Ainda pouco se sabe sobre as causas do transtorno do espectro autista (TEA) e seus efeitos na funcionalidade cerebral, porém, diversas pesquisas apontam que a condição esteja relacionada à uma conectividade diferenciada entre regiões do cérebro. A conectividade córtico-cerebelar tem sido tema de pesquisas nas últimas décadas em decorrência de novos achados que indicam que esta conectividade está relacionada ao aprendizado e refinamento de diversas funcionalidades do córtex. Acredita-se que uma falha na conectividade córtico-cerebelar poderia estar relacionada à falhas em funções sensorimotoras, cognitivas e emocionais. A investigação de regiões cuja conectividade córtico-cerebelar está alterada no TEA contribui para uma melhor compreensão deste transtorno. Assim, o objetivo deste trabalho é identificar regiões do cérebro cuja conectividade funcional com o cerebelo seja diferente entre indivíduos com desenvolvimento típico (DT) e diagnosticados com TEA. Para isto, utilizamos imagens de ressonância magnética funcional (fMRI) de 708 indivíduos em estado de repouso (432 DT e 276 TEA) com idades entre 6 e 58 anos coletados pelo consórcio ABIDE. Os dados foram pré-processados e divididos conforme regiões anatômicas do cérebro que foram adotadas como regiões de interesse (ROIs). Para determinar a conectividade funcional de cada região do córtex com o cerebelo, aplicamos o método de análise de componentes principais (PCA) nas ROIs do cerebelo e utilizamos um modelo regressão linear para cada ROI do córtex, sendo a série temporal da ROI do córtex a variável resposta e as componentes principais (PCs) do cerebelo as variáveis preditoras. Em seguida, identificamos as regiões com conectividade funcional diferente entre indivíduos com DT e diagnosticados com TEA através de um modelo linear que inclui como covariáveis, idade, gênero e local de coleta do dado. Identificamos cinco regiões do córtex que apresentam reduzida conectividade funcional com o cerebelo nos indivíduos com TEA, sendo elas: (i) giro fusiforme direito, (ii) giro pós-central direito, (iii) giro temporal superior direito e (iv) giro temporal médio direito e (v) esquerdo. Todas as cinco regiões são parte do sistema sensorimotor, e estão relacionadas à funções ligadas à sintomas característicos do quadro de TEA, como: sensibilidade à estímulos sensoriais, dislexia, prosopagnosia (dificuldade para reconhecer faces), dificuldade de compreensão de linguagem e dificuldade de reconhecimento de emoções em faces. Nossos resultados mostram que existem regiões do sistema sensorimotor que apresentam conectividade funcional com o cerebelo atipicamente reduzida em TEA, como corroborado por estudos de imageamento com tarefa específica e como hipotetizado por estudos de conectividade estrutural. Nós acreditamos que a conectividade córtico-cerebelar reduzida dessas regiões esteja prejudicando o processamento e aprendizado de funções sensorimotoras, levando ao surgimento de sintomas típicos do TEA. / Little is known about the causes of autism spectrum disorder (ASD) and its effects on brain functions. Several researches point that it may be related to differentiated connections between brain regions. The cortico-cerebellar connectivity has been theme of researches over the last decade given the new discoveries suggesting that this connection is associated with learning and tuning of diverse brain functionalities. It is believed that cortico-cerebellar connectivity impairment may be related to impairments on sensorimotor, cognitive, and emotional functions. For a better understanding of the condition, we aim at identifying brain regions that present a impaired cortico-cerebellar connectivity between TD and ASD. Thus, our goal is to identify brain regions where the functional connectivity with the cerebellum is different between subjects with typical development (TD) and ASD. We used functional magnetic resonance images (fMRI) of 708 subjects under resting state protocol (432 TD and 276 ASD) with ages between 6 and 58 years old collected by the ABIDE Consortium. Data was pre-processed, splited in anatomical brain regions, which were adopted as regions of interest (ROIs). To establish the functional connectivity of each cortical ROI to the cerebellum, first we applied the principal component analysis (PCA) on cerebellar ROIs. Next, we used a linear regression model to each cortical ROI time series as response variable and the cerebellum principal components (PCs) as the predictive variables. After that, we identified regions of different connections between TD and ASD subjects applying a linear model including age, gender, and data collection site as covariables. We identified five cortical regions with reduced functional connectivity with cerebellum on the ASD subjects, namely: (i) right fusiform gyrus, (ii) right postcentral gyrus, (iii) right superior temporal gyrus, (iv) right middle temporal gyrus, and (v) left middle temporal gyrus. All five regions are part of the sensorimotor system, and contribute to functions typically associated with ASD, such as: sensitivity to external stimulus, dyslexia, prosopagnosia (difficulty to recognize faces), language comprehension impairments, and recognition of emotional expressions impairment. Our results show that there are brain regions with atypical cortico-cerebellar connectivity impairment on ASD, in accordance with results from previous studies of tactography and task driven fMRI. We believe that the decreased cortico-cerebellar connectivity of these regions are affecting the learning process of sensorimotor functionalities, leading to typical ASD symptoms.

Ainda pouco se sabe sobre as causas do transtorno do espectro autista (TEA) e seus efeitos na funcionalidade cerebral, porém, diversas pesquisas apontam que a condição esteja relacionada à uma conectividade diferenciada entre regiões do cérebro. A conectividade córtico-cerebelar tem sido tema de pesquisas nas últimas décadas em decorrência de novos achados que indicam que esta conectividade está relacionada ao aprendizado e refinamento de diversas funcionalidades do córtex. Acredita-se que uma falha na conectividade córtico-cerebelar poderia estar relacionada à falhas em funções sensorimotoras, cognitivas e emocionais. A investigação de regiões cuja conectividade córtico-cerebelar está alterada no TEA contribui para uma melhor compreensão deste transtorno. Assim, o objetivo deste trabalho é identificar regiões do cérebro cuja conectividade funcional com o cerebelo seja diferente entre indivíduos com desenvolvimento típico (DT) e diagnosticados com TEA. Para isto, utilizamos imagens de ressonância magnética funcional (fMRI) de 708 indivíduos em estado de repouso (432 DT e 276 TEA) com idades entre 6 e 58 anos coletados pelo consórcio ABIDE. Os dados foram pré-processados e divididos conforme regiões anatômicas do cérebro que foram adotadas como regiões de interesse (ROIs). Para determinar a conectividade funcional de cada região do córtex com o cerebelo, aplicamos o método de análise de componentes principais (PCA) nas ROIs do cerebelo e utilizamos um modelo regressão linear para cada ROI do córtex, sendo a série temporal da ROI do córtex a variável resposta e as componentes principais (PCs) do cerebelo as variáveis preditoras. Em seguida, identificamos as regiões com conectividade funcional diferente entre indivíduos com DT e diagnosticados com TEA através de um modelo linear que inclui como covariáveis, idade, gênero e local de coleta do dado. Identificamos cinco regiões do córtex que apresentam reduzida conectividade funcional com o cerebelo nos indivíduos com TEA, sendo elas: (i) giro fusiforme direito, (ii) giro pós-central direito, (iii) giro temporal superior direito e (iv) giro temporal médio direito e (v) esquerdo. Todas as cinco regiões são parte do sistema sensorimotor, e estão relacionadas à funções ligadas à sintomas característicos do quadro de TEA, como: sensibilidade à estímulos sensoriais, dislexia, prosopagnosia (dificuldade para reconhecer faces), dificuldade de compreensão de linguagem e dificuldade de reconhecimento de emoções em faces. Nossos resultados mostram que existem regiões do sistema sensorimotor que apresentam conectividade funcional com o cerebelo atipicamente reduzida em TEA, como corroborado por estudos de imageamento com tarefa específica e como hipotetizado por estudos de conectividade estrutural. Nós acreditamos que a conectividade córtico-cerebelar reduzida dessas regiões esteja prejudicando o processamento e aprendizado de funções sensorimotoras, levando ao surgimento de sintomas típicos do TEA. / Little is known about the causes of autism spectrum disorder (ASD) and its effects on brain functions. Several researches point that it may be related to differentiated connections between brain regions. The cortico-cerebellar connectivity has been theme of researches over the last decade given the new discoveries suggesting that this connection is associated with learning and tuning of diverse brain functionalities. It is believed that cortico-cerebellar connectivity impairment may be related to impairments on sensorimotor, cognitive, and emotional functions. For a better understanding of the condition, we aim at identifying brain regions that present a impaired cortico-cerebellar connectivity between TD and ASD. Thus, our goal is to identify brain regions where the functional connectivity with the cerebellum is different between subjects with typical development (TD) and ASD. We used functional magnetic resonance images (fMRI) of 708 subjects under resting state protocol (432 TD and 276 ASD) with ages between 6 and 58 years old collected by the ABIDE Consortium. Data was pre-processed, splited in anatomical brain regions, which were adopted as regions of interest (ROIs). To establish the functional connectivity of each cortical ROI to the cerebellum, first we applied the principal component analysis (PCA) on cerebellar ROIs. Next, we used a linear regression model to each cortical ROI time series as response variable and the cerebellum principal components (PCs) as the predictive variables. After that, we identified regions of different connections between TD and ASD subjects applying a linear model including age, gender, and data collection site as covariables. We identified five cortical regions with reduced functional connectivity with cerebellum on the ASD subjects, namely: (i) right fusiform gyrus, (ii) right postcentral gyrus, (iii) right superior temporal gyrus, (iv) right middle temporal gyrus, and (v) left middle temporal gyrus. All five regions are part of the sensorimotor system, and contribute to functions typically associated with ASD, such as: sensitivity to external stimulus, dyslexia, prosopagnosia (difficulty to recognize faces), language comprehension impairments, and recognition of emotional expressions impairment. Our results show that there are brain regions with atypical cortico-cerebellar connectivity impairment on ASD, in accordance with results from previous studies of tactography and task driven fMRI. We believe that the decreased cortico-cerebellar connectivity of these regions are affecting the learning process of sensorimotor functionalities, leading to typical ASD symptoms.

La maladie de Parkinson (MP) est la deuxième maladie neurodégénérative la plus commune. Les symptômes principalement observés chez les patients atteints de la MP sont la rigidité, les tremblements, la bradykinésie et une instabilité posturale. Leur sévérité est souvent asymétrique. La cause principale de ces symptômes moteurs est la dégénérescence du circuit dopaminergique nigro-striatal qui mène à un débalancement d’activité du circuit cortico-striatal. Ce débalancement de circuits est le point essentiel de cette thèse. Dans les protocoles de recherche décrits ici, des patients atteints de la MP (avant et après une dose de levodopa) et des participants contrôles sains ont effectué des mouvements auto-initiés ou en réponse à des stimulis externes pendant que l’on mesurait leur activité cérébrale en imagerie par résonance magnétique fonctionnelle (IRMf). Dans cette thèse, nous abordons et mettons en évidence quatre (4) points principaux. En première partie (chapitre 2), nous présentons un recensement de la littérature sur les cicruits cortico-striataux et cortico-cérébelleux dans la MP. En utilisant des méthodes de neuroimagerie, des changements d’activité cérébrale et cérébelleuse ont été observés chez les patients atteints de la MP comparés aux participants sains. Même si les augmentations d’activité du cervelet ont souvent été attribuées à des mécanismes compensatoires, nos résultats suggèrent qu’elles sont plus probablement liées aux changements pathophysiologiques de la MP et à la perturbation du circuit cortico-cérébelleux. En général, nous suggérons (1) que le circuit cortico-cérébelleux est perturbé chez les patients atteints de la MP, et que les changements d’activité du cervelet sont liés à la pathophysiologie de la MP plutôt qu’à des mécanismes compensatoires. En deuxième partie (chapitre 3), nous discutons des effets de la levodopa sur les hausses et baisses d’activité observés chez les patients atteints de la MP, ainsi que sur l’activité du putamen pendant les mouvements d’origine interne et externe. De nombreuses études en neuroimagerie ont montré une baisse d’activité (hypo-activité) préfrontale liée à la déplétion de dopamine. En revanche, l’utilisation de tâches cognitives a montré des augmentations d’activité (hyper-activité) corticale chez les patients atteints de la MP comparés aux participants sains. Nous avons suggéré précédemment que ces hypo- et hyper-activités des régions préfrontales dépendent de l’implication du striatum. Dans cette thèse nous suggérons de plus (2) que la levodopa ne rétablit pas ces hyper-activations, mais plutôt qu’elles sont liées à la perturbation du circuit méso-cortical, et aussi possiblement associées à l’administration de médication dopaminergique à long terme. Nous montrons aussi (3) que la levodopa a un effet non-spécifique à la tâche sur l’activité du circuit cortico-striatal moteur, et qu’elle n’a pas d’effet sur l’activité du circuit cortico-striatal cognitif. Nous montrons enfin (chapitre 4) que la levodopa a un effet asymétrique sur les mouvements de la main droite et gauche. À peu près 50% des patients atteints de la MP démontrent une asymétrie des symptômes moteurs, et ceci persiste à travers la durée de la maladie. Nos résultats suggèrent (4) que la levodopa pourrait avoir un plus grand effet sur les patrons d’activations des mouvements de la main la plus affectée. / Parkinson’s disease (PD) is the second most common neurodegenerative disease, mainly manifested by tremor, rigidity, bradykinesia and postural instability, and often an asymmetry of symptom severity of the left and right sides of the body. The depletion of dopamine of the nigrostriatal pathway is the primary cause of the motor symptoms observed in patients with PD, leading to an imbalance in basal-ganglia prefrontal circuits. In the protocols described here, patients with PD before and after levodopa administration and healthy participants performed self-initiated (SI) and externally triggered (ET) movements with the left and right hand during functional magnetic resonance imaging (fMRI). In the chapters of this thesis, we argue and provide evidence for four main points. The first portion (chapter 2) provides a literature review on cortico-striatal and cortico-cerebellar circuit disruption in PD. Using neuroimaging techniques, changes in cerebral and cerebellar activity have been observed in patients with PD compared with healthy participants. Although increases in activity in the cerebellum have often been interpreted as compensatory mechanisms, we provide evidence that they are more likely to be related to pathophysiological changes of the disease, and the disruption of the cortico- cerebellar circuit. In general, we argue (1) is that activity in the cerebellum is linked to the pathophysiology of PD. In the second section (chapter 3) we discuss the effect of levodopa on the patterns of cortical hypo- and hyper-activity in PD, as well as the activity of the putamen in SI and ET movements. Many studies have shown cortical hypo-activity in relation to nigrostriatal dopamine depletion. In contrast, some cognitive studies have also identified increases in cortical activity in patients with PD as compared with healthy control participants. We have previously suggested that cortical hypo- and hyper-activations depend on striatal recruitment. In this thesis, we further show that hyper-activations in the prefrontal cortex are not reestablished with levodopa administration. We suggest (2) that they are rather associated with mesocortical dopamine circuit dysfunction, and perhaps linked with long- term dopaminergic medication administration. Furthermore, we show (3) that levodopa has a non-task specific effect on the motor cortico-striatal loop, but does not affect the cognitive cortico-striatal circuit. Finally (chapter 4), we show that the effect of levodopa on movements of the left and right hands is not symmetrical. Previous studies have shown that in about 50% of patients, one side of the body is more severely affected, and this asymmetry persists throughout the duration of the disease. Our results suggest (4) that levodopa may have stronger effects on the cerebral hemodynamic patterns related to the movements of the more affected hand than on those of the less affected hand.

La maladie de Parkinson (MP) est la deuxième maladie neurodégénérative la plus commune. Les symptômes principalement observés chez les patients atteints de la MP sont la rigidité, les tremblements, la bradykinésie et une instabilité posturale. Leur sévérité est souvent asymétrique. La cause principale de ces symptômes moteurs est la dégénérescence du circuit dopaminergique nigro-striatal qui mène à un débalancement d’activité du circuit cortico-striatal. Ce débalancement de circuits est le point essentiel de cette thèse. Dans les protocoles de recherche décrits ici, des patients atteints de la MP (avant et après une dose de levodopa) et des participants contrôles sains ont effectué des mouvements auto-initiés ou en réponse à des stimulis externes pendant que l’on mesurait leur activité cérébrale en imagerie par résonance magnétique fonctionnelle (IRMf). Dans cette thèse, nous abordons et mettons en évidence quatre (4) points principaux. En première partie (chapitre 2), nous présentons un recensement de la littérature sur les cicruits cortico-striataux et cortico-cérébelleux dans la MP. En utilisant des méthodes de neuroimagerie, des changements d’activité cérébrale et cérébelleuse ont été observés chez les patients atteints de la MP comparés aux participants sains. Même si les augmentations d’activité du cervelet ont souvent été attribuées à des mécanismes compensatoires, nos résultats suggèrent qu’elles sont plus probablement liées aux changements pathophysiologiques de la MP et à la perturbation du circuit cortico-cérébelleux. En général, nous suggérons (1) que le circuit cortico-cérébelleux est perturbé chez les patients atteints de la MP, et que les changements d’activité du cervelet sont liés à la pathophysiologie de la MP plutôt qu’à des mécanismes compensatoires. En deuxième partie (chapitre 3), nous discutons des effets de la levodopa sur les hausses et baisses d’activité observés chez les patients atteints de la MP, ainsi que sur l’activité du putamen pendant les mouvements d’origine interne et externe. De nombreuses études en neuroimagerie ont montré une baisse d’activité (hypo-activité) préfrontale liée à la déplétion de dopamine. En revanche, l’utilisation de tâches cognitives a montré des augmentations d’activité (hyper-activité) corticale chez les patients atteints de la MP comparés aux participants sains. Nous avons suggéré précédemment que ces hypo- et hyper-activités des régions préfrontales dépendent de l’implication du striatum. Dans cette thèse nous suggérons de plus (2) que la levodopa ne rétablit pas ces hyper-activations, mais plutôt qu’elles sont liées à la perturbation du circuit méso-cortical, et aussi possiblement associées à l’administration de médication dopaminergique à long terme. Nous montrons aussi (3) que la levodopa a un effet non-spécifique à la tâche sur l’activité du circuit cortico-striatal moteur, et qu’elle n’a pas d’effet sur l’activité du circuit cortico-striatal cognitif. Nous montrons enfin (chapitre 4) que la levodopa a un effet asymétrique sur les mouvements de la main droite et gauche. À peu près 50% des patients atteints de la MP démontrent une asymétrie des symptômes moteurs, et ceci persiste à travers la durée de la maladie. Nos résultats suggèrent (4) que la levodopa pourrait avoir un plus grand effet sur les patrons d’activations des mouvements de la main la plus affectée. / Parkinson’s disease (PD) is the second most common neurodegenerative disease, mainly manifested by tremor, rigidity, bradykinesia and postural instability, and often an asymmetry of symptom severity of the left and right sides of the body. The depletion of dopamine of the nigrostriatal pathway is the primary cause of the motor symptoms observed in patients with PD, leading to an imbalance in basal-ganglia prefrontal circuits. In the protocols described here, patients with PD before and after levodopa administration and healthy participants performed self-initiated (SI) and externally triggered (ET) movements with the left and right hand during functional magnetic resonance imaging (fMRI). In the chapters of this thesis, we argue and provide evidence for four main points. The first portion (chapter 2) provides a literature review on cortico-striatal and cortico-cerebellar circuit disruption in PD. Using neuroimaging techniques, changes in cerebral and cerebellar activity have been observed in patients with PD compared with healthy participants. Although increases in activity in the cerebellum have often been interpreted as compensatory mechanisms, we provide evidence that they are more likely to be related to pathophysiological changes of the disease, and the disruption of the cortico- cerebellar circuit. In general, we argue (1) is that activity in the cerebellum is linked to the pathophysiology of PD. In the second section (chapter 3) we discuss the effect of levodopa on the patterns of cortical hypo- and hyper-activity in PD, as well as the activity of the putamen in SI and ET movements. Many studies have shown cortical hypo-activity in relation to nigrostriatal dopamine depletion. In contrast, some cognitive studies have also identified increases in cortical activity in patients with PD as compared with healthy control participants. We have previously suggested that cortical hypo- and hyper-activations depend on striatal recruitment. In this thesis, we further show that hyper-activations in the prefrontal cortex are not reestablished with levodopa administration. We suggest (2) that they are rather associated with mesocortical dopamine circuit dysfunction, and perhaps linked with long- term dopaminergic medication administration. Furthermore, we show (3) that levodopa has a non-task specific effect on the motor cortico-striatal loop, but does not affect the cognitive cortico-striatal circuit. Finally (chapter 4), we show that the effect of levodopa on movements of the left and right hands is not symmetrical. Previous studies have shown that in about 50% of patients, one side of the body is more severely affected, and this asymmetry persists throughout the duration of the disease. Our results suggest (4) that levodopa may have stronger effects on the cerebral hemodynamic patterns related to the movements of the more affected hand than on those of the less affected hand.