La radiomique suppose que des informations pertinentes non repérables visuellement peuvent être trouvées en calculant une grande quantité d’indices quantitatifs à partir des images médicales. En cancérologie, ces informations pourraient caractériser le phénotype de la tumeur et définir le pronostic du patient. Le GITC est une tumeur pédiatrique rare diagnostiquée d'après des signes cliniques et son apparence en IRM. Cette thèse présente les premières études radiomiques pour des patients atteints de GITC. Comme les intensités en IRM clinique sont exprimées en unités arbitraires, la première étape de l’étude a été la standardisation des images. Une méthode de normalisation basée sur l'estimation de l'intensité dans la matière blanche d'apparence normale s’est avérée efficace sur plus de 1500 volumes d'images. Des études méthodologiques sur le calcul des indices de texture ont abouti aux recommandations suivantes : (a) discrétiser les niveaux de gris avec une largeur constante pour tous les patients, (b) utiliser un volume d'intérêt constant ou faire attention au biais introduit par des volumes de taille et forme différentes. En s’appuyant sur ces recommandations, les indices radiomiques issus de 4 modalités d'IRM ont été systématiquement analysés en vue de prédire les principales mutations génétiques associées aux GITC et la survie globale des patients au moment du diagnostic. Un pipeline de sélection d’indices a été proposé et différentes méthodes d’apprentissage automatique avec validation croisée ont été mises en oeuvre pour les deux tâches de prédiction. La combinaison des indices cliniques avec les indices d’imagerie est plus efficace que les indices cliniques ou d’imagerie seuls pour la prédiction des deux principales mutations de l’histone H3 (H3.1 versus H3.3) associées au GITC. Comme certaines modalités d'imagerie étaient manquantes, une méthodologie adaptée à l’analyse des bases de données d’imagerie multi-modales avec données manquantes a été proposée pour pallier les limites de recueil des données d'imagerie. Cette approche permet d'intégrer de nouveaux patients. Les résultats du test externe de prédiction des deux principales mutations de l’histone H3 sont encourageants. Concernant la survie, certains indices radiomiques semblent informatifs. Toutefois, le faible nombre de patients n'a pas permis d'établir les performances des prédicteurs proposés. Enfin, ces premières études radiomiques suggèrent la pertinence des indices radiomiques pour la prise en charge des patients atteints de GITC en absence de biopsie mais l’augmentation de la base de données est nécessaire pour confirmer ces résultats. La méthodologie proposée dans cette thèse peut être appliquée à d'autres études cliniques. / Radiomics is based on the assumption that relevant, non-visually identifiable information can be found by calculating a large amount of quantitative indices from medical images. In oncology, this information could characterize the phenotype of the tumor and define the prognosis of the patient. DIPG is a rare pediatric tumor diagnosed by clinical signs and MRI appearance. This work presents the first radiomic studies for patients with DIPG. Since clinical MRI intensities are expressed in arbitrary units, the first step in the study was image standardization. A normalization method based on intensity estimation of the normal-appearing white matter has been shown to be effective on more than 1500 image volumes. Methodological studies on the calculation of texture indices have then defined the following recommendations: (a) discretize gray levels with a constant width for all patients, (b) use a constant volume of interest or pay attention to the bias introduced by volumes of different size and shape. Based on these recommendations, radiomic indices from four MRI modalities were systematically analyzed to predict the main genetic mutations associated with DIPG and the overall survival of patients at the time of diagnosis. An index selection pipeline was proposed and different cross-validated machine learning methods were implemented for both prediction tasks. The combination of clinical indices with imaging indices is more effective than the clinical or imaging indices alone for the prediction of the two main mutations in histone H3 (H3.1 versus H3.3) associated with DIPG. As some imaging modalities were missing, a methodology adapted to the analysis of multi-modal imaging databases with missing data was proposed to overcome the limitations of the collection of imaging data. This approach made it possible to integrate new patients. The results of the external prediction test for the two main mutations of H3 histone are encouraging. Regarding survival, some radiomic indices seem to be informative. However, the small number of patients did not make it possible to establish the performance of the proposed predictors. Finally, these first radiomic studies suggest the relevance of the radiomic indices for the management of patients with DIPG in the absence of biopsy but the database need to be increased in order to confirm these results. The proposed methodology can be applied to other studies.

Since mid to late 2010 image synthesizing using neural networks has become a trending research topic. And the framework mostly used for solving these tasks is the Generative adversarial network (GAN). GAN works by using two networks, a generator and a discriminator that trains and competes alongside each other. In today’s research regarding image synthesis, it is mostly about generating or altering images in any way which could be used in many fields, for example creating virtual environments. The topic is however still in quite an early stage of its development and there are fields where image synthesizing using Generative adversarial networks fails. In this work, we will answer one thesis question regarding the limitations and discuss for example the limitation causing GAN networks to get stuck during training. In addition to some limitations with existing GAN models, the research also lacks more experimental GAN variants. It exists today a lot of different variants, where GAN has been further developed and modified. But when it comes to GAN models where the discriminator has been changed to a different network, the number of existing works reduces drastically. In this work, we will experiment and compare an existing deep convolutional generative adversarial network (DCGAN), which is a GAN variant, with one that we have modified using a deep neuro-fuzzy system. We have created the first DCGAN model that uses a deep neuro-fuzzy system as a discriminator. When comparing these models, we concluded that the performance differences are not big. But we strongly believe that with some further improvements our model can outperform the DCGAN model. This work will therefore contribute to the research with the result and knowledge of a possible improvement to DCGAN models which in the future might cause similar research to be conducted on other GANmodels.

L'un des enjeux majeurs actuellement en neurosciences est l'élaboration de modèles computationnels capables de reproduire les données obtenues expérimentalement par des méthodes d'imagerie et permettant l'étude de la relation structure-fonction dans le cerveau. Les travaux de modélisation dans cette thèse se situent à deux échelles et l'analyse des modèles a nécessité le développement d'outils théoriques et numériques dédiés. À l'échelle locale, nous avons proposé un nouveau modèle d'équations différentielles ordinaires générant des activités neuronales, caractérisé et classifié l'ensemble des comportements générés, comparé les sorties du modèle avec des données expérimentales et identifié les structures dynamiques sous-tendant la génération de comportements pathologiques. Ce modèle a ensuite été couplé bilatéralement à un nouveau compartiment modélisant les dynamiques de neuromédiateurs et leurs rétroactions sur l'activité neuronale. La caractérisation théorique de l'impact de ces rétroactions sur l'excitabilité a été obtenue en formalisant l'étude des variations d'une valeur de bifurcation en un problème d'optimisation sous contrainte. Nous avons enfin proposé un modèle de réseau, pour lequel la dynamique des noeuds est fondée sur le modèle local, incorporant deux couplages: neuronal et astrocytaire. Nous avons observé la propagation d'informations différentiellement selon ces deux couplages et leurs influences cumulées, révélé les différences qualitatives des profils d'activité neuronale et gliale de chaque noeud, et interprété les transitions entre comportements au cours du temps grâce aux structures dynamiques identifiées dans les modèles locaux. / A current issue in neuroscience is to elaborate computational models that are able to reproduce experimental data recorded with various imaging methods, and allowing us to study the relationship between structure and function in the human brain. The modeling objectives of this work are two scales and the model analysis need the development of specific theoretical and numerical tools. At the local scale, we propose a new ordinary differential equations model generating neuronal activities. We characterize and classify the behaviors the model can generate, we compare the model outputs to experimental data and we identify the dynamical structures of the neural compartment underlying the generation of pathological patterns. We then extend this approach to a new neuro-glial mass model: a bilateral coupling between the neural compartment and a new one modeling the impact of astrocytes on neurotransmitter concentrations and the feedback of these concentrations on neural activity is developed. We obtain a theoretical characterization of these feedbacks impact on neuronal excitability by formalizing the variation of a bifurcation value as a problem of optimization under constraint. Finally, we propose a network model, which node dynamics are based on the local neuro-glial mass model, embedding a neuronal coupling and a glial one. We numerically observe the differential propagations of information according to each of these coupling types and their cumulated impact, we highlight qualitatively distinct patterns of neural and glial activities of each node, and link the transitions between behaviors with the dynamical structures identified in the local models.

De nos jours, on attribue une myriade d’effets bénéfiques aux polyphénols alimentaires. Ces affirmations reposent la plupart du temps sur des études in vitro, sur quelques études in vivo et presque jamais sur des essais cliniques. On se rend compte de plus en plus que l’intérêt des polyphénols ne résiderait pas uniquement dans leur pouvoir antioxydant mais également dans leur capacité à cibler de multiples cibles moléculaires comme l’inflammation. De plus, de nombreuses études commencent à prendre en compte la biodisponibilité des polyphénols dans l’organisme et leur interaction avec le microbiote intestinal. C’est pourquoi en plus de nous intéresser à deux polyphénols issus de la Pomme-grenade, nous nous sommes intéressés à deux de leurs dérivés métaboliques. Pratiquement aucune étude ne s’était penchée sur l’effet de ces molécules dans un contexte de maladies neuro-dégénératives. Pour se faire, nous avons testé la punicalagine, l’acide ellagique, l’urolithin A et l’urolithin B sur des modèles C. elegans de la SLA et de la maladie de Huntington qui présentaient des phénotypes moteurs, de la neuro-dégénérescence et de l’inflammation. Enfin, nous avons utilisé un modèle de souris ayant subi une axotomie du nerf optique pour confirmer le pouvoir neuro-protecteur de l’urolithin A. Nos résultats ont montré que ces composés dans des proportions différentes étaient en mesure de réduire la toxicité neuronale de protéines liées à la SLA et HD et ainsi diminuer les niveaux de paralysie et de neuro-dégénérescence de nos modèles C. elegans. En parallèle, nous avons observé que cette neuro-protection se faisait au travers une diminution de l’inflammation et pour l’urolithin A une amélioration de la morphologie des mitochondries via la mitophagie. En dernier, nous avons constaté que l’urolithin A était en mesure de promouvoir la survie neuronale chez la souris à la suite d’une lésion du nerf optique. Pour conclure, cette étude par son approche in vivo de multiples maladies neuro-dégénératives renforce les preuves existantes de l’effet bénéfique de la consommation de Pomme-grenade et encourage l’utilisation pharmacologique de l’urolithin A. / Nowadays, a myriad of beneficial effects is attributed to dietary polyphenols. Most of these claims are based on in vitro studies, some in vivo studies, and almost never on clinical trials. It is increasingly realized that the interest of polyphenols does not only lie in their antioxidant power but also in their ability to target multiple molecular pathways such as inflammation. In addition, many studies are beginning to take into account the bioavailability of polyphenols in the body and their interaction with the gut microbiota. That's why in addition to two polyphenols from Pomegranate, we looked at two of their metabolic derivatives. Almost no study has examined the effect of these molecules in the context of neurodegenerative diseases. To do this, we tested punicalagin, ellagic acid, urolithin A and urolithin B on C. elegans models of ALS and Huntington's disease that had motor phenotypes, neurodegeneration and inflammation. Furthermore, we used a mouse model that underwent an axotomy of the optic nerve to confirm the neuroprotective power of urolithin A. Our results showed that these compounds in different proportions were able to reduce the neuronal toxicity of proteins. related to ALS and HD and thus decrease the levels of paralysis and neuro-degeneration of our C. elegans models. In parallel, we observed that this neuroprotection was done through a reduction of the inflammation and for urolithin A an improvement of the morphology of mitochondria via mitophagy. Lastly, we found that urolithin A was able to promote neuronal survival in mice as a result of optic nerve injury. To conclude, this study by its in vivo approach to multiple neuro-degenerative diseases reinforces existing evidence of the beneficial effect of pomegranate consumption and encourages the pharmacological use of urolithin A.

No description available.

Syftet med studien är att belysa på vilket sätt lärares icke - verbala ledarskap kan påverka kommunikationen med eleverna och den miljö de befinner sig i. I studien har observationer och intervjuer genomförts med två erfarna pedagoger. För att läraren ska lyckas skapa en god klassrumsmiljö och få med sig eleverna i undervisningen krävs att läraren lyckas skapa förtroendefulla relationer med eleven. Lärarens engagemang, skicklighet i att använda sitt kroppspråk och planering av miljön i klassrummet är faktorer som påverkar elevers lärande. Om eleverna känner förtroende i sin lärandemiljö, att de är en del i den ökar förtroendet för läraren, vilket i sin tur bidrar till goda resultat.

Research indicates that Developmental Coordination Disorder (DCD) is associated with a poor self-concept and high levels of anxiety (peens et al., 2004; Piek et al., 2000; Skinner & Piek, 2001). Research also substantiates that participation in a well planned motor intervention programme can enhance the self-concept of a child with DCD (Colchico et al., 2005). Literature further indicates that DCD is associated with neuro-motor problems which may vary in severity (Sigmundsson & Hopkins, 2005). It is further indicated that more boys than girls are diagnosed with DCD and also that, in general, boys have a higher self-concept than girls (Maldonado-Duran, 2002; Stein et al., 1998). The aim of this study was firstly, to determine the influence of DCD on the self-concept and anxiety of 7-9 year old children in the Potchefstroom district. Secondly, the study aimed to determine whether gender and the ethnic group of DCD children have an effect on the success of different intervention programmes. A third aim was to determine whether a motor based intervention programme, a self-concept enhancing programme or a combination of the two (psycho-motor intervention programme) would have the best effect on enhancing children's self-concept and motor proficiency. Lastly, the study attempted to determine whether neuro-motor problems could have a negative influence on an intervention programme for DCD children. The Movement Assessment Battery for Children (MABC), Bruininks-Oseretsky Test for Motor Proficiency (BOTMP-SF), Sensory Input Measurement Instrument (SIM) and Quick Neurological Screening Test II (QNST) were used to determine children's motor proficiency as well as possible neuro-motor problems. The Tennessee Self-Concept Scale (Child Form) (TSCS-CF) and Child Anxiety Scale (CAS) were used to determine the children's self-concept and anxiety respectively. One way variance of analysis, repeated measures analysis, independent t-testing, co-variance of analysis as well as correlational coefficients (r) were conducted, using the Statistica computer package in order to analyze the data according to the above-mentioned aims. A p-value of smaller than or equal to 0.05 was accepted as a significant difference. From the results of the study it seemed that the self-concept and anxiety of randomly selected 7-9 year old children (N=58) diagnosed with DCD are negatively influenced and that girls are more vulnerable to these influences. Repeated measure analyses over a period of one year showed that of the three programmes the motor intervention programme showed the best results at improving the children's motor proficiency while, on the other hand, the psychomotor intervention programme improved their self-concept most. Ethnic group and gender did not have a significant effect on the success of intervention programmes. Lastly, it was found that underlying neuro-motor problems could influence the effect of an intervention programme negatively. It is clear from this study that DCD has a negative effect on children, but that participation in a well planned intervention programme will have positive effects on both their motor proficiency and self-concept. / Thesis (Ph.D. (Human Movement Science))--North-West University, Potchefstroom Campus, 2006.

This thesis aims to design an appropriate human-following solution for a mobile robot. The research can be characterised as interactive model building for a Human Robot Interaction (HRI) scenario. It studies possible proposals for the robot system that learns to accomplish the task autonomously, based on the human preference about the positions and movements of the robot during the interaction. A multilayered feedforward network framework with backpropagation is the adopted learning strategy. The research breaks the task of following a human into three independent behaviours: social positioning, human avoidance and obstacle avoidance. Social positioning is the behaviour that moves the robot, via reasonable paths, to the most appropriate location to follow the human. Both the location and the paths reflect the preference of the human, which varies by individual. The main body of the research therefore proposes a using-while-learning system for this behaviour such that the robot can adapt to the human’s preference autonomously. This research investigated multilayered feedforward networks with backpropagation learning to fulfil the social learning task. This learning model is less used in HRI because a complete set of correct training data doesn’t exist as the human preference is initially unknown. The research proposes a novel method to generate the training data during the operation of learning and introduces the concept of adaptive and reactive learning. A novel training scheme that combines the two learning threads has been proposed, in which the learning is fast, robust and able to adapt to new features of the human preference online. The system enables the behaviour to be a real using-while-learning system as no pre-training of any form is needed to ensure the successful performance of the behaviour. Extensive simulations and interactive experiments with humans have also been conducted to prove the robustness of the system.

Medulloblastoma is the most common brain tumor in children, accounting for 10-20% of primary central nervous system (CNS) neoplasms and approximately 40% of all posterior fossa tumors. It is a highly invasive embryonal neuroepithelial tumor that typically arises in the cerebellar vermis and has a tendency to disseminate throughout the CNS early in its course. The molecular mechanisms of the disease largely remain uncharacterized, as the clinical treatment is still associated with mortality and severe side effects. The development of a clinically relevant in vivo model is important not only to further understand the disease but also to provide a method with which to test novel therapeutics. This study quantified the volumetric growth of a human medulloblastoma (VC312) in the athymic nude mouse cerebellum using Gd- enhanced T1-weighed MRI scans. Additionally, a medulloblastoma flank tumor model was used to explore the in vivo effect of the oral anti-cancer agent that inhibits Akt activation in the phosphoinositide 3-kinase (PI3K) pathway. In the orthotopic intracerebellar tumor model, perifosine significantly increased the survival of treated mice while qualitatively reducing leptomeningeal dissemination. In the flank model, perifosine effectively suppressed the volumetric growth, decreased activation of the AKT pathway and reduced cellular proliferation in treated mice.

The quantitative study of conflict management is concerned with finding models which are accurate and also capable of providing a causal interpretation of results. This dissertation applies computational intelligence methods to study interstate disputes. Both multilayer perceptron neural networks and Takagi-Sugeno neuro-fuzzy models are used to model interstate interactions. The multilayer perceptron neural network is trained in the Bayesian framework, using the Hybrid Monte Carlo method to sample from the posterior probabilities. It is found that the network is able to forecast conflict with an accuracy of 77.3%. A hybrid machine learning method using the neural network and the genetic algorithm is then presented as a method of suggesting how conflict can be brought under control. The automatic relevance determination approach and the sensitivity analysis are used as methods of extracting causal information from the neural network. The Takagi-Sugeno neuro-fuzzy model is optimised, using the Gustafson-Kessel clustering algorithm to partion the input space. It is found that the neuro-fuzzy model predicts conflict with an accuracy of 80.1%. The neuro-fuzzy model is also incorporated into the hybrid machine learning method to suggest how the identified conflict cases can be avoided. The casual interpretation is then formulated by a linguistic approximation of the fuzzy rules extracted from the neuro-fuzzy model. The major finding in this work is that the interpretations drawn from both the neural network and the neuro-fuzzy model are consistent.