Le maintien de performances cognitives optimales au cours du vieillissement a été associé à des changements adaptatifs au niveau de l’activité cérébrale relative à diverses habiletés qui tendent à décliner avec l’âge. Peu d’études ont toutefois évalué cette réorganisation neurofonctionnelle dans le cadre des habiletés de communication, notamment en ce qui concerne le langage expressif. Or, considérant que plusieurs composantes langagières demeurent généralement intègres chez les aînés, il s’avère nécessaire d’explorer davantage les mécanismes cérébraux sous-jacents afin de mieux cerner les déterminants du vieillissement cognitif réussi. L’objectif de la présente thèse est d’examiner en détail les effets de l’âge sur les patrons d’activations et les interactions fonctionnelles entre les substrats neuraux contribuant aux habiletés de communication expressive. Deux études en neuroimagerie fonctionnelle, ayant recours à des approches méthodologiques distinctes, ont ainsi été menées à l’aide d’un paradigme mixte novateur et d’une tâche auto-rythmée d’évocation lexicale sémantique et orthographique, effectuée par des participants jeunes et âgés présentant plusieurs années de scolarisation. S’intéressant spécifiquement aux patrons d’activations associés à un rendement élevé à cette tâche, la première étude révèle que le maintien des habiletés d’évocation lexicale lors du vieillissement s’accompagne de changements neurofonctionnels superficiels chez les adultes âgés performants. Par contre, la seconde étude indique que les interactions fonctionnelles entre les régions corticales contribuant aux productions lexicales déclinent considérablement avec l’âge, sans qu’il y ait toutefois d’impact au plan comportemental. Cet effet du vieillissement sur l’intégration fonctionnelle du réseau de l’évocation sémantique et orthographique est aussi exacerbé par la difficulté de la tâche, ce qui s’exprime par des perturbations locales de la connectivité fonctionnelle. Somme toute, cette thèse démontre qu’une réorganisation neurofonctionnelle afin de maintenir les habiletés d’évocation lexicale au cours du vieillissement s’avère superflue chez les adultes âgés instruits et performants, et ce, malgré une diminution des interactions fonctionnelles au sein des réseaux corticaux sous-jacents. Ces résultats reflètent possiblement une perte d’efficience neurale avec l’âge, toutefois insuffisante pour avoir un impact comportemental chez des individus bénéficiant de facteurs de protection susceptibles de favoriser le vieillissement réussi, ce qui est discuté à la lumière du concept de réserve cognitive. / The maintenance of optimal cognitive performance in aging has been associated with adaptive changes in cerebral activation patterns for various abilities that tend to decline with age. Yet, few studies have investigated this neurofunctional reorganization based on communication abilities, particularly with regard to the expressive side of language. Considering that a number of language components usually remain well preserved in older adults, it is, however, necessary to further explore the underlying cerebral mechanisms to better understand the determinants of successful cognitive aging. The objective of this thesis is to examine in detail the effects of aging on activation patterns and functional interactions among neural substrates contributing to expressive communication abilities. Two functional neuroimaging studies, using distinct methodological approaches, have been conducted by means of an innovative mixed design and an overt self-paced semantic and orthographic verbal fluency task, performed by well-educated young and older adults. Specifically focusing of cerebral activation patterns associated with high levels of verbal fluency performance, the first study shows that the relative preservation of lexical speech production abilities in aging is characterized by marginal neurofunctional changes in high-performing older adults. Yet, the second study indicates that functional interactions between cortical areas contributing to speech productions significantly decrease with age, without, however, having an impact at the behavioral level. In addition, this effect of aging on the functional integration of the network engaged during semantic and orthographic verbal fluency was found to be exacerbated by task demands, which is expressed by local functional connectivity disruptions. In sum, this thesis demonstrates that a neurofunctional reorganization to maintain lexical speech production abilities in aging is unnecessary in well-educated and high-performing older adults, despite an age-related decrease in functional networks integration. These results may reflect a loss of neural efficiency with age, although insufficient to have behavioral outcomes in individuals who benefited from protective factors known to promote successful aging, which is discussed in light of the concept of cognitive reserve.

The physical complications of Type 1 diabetes mellitus (T1DM) have been understood as an accelerated ageing process (Morley, 2008). Do people with T1DM also experience accelerated cognitive and brain ageing? Using findings from research of the normal cognitive and brain ageing process and conceptualized in theories of the functional brain changes in cognitive ageing, a combination of cognitive testing and functional magnetic resonance imaging (fMRI) techniques were used to evaluate evidence of accelerated cognitive and brain ageing in middle-aged adults with T1DM. The first part of this thesis comprises a cognitive study of 94 adults (≥ 45 years of age) with long duration (≥ 10 years) of T1DM. Participants completed cognitive assessment and questionnaires on general mood and feelings about living with diabetes. Findings highlighted the importance of microvascular disease (specifically retinopathy) as an independent predictor of cognitive function. The incidence and predictors of mild cognitive impairment (MCI) were then explored. Results indicate a higher percentage of the group met criteria for MCI than expected based on incidence rates in the general population, providing initial evidence of accelerated cognitive ageing. Psychological factors were explored next. The relationship between the measures of well-being, diabetes health, and cognitive function highlighted the need for attention to patient's psychological well-being in diabetes care. Finally, a subgroup of 30 participants between the ages of 45 and 65 who differed on severity of retinopathy were selected to take part in an fMRI study. Blood oxygen level dependent (BOLD) activity was evaluated while participants were engaged in cognitive tasks and during rest. The findings provided evidence that the pattern of BOLD activation and functional connectivity for those with high severity of retinopathy are similar to patterns found in adults over the age of 65. In line with the theories of cognitive ageing, functional brain changes appear to maintain a level of cognitive function. Evidence of accelerated brain ageing in this primarily middle-aged group, emphasizes the importance of treatments and regimens to prevent or minimize microvascular complications.

No description available.

A inferência do curso temporal da atividade neural a partir do efeito BOLD é um importante problema, ainda em aberto. A forma da curva BOLD não reflete diretamente as características temporais da atividade eletrofisiológica dos neurônios. Nessa tese, é introduzido o conceito de tempo de processamento neural (TPN) como um dos parâmetros do modelo biofísico da função de resposta hemodinâmica (HRF). O objetivo da introdução desse conceito é obter estimativas mais acuradas da duração da atividade neural a partir do efeito BOLD, que possui auto grau de nãolinearidade. Duas formas de estimar os parâmetros do modelo do efeito BOLD foram desenvolvidas. A validade e aplicabilidade do conceito de TPN e das rotinas de estimação foram avaliadas por simulações computacionais e análise de séries temporais experimentais. Os resultados das simulações e da aplicação foram comparados com medidas da forma da HRF. O experimento analisado consistiu em um paradigma de tomada de decisão na presença de distratores emocionais. Esperase que o TPN em áreas sensoriais primárias seja equivalente ao tempo de apresentação de estímulos. Por outro lado, o TPN em áreas relacionadas com a tomada de decisão deve ser menor que a duração dos estímulos. Além disso, o TPN deve depender da condição experimental em áreas relacionadas ao controle de distratores emocionais. Como predito, o valores estimados do TPN no giro fusiforme foram equivalentes à duração dos estímulos e o TPN no giro do cíngulo dorsal variou com a presença de distrator emocional. Observou-se ainda lateralidade do TPN no córtex pré-frontal dorsolateral. As medidas da forma da HRF obtidas por um método convencional não dectectaram as variações observadas no TPN / The extraction of information about neural activity dynamics related to the BOLD signal is a challenging task. The temporal evolution of the BOLD signal does not directly reflect the temporal characteristics of electrical activity of neurons. In this work, we introduce the concept of neural processing time (NPT) as a parameter of the biophysical model of the hemodynamic response function (HRF). Through this new concept we aim to infer more accurately the duration of neuronal response from the highly nonlinear BOLD effect. We describe two routines to estimate the parameters of the HRF model. The face validity and applicability of the concept of NPT and the estimation procedures are evaluated through simulations and analysis of experimental time series. The results of both simulation and application were compared with summary measures of HRF shape. We analysed an experiment based on a decision-making paradigm with simultaneous emotional distracters. We hypothesize that the NPT in primary sensory areas is approximately the stimulus presentation duration. On the other hand, the NPT in brain areas related to decisionmaking processes should be less than the stimulus duration. Moreover, in areas related to processing of an emotional distracter, the NPT should depend on the experimental condition. As predicted, the NPT in fusiform gyrus is close to the stimulus duration and the NPT in dorsal anterior cingulate gyrus depends on the presence of an emotional distracter. Interestingly, the estimated NPTs in the dorsolateral prefrontal cortex indicate functional laterality of this region. The analysis using standard measures of HRF did not detect the variations observed in our method (NPT)

Les fonctions cognitives reposent sur la communication dynamique de régions cérébrales interconnectées. Dans la maladie d’Alzheimer (MA), les travaux antérieurs suggèrent que le processus neuropathologique cible de façon précoce un ou plusieurs systèmes anatomo-fonctionnels spécifiques. La dysfonction du réseau par défaut a été objectivée de façon consistante. Cependant, ses relations avec les symptômes cliniques et avec l’atteinte des régions du lobe temporal interne qui lui sont fonctionnellement connectées restent à clarifier. L’IRM fonctionnelle de repos est une technique pertinente pour caractériser in vivo chez l’Homme la connectivité cérébrale.Par une approche des systèmes neuraux, ce travail de thèse a pour objectif de caractériser la réorganisation fonctionnelle neuronale dans la MA, ses corrélats cliniques, ainsi que l’influence de l’âge de début des symptômes. Par le recueil et l’analyse des données neuropsychologiques, en IRMf de repos et en IRM structurale, acquises chez des sujets avec des troubles de la mémoire et avec une forme mnésique légère de MA, notre travail apporte des éclairages : i) sur l’implication du réseau temporal antérieur dans la mémoire déclarative décontextualisée et ses modifications dans le cours de la MA ; ii) sur les similitudes et spécificités des systèmes anatomo-fonctionnels ciblés dans les deux formes cliniques distinctes - à début précoce et tardif - de la MA ; iii) sur la réorganisation de l’organisation topologique cérébrale dans son ensemble de ces deux formes de la maladie. / Cognitive functions rely on the dynamic interplay of connected brain regions. Previous studies suggest that in Alzheimer disease (AD), early pathological changes target one or several specific anatomo-functional networks. Dysfunction of the default mode network is a consistent finding. However, its relationship with clinical symptoms and interconnected medial temporal regions remains to be clarified. Resting state functional MRI (fMRI) is an emerging method aimed at characterizing in vivo brain connectivity in the Human.Using a neural system approach, the aim of this thesis was to characterize neuronal functional reorganization in AD, its clinical correlates, and to determine the influence of age at onset. Neuropsychological data, structural and fMRI were obtained in subjects with early memory impairment and mild “amnestic” AD. This work provides new insights into : i) the functional role of the anterior temporal network in context-free declarative memory and its changes throughout the course of AD; ii) the common and specific features in targeted anatomo-functional networks between early and late onset AD ; iii) the reorganization of whole brain topological properties in the two forms of the disease.

Hitherto, previous studies on the cerebral correlates of fluid intelligence (fluIQ) used tasks that did not exclusively demand fluIQ, or were restricted to participants of average fluIQ (ave-fluIQ) solving intelligence test items of varying difficulty, thus not allowing assumptions on interindividual differences in fluIQ. Geometric analogical reasoning (GAR) demands fluIQ very purely and thus is an eligible approach for research on interindividual differences in fluIQ. In a first study, we examined the cerebral correlates of GAR, and showed the involvement of parietal and frontal brain regions. This is in line with the assumptions of the parieto-frontal integration theory (P-FIT) of intelligence and with literature reports for other visuo-spatial tasks. Building upon these findings, we report results from a second study with high fluIQ (hi-fluIQ) and ave-fluIQ school students solving a GAR task. Again in line with the P-FIT model, we demonstrated that the parieto-frontal network is involved in GAR in both groups. However, the extent of task-related brain activation in parietal and frontal brain regions was differentially modulated by fluIQ. Our results thus partly run counter to the postulates of the neural efficiency hypothesis, which assumes a negative brain activation-intelligence relationship. We conclude that this relationship is not generally unitary; rather, it can be conjectured that the adaptive and flexible modulation of brain activation is characteristic of hi-fluIQ. Knowledge on the stability of the cerebral correlates of hi-fluIQ during adolescence had been sparse. To elucidate this field, we examined the follow-up stability of the cerebral correlates of GAR in hi-fluIQ in a third study. We demonstrated that the relevant brain network is in place already at age 17 and that improvements in behavioral performance at age 18 due to task familiarity are indicative of more efficient use of the cerebral resources available. / Bisherige Studien zu zerebralen Korrelaten fluider Intelligenz (fluIQ) haben Aufgaben verwendet, die fluIQ nicht in Reinform erfordern oder haben Probanden mit durchschnittlicher fluIQ (ave-fluIQ) beim Lösen von Intelligenztestaufgaben mit variierenden Schwierigkeitsstufen untersucht und ermöglichen daher keine Aussagen zu interindividuellen Unterschieden in fluIQ. Geometrisches analoges Schließen (GA) beansprucht fluIQ in Reinform und eignet sich daher als differentielles Untersuchungsparadigma. In einer ersten Studie haben wir die zerebralen Korrelate des GA untersucht und nachgewiesen, dass parietale und frontale Hirnregionen involviert sind. Dies steht im Einklang mit der parieto-frontalen Integrationstheorie (P-FIT) der Intelligenz und mit Literaturberichten zu anderen visuell-räumlichen Aufgaben. Aufbauend auf diesen Befunden berichten wir Ergebnisse einer zweiten Studie, in der Schüler mit hoher fluIQ (hi-fluIQ) und ave-fluIQ GA-Aufgaben lösten. In Übereinstimmung mit den Annahmen des P-FIT-Modells konnten wir zeigen, dass GA in beiden Gruppen das parieto-frontale Netzwerk beansprucht. Das Ausmaß der Hirnaktivierung wurde jedoch differentiell durch fluIQ moduliert. Unsere Ergebnisse widersprechen damit teilweise den Postulaten der neuralen Effizienztheorie, die einen negativen Zusammenhang zwischen Hirnaktivierung und Intelligenz annimmt. Wir schlussfolgern, dass dieser Zusammenhang nicht generell einseitig gerichtet ist, sondern die flexible Modulation von Hirnaktivierung charakteristisch für hi-fluIQ ist. Befunde zur Stabilität zerebraler Korrelate von hi-fluIQ in der Jugend waren bisher rar. Um dieses Feld zu beleuchten, haben wir die follow-up-Stabilität zerebraler Korrelate des GA in der hi-fluIQ Gruppe in einer dritten Studie untersucht. Wir konnten zeigen, dass das relevante zerebrale Netzwerk schon mit 17 Jahren etabliert ist und Performanzverbesserungen über die Zeit für eine effizientere Nutzung der verfügbaren zerebralen Ressourcen sprechen.

A inferência do curso temporal da atividade neural a partir do efeito BOLD é um importante problema, ainda em aberto. A forma da curva BOLD não reflete diretamente as características temporais da atividade eletrofisiológica dos neurônios. Nessa tese, é introduzido o conceito de tempo de processamento neural (TPN) como um dos parâmetros do modelo biofísico da função de resposta hemodinâmica (HRF). O objetivo da introdução desse conceito é obter estimativas mais acuradas da duração da atividade neural a partir do efeito BOLD, que possui auto grau de nãolinearidade. Duas formas de estimar os parâmetros do modelo do efeito BOLD foram desenvolvidas. A validade e aplicabilidade do conceito de TPN e das rotinas de estimação foram avaliadas por simulações computacionais e análise de séries temporais experimentais. Os resultados das simulações e da aplicação foram comparados com medidas da forma da HRF. O experimento analisado consistiu em um paradigma de tomada de decisão na presença de distratores emocionais. Esperase que o TPN em áreas sensoriais primárias seja equivalente ao tempo de apresentação de estímulos. Por outro lado, o TPN em áreas relacionadas com a tomada de decisão deve ser menor que a duração dos estímulos. Além disso, o TPN deve depender da condição experimental em áreas relacionadas ao controle de distratores emocionais. Como predito, o valores estimados do TPN no giro fusiforme foram equivalentes à duração dos estímulos e o TPN no giro do cíngulo dorsal variou com a presença de distrator emocional. Observou-se ainda lateralidade do TPN no córtex pré-frontal dorsolateral. As medidas da forma da HRF obtidas por um método convencional não dectectaram as variações observadas no TPN / The extraction of information about neural activity dynamics related to the BOLD signal is a challenging task. The temporal evolution of the BOLD signal does not directly reflect the temporal characteristics of electrical activity of neurons. In this work, we introduce the concept of neural processing time (NPT) as a parameter of the biophysical model of the hemodynamic response function (HRF). Through this new concept we aim to infer more accurately the duration of neuronal response from the highly nonlinear BOLD effect. We describe two routines to estimate the parameters of the HRF model. The face validity and applicability of the concept of NPT and the estimation procedures are evaluated through simulations and analysis of experimental time series. The results of both simulation and application were compared with summary measures of HRF shape. We analysed an experiment based on a decision-making paradigm with simultaneous emotional distracters. We hypothesize that the NPT in primary sensory areas is approximately the stimulus presentation duration. On the other hand, the NPT in brain areas related to decisionmaking processes should be less than the stimulus duration. Moreover, in areas related to processing of an emotional distracter, the NPT should depend on the experimental condition. As predicted, the NPT in fusiform gyrus is close to the stimulus duration and the NPT in dorsal anterior cingulate gyrus depends on the presence of an emotional distracter. Interestingly, the estimated NPTs in the dorsolateral prefrontal cortex indicate functional laterality of this region. The analysis using standard measures of HRF did not detect the variations observed in our method (NPT)

Le problème d'estimer une fonction de plusieurs variables à partir d'une observation corrompue surgit dans de nombreux domaines d'ingénierie. Par exemple, en imagerie médicale cette tâche a attiré une attention particulière et a, même, motivé l'introduction de nouveaux concepts qui ont trouvé des applications dans de nombreux autres domaines. Cet intérêt est principalement du au fait que l'analyse des données médicales est souvent effectuée dans des conditions difficiles car on doit faire face au bruit, au faible contraste et aux transformations indésirables inhérents aux systèmes d'acquisition. D'autre part , le concept de parcimonie a eu un fort impact sur la reconstruction et la restauration d'images au cours des deux dernières décennies. La parcimonie stipule que certains signaux et images ont des représentations impliquant seulement quelques coefficients non nuls. Cela est avéré être vérifiable dans de nombreux problèmes pratiques. La thèse introduit de nouvelles constructions d'a priori de parcimonie dans le cas des ondelettes et de la variation totale. Ces constructions utilisent une notion d'anisotopie généralisée qui permet de regrouper des variables ayant des comportements similaires : ces comportement peuvent peut être liée à la régularité de la fonction, au sens physique des variables ou bien au modèle d'observation. Nous utilisons ces constructions pour l'estimation non-paramétriques de fonctions. Dans le cas des ondelettes, nous montrons l'optimalité de l'approche sur les espaces fonctionnelles habituels avant de présenter quelques exemples d’applications en débruitage de séquences d'images, de données spectrales et hyper-spectrales, écoulements incompressibles ou encore des images ultrasonores. En suite, nous modélisons un problème déconvolution de données d'imagerie par résonance magnétique fonctionnelle comme un problème de minimisation faisant apparaître un a priori de variation totale structuré en espace-temps. Nous adaptons une généralisation de l'éclatement explicite-implicite pour trouver une solution au problème de minimisation. / The problem of estimating a multivariate function from corrupted observations arises throughout many areas of engineering. For instance, in the particular field of medical signal and image processing, this task has attracted special attention and even triggered new concepts and notions that have found applications in many other fields. This interest is mainly due to the fact that the medical data analysis pipeline is often carried out in challenging conditions, since one has to deal with noise, low contrast and undesirable transformations operated by acquisition systems. On the other hand, the concept of sparsity had a tremendous impact on data reconstruction and restoration in the last two decades. Sparsity stipulates that some signals and images have representations involving only a few non-zero coefficients. The present PhD dissertation introduces new constructions of sparsity priors for wavelets and total variation. These construction harness notions of generalized anisotropy that enables grouping variables into sub-sets having similar behaviour; this behaviour can be related to the regularity of the unknown function, the physical meaning of the variables or the observation model. We use these constructions for non-parametric estimation of multivariate functions. In the case of wavelet thresholding, we show the optimality of the procedure over usual functional spaces before presenting some applications on denoising of image sequence, spectral and hyperspectral data, incompressible flows and ultrasound images. Afterwards, we study the problem of retrieving activity patterns from functional Magnetic Resonance Imaging data without incorporating priors on the timing, durations and atlas-based spatial structure of the activation. We model this challenge as a spatio-temporal deconvolution problem. We propose the corresponding variational formulation and we adapt the generalized forward-backward splitting algorithm to solve it.

SUMMARY Non Rapid Eye Movement (NREM) sleep in humans is defined by spontaneous neural activities organized by specific rhythms or oscillations. The aim of this thesis is to characterize, by means of neuroimaging techniques, the shaping of brain function by these physiological rhythms. The studied oscillations are sleep spindles, delta waves and slow oscillation, representing the main identifiable neurophysiological events of human NREM sleep. Sleep spindles are a hallmark of light NREM sleep. They are commonly described on electroencephalographic (EEG) recordings as 11-15 Hz oscillations, lasting more than 0.5 sec and with a typical waxing-and-waning waveform. During deeper stages of NREM sleep, spindles are progressively replaced by a slow wave activity (SWA; 0.5-4 Hz), which encompasses delta waves (1-4 Hz) and slow oscillations (0.5-1 Hz). In combination with EEG, we studied these rhythms using two different functional brain imaging techniques : positron emission tomography (PET) and functional magnetic resonance imaging (fMRI). These studies originally contribute to the understanding of the generating mechanisms and functional roles of NREM sleep oscillations, which are a hallmark of sleep architecture in healthy humans. Neural correlates of NREM sleep oscillations assessed by EEG / PET In this section, we report the analyses of PET data devoted to the study of NREM sleep oscillations. We characterized the brain areas in which activity, measured in terms of regional cerebral blood flow (rCBF), was correlated with EEG spectral power in the spindle (11-15 Hz), delta waves (1-4 Hz) and slow oscillation (0.5-1 Hz) frequency bands, in 23 non-sleep-deprived young healthy volunteers. EEG activity in the spindle frequency band was negatively correlated with rCBF in the thalamus. This result was in agreement with data suggesting the generation of spindles within cortico-thalamo-cortical loops (Steriade, 2006). Spectral power in the delta band was negatively correlated with rCBF in the medial prefrontal cortex, striatum, insula, anterior cingulate cortex, precuneus and basal forebrain, which are structures potentially involved in the modulation of cortical delta waves (Dang-Vu et al., 2005b). The functional brain mapping of slow oscillations was highly similar to the one of delta waves, in keeping with the hypothesis that both types of oscillations share common physiological mechanisms. These results consisted in negative correlations, which means that the cerebral blood flow in these areas was lower when the power in the corresponding frequency band was higher. The different rhythms of NREM sleep are synchronized by the slow oscillation, which alternates a hyperpolarization phase during which cortical neurons remain silent, and a depolarization phase associated with important neuronal firing. The prominent effect of hyperpolarization phases could account for the decrease in blood flow found in PET studies. Indeed, PET has a limited temporal resolution, around one minute, and therefore averages brain activity over relatively long periods, during which hyperpolarization phases predominate. Thus PET imaging does not allow to directly study brief events, lasting one second or so, such as NREM sleep oscillations. Besides, the spectral power values used in PET studies are just an indirect reflection of the appearance of these rhythms during sleep. These considerations justify the use of fMRI because, together with improved spatial resolution, its temporal resolution around one second allows to assess brain responses associated to the occurrence of NREM sleep oscillations, taken as identifiable events. Neural correlates of NREM sleep oscillations assessed by EEG / fMRI The largest section of the thesis is devoted to the use of fMRI in the study of NREM sleep oscillations. We characterized the brain areas in which activity, measured in terms of blood oxygen level dependent (BOLD) signal, was correlated with the occurrence of NREM sleep oscillations. Compared to EEG with PET, EEG recording with simultaneous fMRI was technically much more challenging. In particular, the analysis of EEG data acquired simultaneously with fMRI required a complex signal processing in order to remove all artefacts induced during the scanning procedure. After clean EEG data had been obtained, automatic detection of spindles (Molle et al., 2002), delta waves and slow oscillations (Massimini et al., 2004) was performed according to published criteria, and provided the series of events to be used as regressors in the statistical analysis of fMRI data. The latter assessed the main effects of spindles, delta waves and slow oscillations on BOLD signal changes across the 14 non-sleep-deprived young healthy volunteers selected for this study. Spindles were analysed considering 2 potential subtypes. Indeed, in humans, while most spindles are recorded in central and parietal regions and display a frequency around 14 Hz (fast spindles), others are prominent on frontal derivations with a frequency around 12 Hz (slow spindles). Previous data also show differences between both subtypes in their modulation by age, circadian and homeostatic factors, menstrual cycle, pregnancy and drugs (De Gennaro and Ferrara, 2003). However, no clear evidence of a distinct neurobiological basis for these two subtypes of spindles has been demonstrated so far. After automatic detection of spindles and their differentiation as fast and slow, we showed that the two subtypes were associated with activation of partially distinct thalamo-cortical networks. These data further support the existence of 2 subtypes of sleep spindles modulated by segregated neural networks (Schabus et al., 2007). Slow oscillation has initially been described at the cellular level in animals as an oscillation <1 Hz of membrane potential, alternating a hyperpolarization phase (down) during which cortical neurons are silent and a depolarization phase (up) associated with intense neuronal firing (Steriade, 2006). At the macroscopic level, this slow rhythm is found on human EEG recordings as high amplitude slow waves, defined by a peak-to-peak amplitude of more than 140 µV (Massimini et al., 2004). The slow oscillation also synchronizes other NREM sleep rhythms such as spindles (Molle et al., 2002) and delta waves (defined here as waves of lower peak-to-peak amplitude : between 75 and 140 µV). The organization of NREM sleep by the slow oscillation suggests that NREM sleep should be characterized by increased brain activities associated with the up state of slow oscillation. Indeed, we observed significant BOLD signal changes in relation to both slow waves and delta waves in specific brain areas including inferior and medial frontal gyrus, parahippocampal gyrus, precuneus, posterior cingulate cortex, ponto-mesencephalic tegmentum and cerebellum. All these responses consisted in brain activity increases. These results stand in sharp contrast with earlier sleep studies, in particular PET studies, reporting decreases in brain activity during NREM sleep. Here we showed that NREM sleep cannot be reduced to a state of global and regional brain activity decrease, but is actually an active state during which phasic increases in brain activity are synchronized to the slow oscillation. We then compared brain responses to delta and slow waves respectively and found no significant difference. In agreement with our PET data, this result suggests that slow waves and delta waves share common neurobiological mechanisms. However, when effects of slow and delta waves were tested separately, we observed that slow waves were specifically associated with activation of brainstem and mesio-temporal areas, while delta waves were associated with activation of inferior and medial frontal areas. This result is important in regard to the potential role of slow oscillation in memory consolidation during sleep (Marshall et al., 2006). Indeed, the preferential activation of mesio-temporal areas with high amplitude slow waves suggests that the amplitude of the wave is a crucial factor in the recruitment during sleep of brain structures involved in the processing of memory traces. RESUME Le sommeil lent de lhomme est défini par la présence dactivités neuronales spontanées, organisées sous forme de rythmes ou oscillations spécifiques. Lobjectif des travaux réalisés dans le cadre de cette thèse est de caractériser, par des méthodes de neuroimagerie, le fonctionnement cérébral au cours de ces rythmes physiologiques. Les oscillations que nous avons étudiées sont les fuseaux du sommeil, les ondes delta et les oscillations lentes, représentant les principales activités neurophysiologiques identifiables chez lhomme au cours du sommeil lent. Les fuseaux du sommeil constituent un élément essentiel du sommeil lent léger. Ils sont communément décrits sur les enregistrements électroencéphalographiques (EEG) comme des oscillations de fréquence comprise entre 11 et 15 Hz, dune durée dau moins 0,5 sec, et de morphologie caractéristique daugmentation puis de diminution damplitude. Au cours des stades plus profonds de sommeil lent, les fuseaux sont en grande partie remplacés par une activité donde lente (SWA; 0,5-4 Hz) qui recouvre les ondes delta (1-4 Hz) et les oscillations lentes (0,5-1 Hz). En combinaison à lEEG, nous avons utilisé deux techniques dimagerie fonctionnelle différentes pour étudier ces rythmes: la tomographie par émission de positons (PET) et limagerie en résonance magnétique fonctionnelle (fMRI). Ces études apportent une contribution originale à notre compréhension du sommeil lent chez lhomme sain, par lexploration des mécanismes générationnels de ces oscillations, piliers de larchitecture du sommeil. Corrélats cérébraux des rythmes du sommeil lent en EEG / PET Dans cette section, nous décrivons lutilisation de la PET dans létude des rythmes du sommeil lent. Nous avons caractérisé les régions cérébrales dans lesquelles lactivité, mesurée en terme de débit sanguin cérébral régional (rCBF), était corrélée à la puissance spectrale EEG dans la bande de fréquence des fuseaux (11-15 Hz), des ondes delta (1-4 Hz) et des oscillations lentes (0.5-1 Hz), chez 23 jeunes volontaires sains et non privés de sommeil. Lactivité EEG dans la bande des fuseaux était corrélée négativement avec le rCBF dans le thalamus. Ce résultat est en accord avec les données suggérant la genèse des fuseaux par des boucles dinteraction cortico-thalamo-corticale (Steriade, 2006). La puissance spectrale dans la bande delta était négativement corrélée avec le rCBF au niveau du cortex préfrontal médial, du striatum, de linsula, du cortex cingulaire antérieur, du précuneus et du télencéphale basal, régions potentiellement impliquées dans la modulation des ondes delta corticales (Dang-Vu et al., 2005b). La carte des oscillations lentes était superposable à celle des ondes delta, ce qui suggère que ces deux types doscillations relèvent chez lhomme de mécanismes physiologiques communs. Ces résultats démontraient donc des corrélations négatives, ce qui signifie que le débit sanguin cérébral dans ces régions était dautant plus faible que la puissance dans la bande de fréquence correspondante était élevée. Linterprétation de ce phénomène doit intégrer le fait que les différents rythmes du sommeil lent sont sculptés par loscillation lente, laquelle alterne une phase dhyperpolarisation au cours de laquelle les neurones corticaux sont silencieux, et une phase de dépolarisation au cours de laquelle ils déchargent en bouffées. Leffet prépondérant des phases dhyperpolarisation pourrait expliquer la baisse de débit cérébral démontrée en PET. En effet, cette dernière présente une résolution temporelle limitée, de lordre de la minute, ce qui a pour effet dintégrer lactivité cérébrale sur des périodes de temps relativement longues, au cours desquelles les phases dhyperpolarisation corticale prédominent. Limagerie en PET ne permet pas donc pas détudier directement des événements brefs de lordre de la seconde, tels que les oscillations du sommeil lent. En outre, les valeurs de puissance spectrale utilisées pour caractériser ces rythmes en PET ne reflètent quindirectement leur survenue au cours du sommeil. Ces considérations justifient le recours à limagerie en fMRI, dont la résolution temporelle de lordre de la seconde permet dévaluer les réponses cérébrales associées à la survenue des oscillations du sommeil lent, considérées cette fois comme des événements identifiables. Corrélats cérébraux des rythmes du sommeil lent en EEG / fMRI Dans cette partie, la plus importante, nous décrivons lanalyse en fMRI des rythmes du sommeil lent. Nous avons caractérisé les régions cérébrales dont l'activité, mesurée par le signal BOLD, était corrélée à la survenue des oscillations du sommeil lent. Par rapport à la situation rencontrée en PET, lenregistrement des données EEG nécessaire à la détection des rythmes du sommeil lent, simultanément à lacquisition fMRI, a posé des difficultés techniques considérablement plus grandes. En particulier, linterprétation de lEEG dans ces conditions a nécessité un traitement précis du signal afin den éliminer les éléments artéfactuels qui le contaminent. Ce nest quaprès ce processus que la détection automatique des fuseaux (Molle et al., 2002), des ondes delta et des oscillations lentes (Massimini et al., 2004) selon des critères publiés a pu seffectuer, permettant dobtenir les séries dévénements qui furent entrés comme régresseurs dans lanalyse statistique des données fMRI. Cette dernière évalue leffet principal des fuseaux, ondes delta et oscillations lentes sur les variations du signal BOLD chez lensemble des 14 jeunes volontaires sains et non privés de sommeil sélectionnés pour létude. En ce qui concerne les fuseaux, ils furent subdivisés en 2 sous-types. Chez lhomme en effet, alors que la grande majorité des fuseaux sont enregistrés dans les régions centrales et pariétales, avec une fréquence denviron 14 Hz (fuseaux rapides), dautres fuseaux dits lents (environ 12 Hz) prédominent dans les régions frontales. Des données antérieures rapportent également des différences entre ces deux sous-types en ce qui concerne leur modulation par des paramètres comme lâge, les facteurs circadiens et homéostatiques, la phase du cycle menstruel, la grossesse et certains agents pharmacologiques (De Gennaro and Ferrara, 2003). Cependant, aucune description formelle dun substrat biologique distinct navait encore été établie pour ces 2 sous-types de fuseaux. Après détection automatique des fuseaux et leur ségrégation en fuseaux rapides et lents, nous avons pu démontrer que les 2 sous-types de fuseaux étaient associés à des activations dans des réseaux thalamo-corticaux partiellement distincts. Ces données apportent donc des arguments pour établir lexistence de 2 sous-types biologiquement différenciés de fuseaux du sommeil (Schabus et al., 2007). Loscillation lente du sommeil lent a été décrite initialement au niveau cellulaire chez lanimal comme une oscillation de fréquence <1Hz et qui alterne une phase dhyperpolarisation (ou down), au cours de laquelle les neurones corticaux sont silencieux, et une phase de dépolarisation (ou up) qui correspond à une période de décharges neuronales intenses (Steriade, 2006). Chez lhomme, cette oscillation lente est également retrouvée sur les enregistrements EEG de surface sous forme dondes lentes de haute amplitude, définies par une amplitude pic-à-pic de plus de 140 µV (Massimini et al., 2004). Loscillation lente synchronise aussi dautres rythmes du sommeil lent tels les fuseaux (Molle et al., 2002) et les ondes delta (définies ici par des ondes de plus basse amplitude pic-à-pic : entre 75 et 140 µV). Lorganisation du sommeil lent par ces oscillations lentes suggère que le sommeil lent devrait être marqué par des activations cérébrales survenant en synchronie avec les phases up des oscillations lentes. De fait, nous avons observé des variations significatives de signal BOLD en association avec les ondes lentes et delta dans des régions cérébrales spécifiques incluant le gyrus frontal inférieur et médial, le gyrus parahippocampique, le precuneus, le cortex cingulaire postérieur, le tegmentum ponto-mésencéphalique et le cervelet. Ces variations étaient positives dans toutes les régions mises en évidence, ce qui traduit une augmentation dactivité. Ces résultats sont originaux en ce quils suggèrent que le sommeil lent, contrairement à ce qui était conclu des précédentes études du sommeil chez lhomme (particulièrement en PET), ne se réduit pas à une hypoactivation cérébrale globale et régionale. Au contraire, nos données montrent que le sommeil lent saccompagne dune activation cérébrale phasique rythmée par la phase de dépolarisation des oscillations lentes. Nous avons ensuite comparé les réponses cérébrales aux ondes delta et celles aux ondes lentes. Aucune région cérébrale ne présentait dactivité significativement différente en fonction des 2 types dondes. En accord avec nos données PET, ce résultat suggère quil ny a pas de différence formelle sur le plan des mécanismes neurobiologiques entre ondes lentes et ondes delta. Toutefois, lorsque les effets des ondes lentes et delta furent testés séparément, nous avons observé que les ondes lentes activaient spécifiquement le tronc cérébral et le cortex mésio-temporal alors que les ondes delta activaient les aires frontales inférieure et médiale. Cet résultat est important si lon considère en particulier le rôle potentiel des oscillations lentes dans la consolidation des traces mnésiques au cours du sommeil (Marshall et al., 2006). Lactivation préférentielle des aires mésio-temporales avec les ondes lentes de haute amplitude suggère en effet que lamplitude de londe est un paramètre déterminant dans le recrutement au cours du sommeil de structures cérébrales impliquées dans le traitement des traces mnésiques.

Theory of Mind (ToM) is the ability to infer others’ mental states and thus to predict their behavior (Perner, 1991). Therefore, ToM is essential for the adequate adjustment of behavior in social situations. ToM can be divided into: 1) cognitive ToM encompassing inferences about intentions and beliefs and 2) affective ToM encompassing inferences about emotions (Shamay-Tsoory, Harari, Aharon-Peretz, & Levkovitz, 2010). Well-functioning skills of both ToM aspects are much-needed in the developmental period of adolescence because in this age phase peer relationships become more important and romantic relationships arise (Steinberg & Morris, 2001). Importantly, affective psychopathological disorders often have their onset in adolescence. ToM development in adolescence might be based on underlying cognitive mechanisms such as the ability to inhibit one’s own thoughts in order to understand another person’s thoughts (Carlson & Moses, 2001). Another possible mechanism relates to functional brain development across adolescence (Blakemore, 2008). Therefore, neurocognitive mechanisms may underlie ongoing ToM development in adolescence. First studies indicate an ongoing behavioral and functional brain development of ToM (e.g. Blakemore, 2008). However, ToM development in adolescence and how this might relate to underlying (neuro)cognitive functions remains largely underexamined. The major aims of the current thesis were first to answer the overall question whether there is an ongoing development of ToM in adolescence. This question relates to both behavioral and functional brain development. As a second major aim, the present work sought to elucidate possible (neuro)cognitive mechanisms of ongoing ToM development across adolescence. Specifically, these cognitive mechanisms might be basic cognitive functions as well as executive functions. Additionally, the present work aimed at exploring potential (neuro)cognitive mechanisms through an integration of both behavioral and functional brain studies. The current experimental work spans three cross-sectional studies investigating adolescents (aged around 12-15 years) and young adults (aged around 18-22 years) to examine for the first time both the behavioral (studies I and II) and functional brain development of ToM (study III) in adolescence and its underlying (neuro)cognitive mechanisms. In all three studies, more complex, advanced ToM tasks were employed to avoid ceiling effects. Study I was aimed at investigating if cognitive and affective ToM continues to develop in adolescence and at exploring if basic cognitive variables such as verbal ability, speed of processing, and working memory capacity underlie such development. Hence, two groups of adolescents and young adults completed tasks of ToM and basic cognitive abilities. Large age effects were revealed on both measures of ToM: adolescents performed lower than adults. These age differences remained significant after controlling for basic cognitive variables. However, verbal ability covaried with performance in affective ToM. Overall, results support the hypothesis of an ongoing development of ToM from adolescence to adulthood on both cognitive and affective aspects. Results may further indicate verbal ability being a basic cognitive mechanism of affective ToM. Study II was designed to further explore if affective ToM, as measured with a dynamic realistic task, continues to develop across adolescence. Importantly, this study sought to explore executive functions as higher cognitive mechanisms of developing affective ToM across adolescence. A large group spanning adolescents and young adults evaluated affective mental states depicted by actors in video clips. Additionally, participants were examined with three subcomponents of executive functions, inhibition, updating, and shifting following the classification of Miyake et al. (2000). Affective ToM performance was positively related to age and all three executive functions. Specifically, inhibition explained the largest amount of variance in age related differences of affective ToM performance. Overall, these results indicate the importance of inhibition as key underlying mechanism of developing an advanced affective ToM in adolescence. Study III set out to explore the functional brain development of affective ToM in adolescence by using functional magnetic resonance imaging (fMRI). The affective ToM measure was the behavioral developmentally sensitive task from study II. An additional control condition consisted of the same emotional stimuli with the instruction to focus on physical information. This study faced methodical challenges of developmental fMRI studies by matching performance of groups. The ventromedial prefrontal cortex (vMPFC) was significantly less deactivated in adolescents in comparison to adults, which might suggest that adolescents seem to rely more on self-referential processes for affective ToM. Furthermore, adolescents compared to adults showed greater activation in the dorsolateral prefrontal cortex (DLPFC) in the control condition, indicating that adolescents might be distracted by the emotional content and therefore needed to focus more on the physical content of the stimulus. These findings suggest affective ToM continues to develop on the functional brain level and reveals different underlying neurocognitive strategies for adolescents in contrast to adults. In summary, the current thesis investigated whether ToM continues to develop in adolescence until young adulthood and explored underlying (neuro)cognitive mechanisms. Findings suggest that there is indeed an ongoing development of both the cognitive and affective aspect of ToM, which importantly contributes to the conceptual debate. Moreover, the second benefit to the debate is to demonstrate how this change may occur. As a basic cognitive mechanism verbal ability and as an executive functioning mechanism inhibition was revealed. Furthermore, neurocognitive mechanisms in form of different underlying neurocognitive strategies of adolescents compared to adults were shown. Taken together, ToM development in adolescence seems to mirror a different adaptive cognitive style in adolescence (Crone & Dahl, 2012). This seems to be important for solving the wealth of socio-emotional developmental tasks that are relevant for this age span.