supplementary data sources containing a much richer set of covariates. Our methods avoid the need to specify the full joint distribution of all covariates.

Background and purpose: Historically HDR brachytherapy treatment planning systems ignore the transit dose in the computation of patient dose. However, the total radiation dose delivered during each treatment cycle is equal to the sum of the static dose and the transit dose and every HDR application therefore results in two radiation doses. Consequently, the absorbed dose to the target volume is more than the prescribed dose as computed during treatment planning. The aim of this study was to determine the magnitude of the transit dose component of two 192Ir HDR brachytherapy units and assess its dosimetric significance. Materials and Methods: Ionization chamber dosimetry systems (well-type and Farmertype ionization chambers) were used to measure the charge generated during the transit of the 192Ir source from a GammaMed and a Nucletron MicroSelectron HDR afterloader using single catheters of lengths 120 cm. Different source configurations were used for the measurements of integrated charge. Two analysis techniques were used for transit time determination: the multiple exposure technique and the graphical solution of zero exposure. The transit time was measured for the total transit of the radioactive source into (entry) and out of (exit) the catheters. Results: A maximum source transit time of 1.7 s was measured. The transit dose depends on the source activity, source configuration, number of treatment fractions, prescription dose and the type of remote afterloader used. It does not depend on the measurement technique, measurement distance or the analysis technique used for transit time determination. Conclusion: A finite transit time increases the radiation dose beyond that due to the programmed source dwell time alone. The significance of the transit dose would increase with a decrease in source dwell time or a higher activity source.

Objetivo Avaliar o perfil da exposição ocupacional a agentes químicos, em laboratórios, e verificar os impactos na avaliação de risco decorrentes das mudanças nos limites de exposição ocupacional (LEO) e na legislação federal brasileira. Metodologia Revisão de literatura relativa à exposição ocupacional a substâncias químicas, referentes a: conceituação de risco; desenho das curvas dose-resposta; identificação dos fatores que interferem na relação dose-efeito; conceitos de exposição por longo período, da exposição múltipla e em baixas concentrações; e fatores de confusão na monitoração ocupacional. Realização de estudo transversal no Centro de Pesquisas da Petrobras, para avaliar o perfil de exposição a agentes químicos de 3.000 trabalhadores, com atividades preponderantes em laboratórios, metade das quais com exposição em baixas concentrações e por longo período, tendo por base as avaliações ambientais do ano de 2004. Análise temporal dos valores de LEO da American Conferrence of Governmental Industrial Hygienists (ACGIH) e levantamento das mudanças nas legislações federal previdenciária, trabalhista e de saúde nos aspectos relacionados à exposição a agentes químicos. Resultados Foram identificadas 484 substâncias químicas em 243 postos de trabalho, resultando em 2.550 situações de exposição, com média de 3,73 substâncias químicas por local. Os 1.563 trabalhadores com exposição formaram 168 Grupos Homogêneos de Exposição (GHE), cuja composição variou de 1 a 44 trabalhadores (média de 4,55, mediana de 3 e moda de 1). Em cada local de trabalho foram identificados, em média, 4,91 GHE. Foram medidas 977 amostras. Com relação ao GHE, foi notado que em 91,9% das avaliações ocorreram resultados abaixo do nível de ação, correspondendo a 92,5% dos empregados, configurando a exposição a baixas concentrações. Verificou-se ainda que 49,6% dos GHE (49,9% dos empregados) tinham concentrações inferiores ao limite de detecção das técnicas analíticas, enquanto que em 8,1% dos GHE e 7,5% dos empregados, as concentrações estavam em nível igual ou acima do nível de ação. Comprovou-se que nos últimos dez anos foram implantados ou revisados 135 (18,5%) LEO. O número de substâncias cuja redução do LEO é igual ou superior a 50% corresponde à quase totalidade das reduções e, em todos os períodos, as maiores são iguais ou superiores a 80% ultrapassando, portanto, o nível de ação. Ainda que não exista concordância nos valores dos LEO foi constatado que, entre as diferentes agências têm ocorrido freqüentes mudanças na legislação. Estes fatos têm obrigado os profissionais a incorporarem estes conhecimentos a sua prática de trabalho. Conclusões A exposição a substâncias químicas em laboratórios é predominantemente em baixa concentração, variada e múltiplas. Na proteção da saúde de trabalhadores, a aceitação do risco quando a concentração ambiental da substância química é abaixo do nível de ação, deve ser usado com parcimônia, pois os valores de LEO têm apresentado tendência de redução e exclui parcela significativa da população. Nesta se incluem os hipersensíveis, os expostos a substâncias com efeito estocástico, misturas com efeitos aditivos e com curva dose-resposta bifásica / Objective The aim of this study was to assess the profile of occupational exposure to chemicals in laboratories and check the impacts on risk assessment that derive from changes in occupational exposure limits (OEL) and in Brazilian federal laws applicable thereto. Methodology Review of publications on occupational exposure to chemicals concerning the following: risk concept, dose-response curve drawing, identifying factors that interfere with dose-effect relationship; concepts of lengthy exposure, multiple exposure, and low concentration exposure; confusing factors in occupational monitoring. A transversal study was carried out at the Petrobras Research Center in order to assess the profile of exposure to chemicals among 3,000 employees whose job was predominantly performed inside a laboratory, half of which were low concentration, lengthy exposures (the study was based on environmental assessments carried out in 2004). Time analysis of American Conference of Governmental Industrial Hygienists (ACGIH) OEL values. Survey on the changes made in security, labor, and health federal laws as relates to exposure to chemicals. Results 484 chemicals were identified in 243 workplaces. This resulted in 2,550 exposure situations with an average of 3.73 chemicals per location. The 1,563 workers under exposure were divided into 168 Homogeneous Exposure Groups (HEG) comprised by 1 to 44 subjects (mean = 4.55; median = 3; mode = 1). On average, 4.91 HEG were identified in each workplace. 977 samples were measured. Regarding the HEG, it was observed that 91.9% of the assessments showed results below action level. This is equivalent to 92.5% of the amount of workers and fits into the low concentration exposure category. It was also observed that 49.6% of HEG (i.e. 49.9% of workers) showed concentrations lower than detection limit in analytical techniques, whereas concentrations were equal to or greater than action level among 8.1% of HEG and 7.5% of workers. 135 (i.e. 18.5%) OEL were proven to have been implemented or revised. The amount of chemicals whose OEL decrease is equal to or greater than 50% is equivalent to nearly all decreases. Moreover, the highest decreases are equal to or greater than 80%, and therefore exceeded action level. Although there is no common agreement on OEL values, it was observed that applicable laws have been changed by several agencies on a regular basis. These facts have led professionals into considering such information in their work practice. Conclusions Exposure to chemicals in laboratories occurs basically under low, varied, multiple concentration. In the field of workers healthcare, one should be careful while considering a risk for environmental concentration of a chemical below action level. This is because OEL levels have shown to be prone to decrease and thus exclude a significant part of the population. Such part includes hypersensitive individuals, people exposed to chemicals with stochastic effect, mixtures with additive effects and biphasic dose-response curve

Objetivo Avaliar o perfil da exposição ocupacional a agentes químicos, em laboratórios, e verificar os impactos na avaliação de risco decorrentes das mudanças nos limites de exposição ocupacional (LEO) e na legislação federal brasileira. Metodologia Revisão de literatura relativa à exposição ocupacional a substâncias químicas, referentes a: conceituação de risco; desenho das curvas dose-resposta; identificação dos fatores que interferem na relação dose-efeito; conceitos de exposição por longo período, da exposição múltipla e em baixas concentrações; e fatores de confusão na monitoração ocupacional. Realização de estudo transversal no Centro de Pesquisas da Petrobras, para avaliar o perfil de exposição a agentes químicos de 3.000 trabalhadores, com atividades preponderantes em laboratórios, metade das quais com exposição em baixas concentrações e por longo período, tendo por base as avaliações ambientais do ano de 2004. Análise temporal dos valores de LEO da American Conferrence of Governmental Industrial Hygienists (ACGIH) e levantamento das mudanças nas legislações federal previdenciária, trabalhista e de saúde nos aspectos relacionados à exposição a agentes químicos. Resultados Foram identificadas 484 substâncias químicas em 243 postos de trabalho, resultando em 2.550 situações de exposição, com média de 3,73 substâncias químicas por local. Os 1.563 trabalhadores com exposição formaram 168 Grupos Homogêneos de Exposição (GHE), cuja composição variou de 1 a 44 trabalhadores (média de 4,55, mediana de 3 e moda de 1). Em cada local de trabalho foram identificados, em média, 4,91 GHE. Foram medidas 977 amostras. Com relação ao GHE, foi notado que em 91,9% das avaliações ocorreram resultados abaixo do nível de ação, correspondendo a 92,5% dos empregados, configurando a exposição a baixas concentrações. Verificou-se ainda que 49,6% dos GHE (49,9% dos empregados) tinham concentrações inferiores ao limite de detecção das técnicas analíticas, enquanto que em 8,1% dos GHE e 7,5% dos empregados, as concentrações estavam em nível igual ou acima do nível de ação. Comprovou-se que nos últimos dez anos foram implantados ou revisados 135 (18,5%) LEO. O número de substâncias cuja redução do LEO é igual ou superior a 50% corresponde à quase totalidade das reduções e, em todos os períodos, as maiores são iguais ou superiores a 80% ultrapassando, portanto, o nível de ação. Ainda que não exista concordância nos valores dos LEO foi constatado que, entre as diferentes agências têm ocorrido freqüentes mudanças na legislação. Estes fatos têm obrigado os profissionais a incorporarem estes conhecimentos a sua prática de trabalho. Conclusões A exposição a substâncias químicas em laboratórios é predominantemente em baixa concentração, variada e múltiplas. Na proteção da saúde de trabalhadores, a aceitação do risco quando a concentração ambiental da substância química é abaixo do nível de ação, deve ser usado com parcimônia, pois os valores de LEO têm apresentado tendência de redução e exclui parcela significativa da população. Nesta se incluem os hipersensíveis, os expostos a substâncias com efeito estocástico, misturas com efeitos aditivos e com curva dose-resposta bifásica / Objective The aim of this study was to assess the profile of occupational exposure to chemicals in laboratories and check the impacts on risk assessment that derive from changes in occupational exposure limits (OEL) and in Brazilian federal laws applicable thereto. Methodology Review of publications on occupational exposure to chemicals concerning the following: risk concept, dose-response curve drawing, identifying factors that interfere with dose-effect relationship; concepts of lengthy exposure, multiple exposure, and low concentration exposure; confusing factors in occupational monitoring. A transversal study was carried out at the Petrobras Research Center in order to assess the profile of exposure to chemicals among 3,000 employees whose job was predominantly performed inside a laboratory, half of which were low concentration, lengthy exposures (the study was based on environmental assessments carried out in 2004). Time analysis of American Conference of Governmental Industrial Hygienists (ACGIH) OEL values. Survey on the changes made in security, labor, and health federal laws as relates to exposure to chemicals. Results 484 chemicals were identified in 243 workplaces. This resulted in 2,550 exposure situations with an average of 3.73 chemicals per location. The 1,563 workers under exposure were divided into 168 Homogeneous Exposure Groups (HEG) comprised by 1 to 44 subjects (mean = 4.55; median = 3; mode = 1). On average, 4.91 HEG were identified in each workplace. 977 samples were measured. Regarding the HEG, it was observed that 91.9% of the assessments showed results below action level. This is equivalent to 92.5% of the amount of workers and fits into the low concentration exposure category. It was also observed that 49.6% of HEG (i.e. 49.9% of workers) showed concentrations lower than detection limit in analytical techniques, whereas concentrations were equal to or greater than action level among 8.1% of HEG and 7.5% of workers. 135 (i.e. 18.5%) OEL were proven to have been implemented or revised. The amount of chemicals whose OEL decrease is equal to or greater than 50% is equivalent to nearly all decreases. Moreover, the highest decreases are equal to or greater than 80%, and therefore exceeded action level. Although there is no common agreement on OEL values, it was observed that applicable laws have been changed by several agencies on a regular basis. These facts have led professionals into considering such information in their work practice. Conclusions Exposure to chemicals in laboratories occurs basically under low, varied, multiple concentration. In the field of workers healthcare, one should be careful while considering a risk for environmental concentration of a chemical below action level. This is because OEL levels have shown to be prone to decrease and thus exclude a significant part of the population. Such part includes hypersensitive individuals, people exposed to chemicals with stochastic effect, mixtures with additive effects and biphasic dose-response curve

This thesis deals with introduction to optical method called laser speckle contrast imaging (LSCI). The thesis contains a description of the basic theory that is related to the method, the LSCI own principle, the characteristics of this technique, its applications and also a description of various variants of LSCI. The second part of the thesis deals with the modification of acquisition software to be suitable for measurement of multi-exposure LSCI (MESI), verification of this method by experiment and conclusions of evaluated of data.

La biosurveillance humaine permet aujourd’hui d’évaluer notre exposition aux produits chimiques par la mesure soit des substances elles-mêmes, soit de leurs métabolites ou marqueurs d'effets sur la santé, à partir des fluides corporels ou des tissus. Les informations recueillies dans le cadre d’enquêtes épidémiologiques renseignent sur l'exposition humaine et constituent des bases précieuses dans la recherche des relations exposition-réponse chez les humains. Les travaux de cette thèse s’inscrivent dans le cadre de l’enquête transversale IMePoGe conduite entre 2008-2010 dans la région Nord–Pas-de-Calais (au nord de la France), incluant 2000 résidents adultes âgés de 20 à 59 ans, et visant à quantifier les niveaux d’imprégnation de la population à 14 métaux et métalloïdes (aluminium, antimoine, arsenic total, béryllium, cadmium, cobalt, chrome, mercure, manganèse, nickel, plomb, thallium, vanadium, zinc) choisis pour leurs effets toxiques et la fréquence de l’exposition professionnelle et environnementale. Les objectifs spécifiques de cette thèse étaient, tout en portant une attention particulière au plomb et au cadmium, deux métaux néphrotoxiques connus dans la littérature : i) d’établir la distribution de l’imprégnation aux métaux et métalloïdes dans la population du Nord de la France et de comparer le niveau régional d’imprégnation avec les données nationales et internationales ; ii) d’identifier les facteurs majeurs de variation de l’imprégnation et les sources d’exposition au plomb et au cadmium dans la population générale ; iii) d’étudier la relation entre la variation du débit de filtration glomérulaire et les niveaux d’imprégnation aux métaux. Globalement, les concentrations sanguines et urinaires de la plupart des métaux et métalloïdes étaient plus élevées que celles rapportées dans l’enquête nationale nutrition santé conduite sur la même période dans la population française, à l’exception du vanadium urinaire et du plomb sanguin. La plombémie moyenne régionale (moyenne géométrique) était de 18,8 μg/L. De nombreuses sources d’exposition au plomb existaient dans la population et étaient à la fois d’origine professionnelle, environnementale et alimentaire. Pour ce qui est du cadmium, le tabagisme se présentait comme la principale source d’exposition récente ou chronique au métal : la cadmiémie moyenne, reflet de l’exposition récente, était de 0,39 μg/L et passait de 0,26 μg/L chez les non-fumeurs à 0,84 μg/L chez les fumeurs ; la cadmiurie moyenne, reflet de l’exposition chronique, était de 0,37 μg/L (0,33 μg/g créatinine) et passait de 0,33 μg/L (0,29 μg/g créatinine) chez les non-fumeurs à 0,46 μg/L (0.37 μg/g créatinine) chez les fumeurs. Enfin, dans le cadre de l’étude de la relation entre l’imprégnation aux métaux et le débit de filtration glomérulaire, notre étude a montré que la prise en compte de la co-exposition à d’autres métaux et métalloïdes potentiellement néphrotoxiques bouleversait considérablement les associations antérieures jusqu’ici rapportées spécifiquement avec le plomb et le cadmium, dans le cadre de faibles niveaux d’imprégnation en population générale. / Human Biomonitoring allows us to evaluate our exposure to chemicals by measuring substances themselves or their metabolites or markers of health effects, from body fluids or tissues. The information collected through epidemiological surveys provide information on human exposure and are valuable databases in the research of exposure-response relationships in humans. This thesis is part of the cross-sectional IMePoGe survey conducted between 2008-2010 in the Nord–Pas-de-Calais region (in northern France), including 2,000 adult residents aged 20 to 59 years old, and aimed to quantify the impregnation levels of the population to 14 metals and metalloids (aluminum, antimony, total arsenic, beryllium, cadmium, cobalt, chromium, mercury, manganese, nickel, lead, thallium, vanadium, zinc) chosen for their toxic effects and the frequency of occupational and environmental exposure. The specific objectives of this thesis were, which a special interest for lead and cadmium, two nephrotoxic metals known in the literature: i) to establish the distribution of impregnation metals into the northern population of France and compare the exposure regional level to metals and metalloids with the national and international data; ii) to identify the major factors of variation of the impregnation and the sources of exposure to lead and cadmium in the general population; iii) to study the relationship between the change in glomerular filtration rate and the impregnation levels to metals. Overall, blood and urinary concentrations of most metals and metalloids were higher than those found in the national nutritional health survey conducted during the same period in the French population, with the exception of urinary vanadium and blood lead. The regional mean of blood lead level (geometric mean) was 18.8 μg/L. Several sources of lead exposure existed in the population and were link to the occupational, environmental and consumption parameters. Regarding cadmium, smoking was the main source of recent or chronic exposure to metal: the geometric mean of blood cadmium, reflecting a recent exposure, was 0.39 μg/L and increased from 0.26 μg/L in non-smokers to 0.84 μg/L in smokers; the geometric mean of urinary cadmium, reflecting the chronic exposure, was 0.37 μg/L (0.33 μg/g creatinine) and increased from 0.33 μg/L (0.29 μg/g creatinine) in non-smokers to 0.46 μg/L (0.37 μg/g creatinine) in smokers. Finally, as part of the study of the relationship between the metal levels and the glomerular filtration rate, our study showed that taking into account the multiple exposure to the other potentially nephrotoxic metals and metalloids upset considerably the previous associations specifically reported with lead and cadmium, in the context of low levels exposure in the general population.

Master's thesis is focused on capturing of low dynamic range images using common devices such as camera and its multiple exposure. The main part of thesis is dedicated to composing these images to HDR image, inclusive sequence of images of static scenes, but also dynamic ones. Next part describes tone mapping used for display HDR image on LDR monitors. Moreover, there is given design and implementation of application solving problems mentioned earlier. In the end, the implemented application is evaluated and the possible continuation of this work is stated.

Chaque jour, 15% des Européens sont exposés à des niveaux moyens de bruit ou de polluants atmosphériquesdépassant les valeurs seuils fixées par la législation européenne. L’intensité de la pollution de chaque polluant et sarépartition à l’échelle d’une agglomération dépendent des capacités d’émission des sources, des phénomènes dedispersion, et de l’environnement urbain. Il en résulte une variabilité spatiale du niveau de multi-exposition qui resteencore largement méconnue. L’objectif de ce travail de thèse est d’identifier et de caractériser les situations de multiexpositionsur la ville de Besançon. Les niveaux sonores et les concentrations de plusieurs polluants atmosphériques denatures différentes (NO2, benzène, PM10 et PM2.5) ont été finement modélisés (2m x 2m). La compatibilité des modèlesa été maximisée par l’introduction d’inputs identiques dans les méthodologies de calcul Mithra-SIG pour le bruit, etCOPERT4/ADMS-urban pour les polluants. Dans un premier temps, l’impact de la définition du voisinage sur lesniveaux d’exposition a été exploré pour les deux pollutions. Douze indicateurs, présentant chacun une définitiondifférente de la zone d’exposition au voisinage du domicile, ont permis de quantifier les niveaux d'exposition autour des10865 bâtiments d’habitations bisontins. Les résultats obtenus mettent en évidence un impact significatif de ladéfinition de la zone d’exposition sur les niveaux. Les différentes définitions du voisinage conduisent à l'existence d'unbiais différentiel, dû à la prise en compte particulière des variables environnementales dans chacun des indicateurs. Cebiais se retrouve pour l'ensemble des polluants étudiés, bien que dans des proportions variables en fonction du polluant.Dans un second temps, l’étude des situations de multi-exposition bruit/NO2 a pu être réalisée. Les résultats obtenusindiquent l'existence d'une relation complexe et permettent de décrire les différentes situations de multi-expositionprésentes dans la ville. Ce travail de thèse constitue une des premières approches de ce type en France et permet demieux comprendre les relations existantes entre définition du voisinage, zones d'exposition, paramètresenvironnementaux et niveaux d'exposition. / Every day, 15% of the European population is exposed to average noise or air pollution levels exceeding theEuropean legal threshold. Levels and distribution of each pollutant across the city depend of individual sourcesemissions, dispersion phenomenon, and urban environment. Consequently, spatial variability of multi-exposure levelremains mostly unknown. The aim of this Ph. D. thesis is to identify and describe multi-exposure situations in the cityof Besançon. Environmental noise levels, and concentrations of several air pollutants (NO2, benzene, PM10 and PM2.5),have been modeled using a fine scale grid (2m x 2m). Model compatibility has been maximized by the use of commoninput into the calculation software Mithra-SIG (for noise) and COPERT4/ADMS-urban (for air pollutants). In a firsttime, impact of the neighborhood definition on exposure levels has been explored for both pollutions. Twelveindicators, each representing a different definition of the exposure area in the dwelling vicinity, have been used toquantify exposure levels around the 10 865 inhabitable building of Besançon. Obtained results show a significantimpact of the neighborhood definition on the exposure levels. Different neighborhood definitions lead to a differentialbias caused by a different accounting for environmental variables. This bias exists for all studied pollutant, but indifferent proportion depending on the pollutant. In a second time, situations of multi-exposure to noise and NO2 havebeen identified in the city. Results indicate a complex relation between pollutants and allow the description of severalmulti-exposure situations across the city. This work constitute one of the firsts approaches of this kind in France andallow a better understanding of the existing relations between neighborhood, exposure area, environmental parametersand exposure levels.