Práce popisuje základy principu funkčnosti neuronů a vytvoření umělých neuronových sítí. Je zde důkladně popsána struktura a funkce neuronů a ukázán nejpoužívanější algoritmus pro učení neuronů. Základy fuzzy logiky, včetně jejich výhod a nevýhod, jsou rovněž prezentovány. Detailněji je popsán algoritmus zpětného šíření chyb a adaptivní neuro-fuzzy inferenční systém. Tyto techniky poskytují efektivní způsoby učení neuronových sítí.

Artificial intelligence and fuzzy logic related to it currently belong to very popular and rapidly expanding technological subjects. It finds use in many areas, which also include the process of prediction of future states based on specific finite input characteristics. This master’s thesis deals with predictions that are done in field of agricultural crops growing. Basic principles that are affecting mentioned agricultural growing are explained here, their meaning and significance are specified, these are later on perceived as a key aspect to creation of fuzzy models that are used for prediction. This process is specifically about finding out the most suitable crop on considered parcel for maximization of income. Second part of design section is dedicated to description of approaches for work with fuzzy models and is also used as demonstration of application created for purpose of this thesis.

Práce popisuje základy principu funkčnosti neuronů a vytvoření umělých neuronových sítí. Je zde důkladně popsána struktura a funkce neuronů a ukázán nejpoužívanější algoritmus pro učení neuronů. Základy fuzzy logiky, včetně jejich výhod a nevýhod, jsou rovněž prezentovány. Detailněji je popsán algoritmus zpětného šíření chyb a adaptivní neuro-fuzzy inferenční systém. Tyto techniky poskytují efektivní způsoby učení neuronových sítí.

Le médulloblastome est une des tumeurs les plus fréquentes du système nerveux central chez l’enfant. Son impact clinique, ainsi que les effets secondaires engendrés par les traitements actuels, sont significatifs en matière de morbidité et de mortalité. La caractérisation moléculaire des tumeurs du système nerveux central a grandement évolué, et ce, particulièrement en ce qui concerne le médulloblastome. Des travaux antérieurs ont permis d’établir qu’un des sous-groupes de médulloblastome est caractérisé par l’activation de la voie sonic hedgehog. La mutation la plus fréquente menant à ce sous-type de médulloblastome est la mutation du gène suppresseur de tumeur PTCH1. Grâce au modèle de souris Ptch1+/-, des données issues de notre laboratoire ont permis de caractériser le développement de cette tumeur comme étant en deux étapes. Ce travail porte sur la caractérisation du mécanisme par lequel cette première étape, soit la perte d’hétérozygotie de Ptch1, survient. Tout d’abord, nous revisitons le rôle in vivo du corécepteur Boc dans la tumorigenèse. Selon nos résultats, la modulation de Boc ne semble pas avoir un impact significatif sur le développement tumoral dans des expériences de transplantation orthotopiques. Ensuite, nous démontrons que le ligand Shh augmente le dommage à l’ADN, ce qui mène à une hausse des évènements de recombinaisons qui peuvent causer une perte d’hétérozygotie. Nous tentons de moduler l’activité de Rad51 en observant une tendance non statistiquement significative des évènements de recombinaison avec des inhibiteurs de Rad51. Nous démontrons ensuite qu’un inhibiteur de Cdc7 permet la diminution des évènements de recombinaisons ainsi qu’une diminution du stress réplicatif de l’ADN. En intervenant sur le gène Mcm2 grâce à un modèle de souris transgénique, nous parvenons à prouver qu’une diminution de l’action de Mcm2 permet une diminution du stress réplicatif de l’ADN. En somme, la première étape du développement du médulloblastome sonic hedgehog-activé est la perte d’hétérozygotie de Ptch1. Celle-ci est caractérisée par une augmentation du dommage à l’ADN engendrant une hausse des évènements de recombinaison. Plusieurs cibles potentielles de modulation s’avèrent prometteuses pour un éventuel traitement ciblé. / Medulloblastoma is one of the most common central nervous system tumors of the child. Its clinical impact, as well as the adverse effects caused by current treatments, are significant in terms of morbidity and mortality. The molecular characterization of tumors of the central nervous system has greatly evolved, particularly in the case of medulloblastoma. Previous work has established that one of the medulloblastoma sub-groups is characterized by the activation of the sonic hedgehog (Shh) pathway. The most common mutation leading to this medulloblastoma subtype is the PTCH1 tumor suppressor gene mutation. Working with the Ptch1+/- mouse model, data from our la-boratory characterized the medulloblastoma tumorigenesis as a two-step process. This work focuses on the characterization of the mechanism by which this first step, the loss of heterozygosity of Ptch1, occurs. First, we revisit the in vivo role of the Boc coreceptor in the medulloblastoma tumor-igenesis. According to our results, Boc modulation does not seem to have a significant impact on tumor development. Next, we show that the Shh ligand increases DNA dam-age. This leads to an increase in recombination events which predispose to loss of het-erozygosity. We attempt to modulate Rad51 activity and observe a non-statistically sig-nificant trend to decrease recombination events with Rad51 inhibitors. We then demonstrate that Cdc7 inhibition reduces recombination events as well as DNA replica-tive stress. Using an Mcm2 transgenic mouse model, we demonstrate that a reduction in the action of Mcm2 reduces DNA replicative stress. To conclude, the first step in the development of Shh-activated medulloblastoma is the loss of heterozygosity of Ptch1. This is characterized by an increase in DNA damage leading to an increase in recombination events. Several potential modulation targets hold promise for possible targeted therapy.

Background: Data on the safety and efficacy of flow diverters (FD) for the treatment of unruptured internal carotid artery (ICA) aneurysms with compressive neuro-ophthalmological symptoms (NOS) are scarce and comprise mainly small case series. Methods: We performed a search of three databases and included series with ≥10 patients, with unruptured aneurysms of the ICA and NOS, treated with FD. Random-effects analysis of treatment results and safety was performed. Results: A total of 22 studies reporting on 594 patients were included. Pooled proportions of NOS recovery, improvement, transient and permanent worsening were: 47.4% (95% CI 35.0% to 60.1%); 74.5% (95% CI 67.9% to 80.2%); 7.1% (95% CI 3.3% to 14.7%); and 4.9% (95% CI 3.2% to 7.4%), respectively. Rates of complete recovery and improvement in patients with isolated visual symptoms were 30.6% (95% CI 12.5% to 57.7%) and 56.6% (95% CI 42.3% to 69.9%). Isolated oculomotor symptoms recovered completely in 47.8% (95% CI 29.9% to 66.3%) and improved in 78% (95% CI 69.2% to 84.9%). Morbidity occurred in 5% (95% CI 2.8% to 9%) and mortality in 3.9% (95% CI 2% to 7.5%) of patients. An increased likelihood of symptom improvement was observed when treatment was performed early (<1 month) after symptom onset (OR=11.22, 95% CI 3.9% to 32.5%). Conclusion: Flow diversion promotes recovery or improvement of compressive symptoms in a large proportion of patients but is associated with significant rates of morbidity and mortality. Transient and permanent NOS worsening is not uncommon. Early treatment is of utmost importance, as it increases the likelihood of symptom improvement more than 10-fold.

Les commotions cérébrales ont longtemps été considérées comme une blessure ne comportant que peu ou pas de conséquences. Cependant, la mise à la retraite forcée de plusieurs athlètes de haut niveau, liée au fait d'avoir subi des commotions cérébrales multiples, a porté cette question au premier plan de la culture scientifique et sportive. Malgré la sensibilisation croissante du public et la compréhension scientifique accrue des commotions cérébrales, il reste encore beaucoup d’inconnus au sujet de ces blessures. En effet, il est difficile de comprendre comment cette atteinte peut avoir des effets si profonds malgré le fait qu’elle n’entraîne apparemment pas de conséquences physiques apparentes lorsque les techniques traditionnelles d’imagerie cérébrale sont utilisées. Les techniques de neuroimagerie fonctionnelle ont cependant contribué à répondre aux nombreuses questions entourant les conséquences des commotions cérébrales ainsi qu'à accroître la compréhension générale de la physiopathologie de commotions cérébrales. Bien que les techniques de base telles que l'imagerie structurelle comme les scans TC et IRM soient incapables de détecter des changements structurels dans la grande majorité des cas (Ellemberg, Henry, Macciocchi, Guskiewicz, & Broglio, 2009; Johnston, Ptito, Chankowsky, & Chen, 2001), d'autres techniques plus précises et plus sensibles ont été en mesure de détecter avec succès des changements dans le cerveau commotionné. Des études d’IRM fonctionelle ont entre autres établi une solide relation entre les altérations fonctionnelles et les symptômes post-commotionels (Chen, Johnston, Collie, McCrory, & Ptito, 2007; Chen et al., 2004; Chen, Johnston, Petrides, & Ptito, 2008; Fazio, Lovell, Pardini, & Collins, 2007). Les mesures électrophysiologiques telles que les potentiels évoqués cognitifs (ERP) (Gaetz, Goodman, & Weinberg, 2000; Gaetz & Weinberg, 2000; Theriault, De Beaumont, Gosselin, Filipinni, & Lassonde, 2009; Theriault, De Beaumont, Tremblay, Lassonde, & Jolicoeur, 2010) et la stimulation magnétique transcrânienne ou SMT (De Beaumont, Brisson, Lassonde, & Jolicoeur, 2007; De Beaumont, Lassonde, Leclerc, & Theoret, 2007; De Beaumont et al., 2009) ont systématiquement démontré des altérations fonctionnelles chez les athlètes commotionnés. Cependant, très peu de recherches ont tenté d'explorer davantage certaines conséquences spécifiques des commotions cérébrales, entre autres sur les plans structural et métabolique. La première étude de cette thèse a évalué les changements structurels chez les athlètes commotionnés à l’aide de l'imagerie en tenseur de diffusion (DTI) qui mesure la diffusion de l'eau dans la matière blanche, permettant ainsi de visualiser des altérations des fibres nerveuses. Nous avons comparé les athlètes commotionnés à des athlètes de contrôle non-commotionnés quelques jours après la commotion et de nouveau six mois plus tard. Nos résultats indiquent un patron constant de diffusion accrue le long des voies cortico-spinales et dans la partie du corps calleux reliant les régions motrices. De plus, ces changements étaient encore présents six mois après la commotion, ce qui suggère que les effets de la commotion cérébrale persistent bien après la phase aiguë. Les deuxième et troisième études ont employé la spectroscopie par résonance magnétique afin d'étudier les changements neurométaboliques qui se produisent dans le cerveau commotionné. La première de ces études a évalué les changements neurométaboliques, les aspects neuropsychologiques, et la symptomatologie dans la phase aiguë post-commotion. Bien que les tests neuropsychologiques aient été incapables de démontrer des différences entre les athlètes commotionnés et non-commotionnés, des altérations neurométaboliques ont été notées dans le cortex préfrontal dorsolatéral ainsi que dans le cortex moteur primaire, lesquelles se sont avérées corréler avec les symptômes rapportés. La deuxième de ces études a comparé les changements neurométaboliques immédiatement après une commotion cérébrale et de nouveau six mois après l’atteinte. Les résultats ont démontré des altérations dans le cortex préfrontal dorsolatéral et moteur primaire dans la phase aiguë post-traumatique, mais seules les altérations du cortex moteur primaire ont persisté six mois après la commotion. Ces résultats indiquent que les commotions cérébrales peuvent affecter les propriétés physiques du cerveau, spécialement au niveau moteur. Il importe donc de mener davantage de recherches afin de mieux caractériser les effets moteurs des commotions cérébrales sur le plan fonctionnel. / Concussions had long been considered an injury of little to no consequence. However, the forced retirement of several high profile athletes due to the impact of having suffered multiple concussions has pushed the issue to the forefront of scientific and sports culture alike. Despite the growing public awareness and the ever-expanding scientific understanding of concussions there is still much that remains unknown about these injuries. Indeed, understanding how an injury can have such profound effects, though mostly transient, without any apparent physical consequence continues to confound how concussions are conceptualized in research. Neuroimaging techniques have helped answer many of the questions surrounding the physical consequences of concussions on the brain as well as increasing the general understanding of the pathophysiology of concussions. While basic structural imaging techniques such as CT scans and MRI are unable to detect any structural changes in the vast majority of cases (Ellemberg, et al., 2009; Johnston, et al., 2001), other more precise and sensitive techniques have been able to successfully detect changes in the concussed brain. Functional MRI studies have further established a strong relationship between functional alterations and post-concussion symptoms (Chen, et al., 2007; Chen, et al., 2004; Chen, et al., 2008; Fazio, et al., 2007). Electrophysiological measures such as ERP (Gaetz, et al., 2000; Gaetz & Weinberg, 2000; Theriault, et al., 2009; Theriault, et al., 2010) and TMS (De Beaumont, Brisson, et al., 2007; De Beaumont, Lassonde, et al., 2007; De Beaumont, et al., 2009) have consistently demonstrated alterations in concussed athletes. However, there has been very little research that has attempted to further explore the specific structural and metabolic aspects of concussion. The first study assessed structural changes in concussed athletes using diffusion tensor imaging which measures water diffusion in white matter. We compared concussed athletes with non-concussed control athletes in the days immediately after injury and again six months later. Our results indicated a consistent pattern of increased diffusion along neural tracts of the cortical spinal tract and in the corpus callosum underlying motor cortex. Furthermore, these changes were still present six months after injury suggesting that the effects of concussion are persistent past the acute phase. The second and third studies employed magnetic resonance spectroscopy as a means of investigating the neurometabolic changes that occur in the concussed brain. The first of these studies investigated the neurometabolic changes, neuropsychological aspects, and symptomatology in the acute post-injury phase. While neuropsychological testing was unable to show differences between concussed and non-concussed athletes, neurometabolic alterations were noted in the dorsal lateral prefrontal cortex as well as in primary motor cortex which correlated with reported symptoms. The second study investigated neurometabolic changes immediately after concussion and again six months after injury. Results indicated alterations in the dorsolateral prefrontal and primary motor cortices in the acute post-injury phase, but only those in primary motor cortex persisted to the six month time point.

There are approximately 6000 to 8000 rare diseases, each with a prevalence of less than 1 / 10 000, but in aggregate affecting 6 to 8% of the population. It is important to evaluate disease development and progression to know the natural course of any disease. This information can be utilized in diagnostics and in assessing effects of therapeutic interventions as they become available. This thesis describes the natural clinical history and evolution of imaging findings of two rare diseases over approximately two decades. Papers I, II and III present clinical, magnetic resonance imaging (MRI), magnetic resonance spectroscopy (MRS) and 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) findings in LMNB1-related autosomal dominant leukodystrophy (ADLD). MRI was found to be very sensitive in finding pathology in patients with LMNB1-related ADLD, even before the onset of clinical symptoms. However, even patients with widespread MRI changes can have a relatively mild symptomatology and present only slight disturbances in metabolic examinations such as MRS and FDG-PET. This is compatible with relatively intact axons, even as myelin impairment is widespread. Paper IV presents clinical and MRI findings in the brain and musculature in SIL1-positive Marinesco-Sjögren syndrome (MSS), and describes a new, mild phenotype of the disease with no intellectual disabilities and only slight motor disabilities. With a 19-year-long radiological follow-up, a slow progressive atrophic process in the cerebellum and brainstem could be demonstrated. MRI of the musculature shows early involvement of the quadriceps and gastrocnemii but not the tibialis anterior, progressing to widespread atrophy in the back and upper and lower limbs at the age of 20 years. In the mildest phenotype, the most severely affected muscles were the m gluteus maximus, m sartorius, m peroneus longus, and the lateral head of the m gastrocnemius.

In the literature, various Mobile Ad hoc NETwork (MANET) routing protocols proposed. Each performs the best under specific context conditions, for example under high mobility or less volatile topologies. In existing MANET, the degradation in the routing protocol performance is always associated with changes in the network context. To date, no MANET routing protocol is able to produce optimal performance under all possible conditions. The core aim of this thesis is to solve the routing problem in mobile Ad hoc networks by introducing an optimum system that is in charge of the selection of the running routing protocol at all times, the system proposed in this thesis aims to address the degradation mentioned above. This optimisation system is a novel approach that can cope with the network performance’s degradation problem by switching to other routing protocol. The optimisation system proposed for MANET in this thesis adaptively selects the best routing protocol using an Artificial Intelligence mechanism according to the network context. In this thesis, MANET modelling helps in understanding the network performance through different contexts, as well as the models’ support to the optimisation system. Therefore, one of the main contributions of this thesis is the utilisation and comparison of various modelling techniques to create representative MANET performance models. Moreover, the proposed system uses an optimisation method to select the optimal communication routing protocol for the network context. Therefore, to build the proposed system, different optimisation techniques were utilised and compared to identify the best optimisation technique for the MANET intelligent system, which is also an important contribution of this thesis. The parameters selected to describe the network context were the network size and average mobility. The proposed system then functions by varying the routing mechanism with the time to keep the network performance at the best level. The selected protocol has been shown to produce a combination of: higher throughput, lower delay, fewer retransmission attempts, less data drop, and lower load, and was thus chosen on this basis. Validation test results indicate that the identified protocol can achieve both a better network performance quality than other routing protocols and a minimum cost function of 4.4%. The Ad hoc On Demand Distance Vector (AODV) protocol comes in second with a cost minimisation function of 27.5%, and the Optimised Link State Routing (OLSR) algorithm comes in third with a cost minimisation function of 29.8%. Finally, The Dynamic Source Routing (DSR) algorithm comes in last with a cost minimisation function of 38.3%.

Nós investigamos o efeito sequencial do treinamento aerório (T) nas vias ocitocinérgicas (OTérgicas) em áreas centrais autonômicas, utilizando o fator neurotrófico derivado do cérebro (BDNF) como um indicador de plasticidade neuronal. Ratos normotensos (WKY) e espontaneamente hipertensos (SHR) foram submetidos a T ou sedentarismo (S) por 8 semanas. T causou bradicardia de repouso (SHR=T2-T8; WKY=T8) e redução da PA (SHR=T8). T promoveu aumento precoce (T1) da expressão de BDNF, sendo em WKY aumento transiente e SHR mantido. A irBDNF foi similarmente aumentado em WKY e SHR de T1-T2 até T8. Tais alterações foram acompanhadas por aumento no conteúdo de OT nas áreas biossintéticas e de OTR nas áreas de projeção. Os resultados indicam que a plasticidade induzida pelo T no sistema OTérgico central ocorre precocemente no PVN comparado com outras áreas autonômicas, é mais rápido em WKY e precede as mudanças da FC. Os dados também mostram maior \"turnover\" do RNAm de BDNF no grupo SHR. / We investigate time-course training (T) effects on oxitocinergic (OTergic) pathways within autonomic areas, using brain-derived neurotrophic factor (BDNF) as a index of neuronal plasticity. Normotensive (WKY) and spontaneously hypertensive rats (SHR) were submitted to low-intensity T or sedentarism (S) for 8 weeks. T caused resting bradycardia (SHR=T2-T8; WKY=T8) and reduced AP (SHR=T8). T induced precocious increase (T1) of BDNF expression, with transient increase in WKY and sustained in SHR. BDNFir was similarly increased in WKY and SHR (T1-T2 until T8). These changes were accompanied by increased content of OT in biosynthetic areas and OTR in projection areas. Results indicate that T-induced plasticity of central OTergic system occurs precociously in the PVN compared to other autonomic areas, is faster in WKY and precedes HR changes. Data also showed higher BDNF mRNA turnover in the SHR group.

Este estudo apresenta o trabalho do psicanalista e hipnoterapeuta americano Milton Hyland Erickson a partir de seus dados biográficos e de sua relevância para a chamada terapia estratégica, propondo, em consonância com sua experiência profissional, uma nova epistemologia para a mudança; propõe ainda uma comparação entre a trajetória de Freud e a de Erickson em relação à hipnose, bem como um apanhado histórico sobre essa técnica. Com base nessa recuperação teórica, os padrões ericksonianos de linguagem são investigados à luz do Metamodelo e do Modelo Milton, criações de Richard Bandler e John Grinder, tomando por base alguns conceitos da Programação Neurolinguística (PNL) tais como sistemas representacionais, filtros, modelagem, espelhamento e rapport. Empreende-se uma análise do discurso ericksoniano a partir de três casos selecionados dentre seus atendimentos clínicos, evidenciando os padrões de linguagem apresentados nas categorias e subcategorias do Metamodelo e do Modelo Milton, com o objetivo de validá-los tanto teórica quanto empiricamente / This study aimed at scrutinizing Milton Hyland Ericksons theoretical framework and therapeutic methodology with a view to the understanding of their relevance to the so-called strategic therapy. That aim was carried out through analyses of issues such as Ericksons professional experience, his shared points and differences with Freudian hypnosis and particularly his patterns of discourse as well as his very clinical technique. Those analyses were accomplished under the guidance of the Metamodel and the Milton Model as proposed by Richard Bandler and John Grinder within the references and concepts comprised in the Neuro-Linguistic Programming such as representational systems, filters, modeling, mirroring and rapport. In order to ground and illustrate the theoretical analyses, this work was enriched by the scrutiny of three Ericksonian clinical cases. This strategy proved to be effective since it has provided evidences about both Ericksons language patterns and empirical data for the observation of the categories and subcategories of the Metamodel and Miltons Model, as a kind of a quest for validation of his theory and methods. The results put into light Ericksons understanding of the therapeutic work