Cette thèse se compose de deux parties, concernant deux études expérimentales sur les mouvements oculaires de poursuite lente d'un stimulus visuel en mouvement (barre inclinée). La première étude aborde l'intégration dynamique de signaux locaux de mouvement visuel provenant de la rétine, tandis que la seconde porte sur l'influence de signaux extra-rétiniens sur l'intégration du mouvement. Un cadre théorique plus général est également proposé, sur la base d'un modèle bayésien récurrent et hiérarchique pour la poursuite lente. Pour la première étude, l'intégration dynamique de mouvement a été analysée en variant le contraste et la vitesse de la barre inclinée. Les résultats montrent que des vitesses plus élevées et des valeurs plus basses de contraste produisent un plus fort biais dans la direction initiale de poursuite et que successivement la dynamique d'intégration de mouvement est plus lente pour les contrastes faibles. Une version en boucle ouverte d'un modèle bayésien est proposée, où un réseau bayésien récurrent est connecté en cascade avec un modèle du système oculomoteur pour générer des réponses de poursuite lente. Les réponses du modèle reproduisent qualitativement les différentes dynamiques observées dans les réponses de poursuite à la barre inclinée en fonction des vitesses et des contrastes différents. La deuxième étude a enquêté sur les interactions dynamiques entre les signaux rétiniens et extra-rétiniens dans l'intégration dynamique de mouvement pour la poursuite lente par le moyen d'une suppression transitoire de la cible à différents moments de la poursuite, et notamment au cours de la phase de boucle ouverte et pendant l'état d'équilibre. / This thesis addresses two studies by studying smooth pursuit eye movements for a translating tilted bar stimulus. First, the dynamic integration of local visual motion signals originating from retina and second, the influence of extra-retinal signals on motion integration. It also proposes a more generalized, hierarchical recurrent bayesian framework for smooth pursuit. The first study involved investigating dynamic motion integration for varying contrasts and speeds using a tilted bar stimuli. Results show that higher speeds and lower contrasts result in higher initial direction bias and subsequent dynamics of motion integration is slower for lower contrasts. It proposes an open-loop version of a recurrent bayesian model where a recurrent bayesian network is cascaded with an oculomotor plant to generate smooth pursuit responses. The model responses qualitatively account for the different dynamics observed in smooth pursuit responses to tilted bar stimulus at different speeds and contrasts. The second study investigated the dynamic interactions between retinal and extra-retinal signals in dynamic motion integration for smooth pursuit by transiently blanking the target at different moments during open-loop and steady-state phases of pursuit. The results suggest that weights to retinal and extra-retinal signals are dynamic in nature and extra-retinal signals dominate retinal signals on target reappearance after a blank introduced during open-loop of pursuit when compared to a blank introduced during steady-state of pursuit. The previous version of the model is updated to a closed-loop version and extended to a hierarchical recurrent bayesian model.

Les «Niayes», zones humides côtières du Sénégal constituent des écosystèmes fragiles, richesen biodiversité tout en étant traditionnellement des zones de maraîchage dans unenvironnement sahélien. La grande sécheresse des années 1970 qui a frappé l’ensemble despays du Sahel, s’est répercutée sur les Niayes par un afflux de populations rurales venant del’intérieur du pays pour y trouver des moyens de subsistance. Outre la sécheresse, cettemigration massive a considérablement augmenté la pression foncière et engendré à la fois uneurbanisation mal maîtrisée, notamment autour des grandes agglomérations, et une mise envaleur agricole; les deux portant atteinte à la valeur environnementale de ces écosystèmescôtiers. Depuis quelques années, on assiste à des hivernages particulièrement pluvieux. Ceretour pluviométrique intervenant dans des zones urbanisées de façon anarchique provoquedes inondations accompagnées de conditions sanitaires précaires pour les populations les plusdéfavorisées. A la variabilité pluviométrique, risque d’apparence naturel, s’joutent desfacteurs de risques anthropiques comme l’aménagement du territoire non maîtrisé, le nonrespectdu cadre réglementaire, diverses pollutions d’origine agricole et industrielle.Cette thèse, après avoir introduit la notion de zone humide et avoir présenté les Niayes de laGrande Côte en général et des trois zones d’étude focus (Pikine, banlieue de Dakar ; Mboro,région de Thiès; Saint-Louis), retrace leurs problèmes environnementaux et les différentespolitiques d’environnement. L’évolution de l’occupation du sol des Niayes est étudiée parrecours à l’imagerie satellitale pour la période 1986-2010 pour la confronter à la gestion deces écosystèmes sensibles et s’interroger sur la pertinence des politiques mises en place dansun esprit de développement durable. / The “Niayes”, coastal wetlands of Senegal, constitute fragile ecosystems that are rich inbiodiversity while traditionally being gardening areas in the “Sahelien” environment. The bigdrought of the 1970s that struck all the Sahel countries also affected the “Niayes” by an influxof rural populations coming from inside the country in search of livelihood. Besides theextreme drought, mass migration considerably increased the proprietary pressure and at thesame time generated an uncontrolled urbanization, especially around the big agglomerationsand agricultural development: both of them undermine the environmental value of thesecoastal ecosystems. In recent years, we have witnessed particularly rainy seasons. This returneof rainfall involved in urbanized areas causes flooding, accompanied by precarious sanitaryconditions for the most unprivileged populations. The rainfall variability, risk of naturalappearance, is added to anthropogenic risk factors such as the non-mastered regionalplanning, the failure of the regulatory framework, and the pollution from various agriculturaland industrial sources.This thesis, after having introduced the concept of wet zones and having presented the Niayesof the Great Coast in general, and of the three study areas (Pikine, outskirts of Dakar; Mboro,Thies region; and Saint-Louis) in particular, traces their environmental problems and variouspolitical differences. Land use of the Niayes is analysed by using remote sensing data for theperiod 1986-2010 to confront its evolution to the management of the sensible ecosystems andquestion the relevance of the implemented politics in the spirit of sustainable development.

Falling in indoor home setting can be dangerous for elderly population (in USA and globally), causing hospitalization, long term reduced mobility, disability or even death. Prevention of fall by monitoring different human activities or identifying the aftermath of fall has greater significance for elderly population. This is possible due to the availability and emergence of miniaturized sensors with advanced electronics and data analytics tools. This thesis aims at developing machine learning models to classify fall activities and non-fall activities. In this thesis, two types of neural networks with different parameters were tested for their capability in dealing with such tasks. A publicly available dataset was used to conduct the experiments. The two types of neural network models, convolution and recurrent neural network, were developed and evaluated. Convolution neural network achieved an accuracy of over 95% for classifying fall and non-fall activities. Recurrent neural network provided an accuracy of over 97% accuracy in predicting fall, non-fall and a third category activity (defined in this study as “pre/postcondition”). Both neural network models show high potential for being used in fall prevention and management activity. Moreover, two theoretical designs of fall detection systems were proposed in this thesis based on the developed convolution and recurrent neural networks.

Orientador: Regina Helena Garcia Martins / Resumo: Mercuri G. Incidência e Prevalência da Papilomatose Laríngea no Estado de São Paulo [dissertação]. Botucatu, SP: Faculdade de Medicina de Botucatu, Universidade Estadual Paulista; 2019. Introdução: A papilomatose respiratória recorrente ou papilomatose laríngea recorrente é a neoplasia benigna da laringe causada pelo Papiloma Vírus Humano, caracterizando-se pela presença de lesões proliferativas exofíticas recorrentes. A ausência de tratamento curativo, os custos relacionados ao diagnóstico, prevenção e tratamento das doenças causadas pelo HPV apresentam importante impacto econômico em todo mundo. No entanto, a falta de dados epidemiológicos nacionais, quanto à incidência e prevalência da papilomatose laríngea no Brasil não nos permite analisar o comportamento da doença em nossa população, bem como os benefícios da vacinação. Objetivo: Estimar a incidência e a prevalência da Papilomatose Laríngea no Estado de São Paulo. Métodos: Estudo populacional que objetivou levantar todos os casos diagnosticados de Papilomatose Laríngea do Estado de São Paulo, considerando os principais centros de diagnóstico desta enfermidade. Foram mapeados e selecionados os principais serviços de residência e especialização em Otorrinolaringologia do Estado de São Paulo, a partir dos dados fornecidos pela Associação Brasileira de Otorrinolaringologia e Cirurgia Cérvico-Facial (ABORL-CCF). Através do preenchimento de um questionário obtivemos os dados para estimar a prevalência e a incidência mínimas p... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Mercuri G. Incidence and Prevalence of Laryngeal Papillomatosis in São Paulo State [thesis]. Botucatu, SP: Faculty of Medicine of Botucatu, Universidade Estadual Paulista; 2019. Introduction: Recurrent respiratory papillomatosis or recurrent laryngeal papillomatosis is a benign laryngeal neoplasm caused by Human Papilloma Virus, characterized by the presence of recurrent exophytic proliferative lesions. Without curative treatment, costs related to the diagnosis, prevention, and treatment of HPV-related diseases have a significant economic impact worldwide. However, the lack of national epidemiological data on the incidence and prevalence of laryngeal papillomatosis in Brazil does not allow us to analyze the behavior of the disease in our population, as well as the benefits of vaccination. Objective: To estimate the incidence and prevalence of Laryngeal Papillomatosis in the State of São Paulo. Methods: A population study was conducted aiming to raise all the diagnosed cases of Laryngeal Papillomatosis of the State of São Paulo, considering the main centers of diagnosis of this disease. The main services of residence and specialization in Otorhinolaryngology of the State of São Paulo were mapped and selected, based on the data provided by the Brazilian Association of Otorhinolaryngology and Cervical-Facial Surgery (ABORL-CCF). By completing a questionnaire we obtained the data to estimate the minimum prevalence and incidence for the State in 2017. Results: The questionnaire wa... (Complete abstract click electronic access below) / Mestre

Aborto espontâneo recorrente (AER) é definido pela presença de três ou mais abortos espontâneos e consecutivos que ocorreram até a 20ª semana de gestação. O AER possui origem multifatorial. Dentre os diversos fatores associados ao AER, alterações na hemostasia podem comprometer o fluxo sanguíneo na placenta e com isso pode aumentar o risco de complicações obstétricas, como o aborto. O objetivo geral deste estudo foi investigar se existe associação entre polimorfismos genéticos (no gene do fibrinogênio (FGB -455G>A e -148C>T), da trombomodulina (THBD 1418C>T), do fator V (F5 1691G>A), da protrombina (F2 20210 G>A), do PAI-1 (SERPINE1 4G/5G) e do TAFI (CPB2<i/> c.505G>A)) e os abortos espontâneos recorrentes (primários e secundários). Os objetivos específicos desse estudo foram: 1- avaliar se existe associação entre os sete polimorfismos e o período em que ocorreram as perdas fetais (precoce ou tardia) e o número de abortos recorrentes; 2- determinar se os haplótipos dos polimorfismos FGB -455G>A e FGB -148C>T estão ou não associados aos abortos primários e secundários. Foram incluídas 256 mulheres com história de abortos espontâneos recorrentes, provenientes do Ambulatório de Obstetrícia da Clínica Obstétrica do Hospital das Clínicas da Faculdade de Medicina da USP e 264 mulheres saudáveis, sem história de aborto espontâneo e que tiveram pelo menos duas gestações normais (grupo controle), pareadas segundo as idades. Amostras de sangue foram obtidas para realização das genotipagens dos polimorfismos por meio de PCR em tempo real (FGB -148C>T, FGB -455G>A, THBD 1418C>T e CBP2 c.505G>A), e PCR-RFLP (SERPINE14G/5G, F5 1691G>A e F2 20210 G>A). As frequências dos genótipos e de alelos para os sete polimorfismos foram semelhantes entre os grupos aborto primário, aborto secundário e grupo controle. Entretanto, quando foi realizada um modelo de regressão logística multivariada saturada, que incluiu as variáveis independentes: F5 1691G>A (referência GG vs GA), F2 20210G>A (referência GG vs GA), CBP2 c.505G>A (referência GG + GA vs AA), THBD 1418C>T (referência CC + CT vs TT), SERPINE1 4G/5G (referência 5G/5G vs 4G/4G + 4G/5G) FGB -455G>A (referência GG vs GA vs AA) e FGB -148C>T (referência CC vs CT vs TT), apenas o polimorfismo FGB -148C>T foi associado ao maior risco de ter aborto primário (OR: 2,91, IC 95% 1,02 - 8,29, p=0,045). Quando os haplótipos para os polimorfismos FGB -455G>A e FGB 148C>T foram considerados, foi observada maior frequência de haplótipo 455G/148T em mulheres com AER primário (3,4%) do que no grupo controle (1,1%), (p=0,030); porém esse efeito não foi observado no AER secundário. Em relação ao número de abortos consecutivos, houve uma tendência (p=0,060) a maior frequência de genótipo TT para o polimorfismo FGB -148C>T no grupo de aborto primário com até três perdas quando comparado com as mulheres do mesmo grupo, porém com número maior de perdas (>3). Em conclusão, os sete polimorfismos quando analisados separadamente, não foram associados ao AER; no entanto, em modelo multivariado de regressão logística, o genótipo TT do polimorfismo FGB 148C>T foi associado com o aumento do risco de ter AER primário. Além disso, foi encontrado maior frequência do haplótipo 455G/148T para os polimorfismos FGB -455G>A e FGB -148C>T em mulheres com aborto primário. / Recurrent pregnancy loss (RPL) is defined by the presence of three or more consecutive losses prior to 20 weeks of gestation. The RPL has multifactorial origin. Among several factors associated with RPL, changes in hemostasis may impair the blood flow in the placenta and thus may increase the risk of obstetric complications, such as pregnancy loss. The general aim of this study was to investigate the association between genetic polymorphisms (in the genes of fibrinogen (FGB -455G>A and -148C>T), thrombomodulin (THBD 1418C>T), factor V (F5 1691G>A), prothrombin (F2 20210 G>A), PAI-1 (SERPINE1 4G/5G) and TAFI (CPB2 c.505G>A)) and recurrent pregnant losses (primary and secondary). The specific aims of this study were: 1 - to evaluate the association between the seven polymorphisms and the period in which the fetal losses occurred (early or late) and the number of recurrent losses; 2 - to determine if the haplotypes of polymorphisms FGB -455G>A and FGB -148C>T present association with primary and secondary pregnant losses. We included 256 women with a RPL history, from the Ambulatory of Obstetrics from Clinical Hospital of the Medical School of USP and 264 healthy women without losses history that have had at least two normal pregnancies (control group), matched according to age. Blood samples were obtained to perform the genotyping of polymorphisms by real-time PCR (FGB -148C>T, FGB -455G>A, THBD 1418C>T and CBP2 c.505G>A), and PCR-RFLP (SERPINE1 4G/5G, F5 1691G>A and F2 20210G>A). The frequencies of genotypes and alleles for the seven genetic polymorphisms were similar in 3 groups. However, when it was performed a model of multivariate logistic regression, which included the independent variables: F5 1691G>A (GG vs GA reference), F2 20210G>A (GG vs GA reference), CBP2 c.505G>A (GG + GA reference vs AA), THBD 1418C>T (reference CC + CT vs TT), SERPINE1 4G/5G (reference 5G/5G + 4G/5G vs 4G/4G), FGB -455G>A (GG reference vs GA vs AA) and FGB - 148C>T (reference CC vs CT vs TT), only the polymorphism FGB 148C>T polymorphism was associated with a higher risk of having primary losses (OR: 2.91, 95% CI 1.02 to 8.29, p = 0.045). When the haplotypes for the polymorphisms FGB -455G>A and FGB -148C>T were considered, had a higher frequency of the haplotype 455G/148T in women with primary RPL (3.4%) than in the control group (1.1%) (p = 0.030); but this effect was not observed in secondary RPL. Regarding the number of successive pregnant losses, there was a trend (p = 0.060) to higher frequency of the TT genotype for FGB -148C>T polymorphism in the group with primary RPL up to three losses when compared with women of the same group, but with loss number higher than three. In conclusion, when the seven genetic polymorphisms were evaluated separately, they do not show association with RPL, however, in multivariate logistic regression analysis, the TT genotype of the FGB -148C>T polymorphism was associated with increased risk for primary RPL. Furthermore, it was found higher frequency of the haplotype 455G/148T for the FGB -455G>A and FGB -148 C>T polymorphisms in women with primary RPL.

Under recent years the gambling industry has been moving towards providing their customer the possibility to gamble online instead of visiting a physical location. Aggressive marketing, fast growth and a multitude of actors within the market have resulted in a spike of customers who have developed a gambling problem. Decision makers are trying to fight back by regulating markets in order to make the companies take responsibility and work towards preventing these problems. One method of working proactively in this regards is to identify vulnerable customers before they develop a destructive habit. In this work a novel method of predicting customers that have a higher risk in regards to gambling-related problems is explored. More concretely, a recurrent neural network with long short-term memory cells is created to process raw behaviour data that are aggregated on a daily basis to classify them as high-risk or not. Supervised training is used in order to learn from historical data, where the usage of permanent self-exclusions due to gambling related problems defines problem gamblers. The work consists of: obtain a local optimal configuration of the network which enhances the performance for identifying problem gam- blers who favour the casino section over sports section, and analyze the model to provide insights in the field. This project was carried out together with LeoVegas Mobile Gaming Group. The group offers both online casino games and sports booking in a number of countries in Europe. This collaboration made both data and expertise within the industry accessible to perform this work. The company currently have a model in production to perform these predictions, but want to explore other approaches. The model that has been developed showed a significant increase in performance compared to the one that is currently used at the company. Specifically, the precision and recall which are two metrics important for a two class classification model, increased by 37% and 21% respectively. Using raw time series data, instead of aggregated data increased the responsiveness regarding customers change in behaviour over time. The model also scaled better with more history compared to the current model, which could be a result of the nature of a recurrent network compared to the current model used.

Nesta tese é analisado situações onde eventos de interesse podem ocorrer mais que uma vez para o mesmo indivíduo. Embora os estudos nessa área tenham recebido considerável atenção nos últimos anos, as técnicas que podem ser aplicadas a esses casos especiais ainda são pouco exploradas. Além disso, em problemas desse tipo, é razoável supor que existe dependência entre as observações. Uma das formas de incorporá-la é introduzir um efeito aleatório na modelagem da função de risco, dando origem aos modelos de fragilidade. Esses modelos, em análise de sobrevivência, visam descrever a heterogeneidade não observada entre as unidades em estudo. Os modelos estatísticos apresentados neste texto são fundamentalmente modelos de sobrevivência baseados em processos de contagem, onde é representado o problema como um processo de Poisson homogêneo e não-homogêneo com um termo de fragilidade, para o qual um indivíduo com um dado vetor de covariável x é acometido pela ocorrência de eventos repetidos. Esses modelos estão divididos em duas classes: modelos de fragilidade multiplicativos e aditivos; ambos visam responder às diferentes formas de avaliar a influência da heterogeneidade entre as unidades na função de intensidade dos processos de contagem. Até agora, a maioria dos estudos tem usado a distribuição gama para o termo de fragilidade, a qual é matematicamente conveniente. Este trabalho mostra que a distribuição gaussiana inversa tem propriedade igualmente simples à distribuição gama. Consequências das diferentes distribuições são examinadas, visando mostrar que a escolha da distribuição de fragilidade é importante. O objetivo deste trabalho é propor alguns métodos estatísticos para a análise de eventos recorrentes e verificar o efeito da introdução do termo aleatório no modelo por meio do estudo do custo, da estimação dos outros parâmetros de interesse. Também um estudo de simulação bootstrap é apresentado para fazer inferências dos parâmetros de interesse. Além disso, uma abordagem Bayesiana é proposta para os modelos de fragilidade multiplicativos e aditivos. Métodos de simulações são utilizados para avaliar as quantidades de interesse a posteriori. Por fim para ilustrar a metodologia, considera-se um conjunto de dados reais sobre um estudo dos resultados experimentais de animais cancerígenos. / In this thesis we analyse situations where events of interest may occur more than once for the same individual and it is reasonable to assume that there is dependency among the observations. A way of incorporating this dependency is to introduce a random effect in the modelling include a frailty term in the intensity function. The statistical methods presented here are intensity models based, where we represent the problem as a homogeneous and nonhomogeneous Poisson process with a frailty term for which an individual with given fixed covariate vector x has reccurent events occuring. These models are divided into two classes: multiplicative and additive models, aiming to answer the different ways of assessing the influence of heterogeneity among individuals in the intensity function of the couting processes. Until now most of the studies have used a frailty gamma distribution, due to mathematical convenience. In this work however we show that a frailty gaussian inverse distribution has equally simple proprieties when compared to a frailty gamma distribution. Methods for regression analysis are presented where we verify the effect of the frailty term in the model through of the study of the cost of estimating the other parameters of interest. We also use the simulation bootstrap method to make inference on the parameters of interest. Besides we develop a Bayesian approach for the homogeneous and nonhomogeneous Poisson process with multiplicative and additive frailty. Simulation methods are used to assess the posterior quantities of interest. In order to ilustrate our methodology we considere a real data set on results of an experimental animal carcinogenesis study.

O presente estudo objetivou estimar parâmetros genéticos visando quantificar a variabilidade genética de uma população de café robusta (Coffea canephora Pierre ex A. Froehner) introduzida da Costa Rica e analisar o seu potencial genético para o desenvolvimento de futuras cultivares clonais para o estado de São Paulo. Outro intuito foi verificar a possibilidade de submeter essa população à seleção recorrente, tornando-a, assim, fonte de alimentação e sustentação de programas de melhoramento genético do café robusta. O experimento foi composto por 25 tratamentos, sendo 21 progênies de C. canephora e quatro cultivares de C. arabica, plantados em Mococa (SP). O delineamento experimental utilizado foi em látice balanceado 5x5 quadruplicado, com seis repetições e uma planta por parcela. Foram realizadas doze colheitas e após a sexta colheita as plantas foram podadas. Em 2004, foi realizada uma seleção fenotípica dessa população a fim de clonar os melhores indivíduos. Essa seleção resultou em novo experimento, instalado em Campinas, seguindo o delineamento de blocos ao acaso, com três repetições, 28 clones e quatro plantas por parcela, sendo realizadas cinco colheitas consecutivas. As análises estatísticas e biométricas foram realizadas considerando os modelos lineares mistos (procedimento REML/BLUP), por meio do software Selegen, cujos componentes de variância são estimados pelo método da máxima verossimilhança restrita (REML) e os valores genotípicos preditos pela melhor predição linear não viesado (BLUP). As análises mostraram que, na população em estudo, observou-se elevada variabilidade genética, passível de ser explorada tanto para a extração de clones quanto para a seleção recorrente. As adversidades climáticas severas fizeram com que a seleção fosse prejudicada. Nessa situação é preferível não considerar o período afetado e analisar os dados após a recuperação das plantas. A seleção baseada em seis colheitas forneceu estimativas de parâmetros e ganhos genéticos similares aos obtidos na seleção baseada em duas colheitas de alta produção. Os ganhos genéticos esperados nas duas formas de propagações foram elevados e a seleção clonal proporcionou maiores ganhos do que a sexual. No experimento clonal foi possível identificar materiais com potencial produtivo e que poderão vir a ser recomendados para o cultivo no estado de São Paulo. Apesar de a interação genótipos x colheitas ter sido do tipo complexa, devido ao veranico ocorrido, esta não afetou significativamente o ordenamento dos melhores clones e nem comprometeu as estimativas dos parâmetros genéticos. Os coeficientes de variação experimental e genético bem como seu valor relativo deverão ser analisados conjuntamente com o número de repetições e a acurácia seletiva. A seleção recorrente deverá ser conduzida concomitantemente com o programa de seleção clonal, a fim de evitar o esgotamento da variabilidade genética e o comprometimento do programa de melhoramento genético visando o desenvolvimento de cultivares clonais. Tendo em vista que a população inicial foi constituída por um pequeno número de progênies, é aconselhável o monitoramento do tamanho efetivo populacional e do grau de endogamia ao longo dos ciclos de seleção recorrente. / The objective of this research was to estimate genetic parameters to quantify the variability of a population of robusta coffee (Coffea canephora Pierre ex A. Froehner) introduced into Brazil from Costa Rica in 1974 aiming at determining its genetic potential for the development of clonal or seedling cultivars for the state of São Paulo, Brazil. The feasibility has also been studied to submit this population to recurrent selection, making it a continuous source of improved base material in support of varietal improvement of robusta. An experiment consisting of 21 open pollinated seedling progenies of robusta and four cultivars of arabica was established in Mococa (SP) in 1975. Yield was observed for twelve harvests and after the sixth harvest the plants were pruned. The experimental lay out was a balanced 5x5 quadruple lattice design, with six replicates and one plant per plot. In 2004 a phenotypic selection of this population for yield was carried out aiming at cloning the best individuals. These 28 clones were planted in an experiment in Campinas in 2005, following a completely randomized block design, with 28 treatments (clones), three replications and four plants per plot. In total, yields were collected over five harvests. Statistical and biometrical analyzes were performed considering the linear mixed models (REML/BLUP), through software Selegen, where the variance components are estimated by restricted maximum likelihood (REML) and genotypic values predicted by best linear unbiased prediction (BLUP). The analyzes showed that the population had high genetic variability, which can be exploited for the extraction of both clones and seedling progenies, used for recurrent selection. Selection was impaired by severe adverse weather conditions. In such situations it is preferable not to consider the affected period and analyze the data after recovery of the plants. Due to the moisture stress that occurred in the clonal trial, genotype x environment interaction was complex. However this did not affect the ranking of superior clones nor compromised the genetic parameter estimates. Selection for yield based on six yield resulted in genetic parameters and genetic gains similar to those obtained by selection based on two high yielding harvesting periods. The expected genetic gains both for clones as for open pollinated progenies were high. However, clonal selection resulted in higher genetic gains for yield than the seedling selection. In the clonal experiment it was possible to identify materials with high yield potential that may become to be recommended for cultivation in São Paulo State. The experimental, genetics and relative coefficients of variation, should be analyzed together with the number of replications and selective accuracy. Recurrent selection should be conducted concurrently with the clonal selection program in order to avoid depletion of genetic variability and to impairthe breeding program aiming at the development of clonal cultivars. Considering that the initial population was composed of a small number of progeny, it will be important monitoring adequately the effective size and the inbreeding coefficient during recurrent selection cycles.

In der chirurgischen Therapie von Schilddrüsenerkrankungen bestehen weiterhin kontroverse Diskussionen hinsichtlich der adäquaten Operationstechnik bei bilateralen Resektionen. Rezidiv- und Komplikationsvermeidung stellen hohe Anforderungen an die Schilddrüsenchirurgie. Darüberhinaus bestehen weltweit divergierende Meinungen über Risikofaktoren für die spezifischen Operationskomplikationen, die postoperative Rekurrensparese und Hypokalzämie. In der chirurgischen Abteilung der Universitätsklinik Charité, Campus Virchow-Klinikum wurden im Rahmen einer prospektiven Schilddrüsenstudie 2728 Patientendaten aus den Jahren 1985-1999 erhoben und statistisch ausgewertet. Anhand der multivariaten Regressionsanalyse wurde gezielt nach Risikofaktoren für spezifische Operationskomplikationen gesucht. Das individuell indizierte chirurgische Vorgehen basiert auf dem operativen Standard, der neben überdurchschnittlich häufig angewandten ausgedehnten Resektionsverfahren ein Operieren in exakter Mikrodissektionstechnik mit Darstellung und Schonung der anatomischen Leitstrukturen beinhält. Bei den spezifischen Komplikationen waren 4,3% transiente und 0,4% permanente Rekurrensparesen sowie 19,5% transiente und 0,9% permanente Hypokalzämien zu verzeichnen. Die multivariate Analyse wurde sowohl für das Gesamtpatientengut als auch nur für Patienten mit benignen Schilddrüsenerkrankungen durchgeführt. Für die erste Gruppe ergab die Thorakotomie einen Risikofaktor für eine postoperative Rekurrensparese und Frauen, Strumen mit retrosternalem Anteil sowie Near Total-Operationsverfahren waren jeweils Risikofaktoren für eine postoperative Hypokalzämie. Für die zweite, benigne Patientengruppe stellten beidseitig voroperierte Patienten ein signifikantes Risiko für eine Rekurrensparese und Morbus Basedow-Patienten, Patienten mit sehr grossen Schilddrüsen (WHO III Struma, Schilddrüsen mit retrosternalem Anteil) sowie Near Total-Verfahren jeweils ein Risikofaktor hinsichtlich einer postoperativen, überwiegend transienten Hypokalzämie dar. Eine solide Interpretation der gewonnen Ergebnisse erlaubt die Schlussfolgerung bei Beachtung einer kleinen Gruppe von Risikopatienten, dass unter Berücksichtigung klarer Standards für Indikation und Operationsverfahren Schilddrüseneingriffe heutzutage sehr sicher und effektiv durchzuführen sind. / The surgical therapy of thyroid diseases gives reason for controverse discussion concearning the appropriate surgical technique by bilateral resection. The avoidance of recurrent goiter and surgical complications is a great demand for thyroid surgery. In addition to that exist worldwide diverge opinions about riskfactors for the specific complications after thyroid operations, recurrent laryngeal nerve palsy and hypocalcaemia. In the department of surgery, universityhospital Charité, campus Virchow-Klinikum, Berlin, Germany were 2728 patientfacts collected and statistically evaluated within a prospective study from 1985-1999. By means of the multivariate logistic regression analysis riskfactors for the specific surgical complications were searched. The individually indicated surgical procedure based upon standardized surgery, that contains beside numerous extended resections than average also the exact mikrodissection technique with preparation and preservation of all important anatomical structures. These study showed 4,3% transient and 0,4% permanent recurrent laryngeal nerve palsies as well as 19,5% transient and 0,9% permanent hypocalcaemiae. The whole number of patients and the patients with benign thyroid diseases were analysed through multivariat logistic regression separately. So, for the first group thoracotomy proved to be a riskfactor for recurrent nerve palsy just as femal patients, retrosternal goiter and near total resections showed to be riskfactors for hypocalcaemia. For the second, the benign group bilateral reoperation for recurrent goiter revealed to be riskfactor for recurrent nerve palsy. Graves´~disease, very large goiter (WHO III and retrosternal goiter) and near total resection showed to be a riskfaktor for a hypocalcaemia (mainly transient). A solid interpretation of the achieved results allows the conclusion by attention to a small group of riskpatients, that today thyroid surgery can be performed very safely and effectively when standards for indication and surgical techniques are considered.

This thesis puts forward a novel way of control for robotic morphologies. Taking inspiration from Behaviour Based robotics and self-organisation principles, we present an interfacing mechanism, capable of adapting both to the user and the robot, while enabling a paradigm of intuitive control for the user. A transparent mechanism is presented, allowing for a seamless integration of control signals and robot behaviours. Instead of the user adapting to the interface and control paradigm, the proposed architecture allows the user to shape the control motifs in their way of preference, moving away from the cases where the user has to read and understand operation manuals or has to learn to operate a specific device. The seminal idea behind the work presented is the coupling of intuitive human behaviours with the dynamics of a machine in order to control and direct the machine dynamics. Starting from a tabula rasa basis, the architectures presented are able to identify control patterns (behaviours) for any given robotic morphology and successfully merge them with control signals from the user, regardless of the input device used. We provide a deep insight in the advantages of behaviour coupling, investigating the proposed system in detail, providing evidence for and quantifying emergent properties of the models proposed. The structural components of the interface are presented and assessed both individually and as a whole, as are inherent properties of the architectures. The proposed system is examined and tested both in vitro and in vivo, and is shown to work even in cases of complicated environments, as well as, complicated robotic morphologies. As a whole, this paradigm of control is found to highlight the potential for a change in the paradigm of robotic control, and a new level in the taxonomy of human in the loop systems.