Biological brains exhibit many interesting and complex behaviours. Understanding of the mechanisms behind brain behaviours is critical for continuing advancement in fields of research such as artificial intelligence and medicine. In particular, synchronisation of neuronal firing is associated with both improvements to and degeneration of the brain's performance; increased synchronisation can lead to enhanced information-processing or neurological disorders such as epilepsy and Parkinson's disease. As a result, it is desirable to research under which conditions synchronisation arises in neural networks and the possibility of controlling its prevalence. Stochastic ensembles of FitzHugh-Nagumo elements are used to model neural networks for numerical simulations and bifurcation analysis. The FitzHugh-Nagumo model is employed because of its realistic representation of the flow of sodium and potassium ions in addition to its advantageous property of allowing phase plane dynamics to be observed. Network characteristics such as connectivity, configuration and size are explored to determine their influences on global synchronisation generation in their respective systems. Oscillations in the mean-field are used to detect the presence of synchronisation over a range of coupling strength values. To ensure simulation efficiency, coupling strengths between neurons that are identical and fixed with time are investigated initially. Such networks where the interaction strengths are fixed are referred to as homogeneously coupled. The capacity of controlling and altering behaviours produced by homogeneously coupled networks is assessed through the application of weak and strong delayed feedback independently with various time delays. To imitate learning, the coupling strengths later deviate from one another and evolve with time in networks that are referred to as heterogeneously coupled. The intensity of coupling strength fluctuations and the rate at which coupling strengths converge to a desired mean value are studied to determine their impact upon synchronisation performance. The stochastic delay differential equations governing the numerically simulated networks are then converted into a finite set of deterministic cumulant equations by virtue of the Gaussian approximation method. Cumulant equations for maximal and sub-maximal connectivity are used to generate two-parameter bifurcation diagrams on the noise intensity and coupling strength plane, which provides qualitative agreement with numerical simulations. Analysis of artificial brain networks, in respect to biological brain networks, are discussed in light of recent research in sleep theory.

L'épilepsie est une maladie fréquente affectant 0,5 à 1% de la population générale. Elle est caractérisée par des crises récurrentes responsables d'un sévère handicap médical et psychosocial. Les causes de l'épilepsie sont multiples et peuvent être liées notamment à des lésions cérébrales anténatales ou acquises, des causes génétiques ou métaboliques. L'épilepsie du lobe temporal (ELT) est la forme la plus répandue chez l’adulte, le plus souvent associée à une sclérose de l'hippocampe et réfractaire aux traitements antiépileptiques. Si pendant longtemps l'ELT a été perçue comme une pathologie focale centrée sur l'hippocampe sclérosé, de nombreux travaux montrent que les atteintes associées à l'ELT s'étendent bien au-delà de l'hippocampe et du lobe temporal, suggérant une altération plus globale du réseau cérébral structurel impactant le fonctionnement du cerveau. Toutefois, ces atteintes sont encore mal connues. Le développement récent des séquences et du traitement de l’imagerie de diffusion permettent l’acquisition d’images anatomiques du cerveau et la modélisation des fibres de substance blanche. L’architecture du réseau cérébral peut alors être représentée mathématiquement par un graphe, appelé « connectome » structurel, définissant la force des liens structurels (fibres de substance blanche) entre différentes régions du cerveau.L’objectif principal de la thèse est d’identifier les altérations du réseau structurel liées à l’épilepsie, avec un intérêt particulier à l’ELT. L’objectif secondaire est de développer de nouvelles méthodes d’extraction du connectome structurel pour en améliorer la précision anatomique et mieux identifier et localiser les altérations du réseau structurel.Ainsi, dans un premier temps, nous établissons l’état de l’art des méthodes d’extraction et d’analyse du connectome structurel et discutons leurs limites. Nous présentons alors une nouvelle méthode d’extraction du connectome structurel haute-résolution couvrant l’ensemble du cortex et incluant certaines régions sous-corticales, baptisée « high-resolution structural connectome ». L’objectif est de définir un cadre d’analyse du connectome structurel avec une très bonne précision anatomique et de fournir les outils nécessaires pour des études individuelles ou de groupe en tenant compte des contraintes de temps de calcul et d’utilisation de la mémoire et du disque.Dans un deuxième temps, nous analysons le connectome structurel de patients ELT avec sclérose hippocampique latéralisée dans le but de mettre en évidence le réseau structurel pathologique et d’en distinguer les caractéristiques en fonction de la latéralité de la lésion. Nous validons la stabilité et la reproductibilité du connectome structurel haute-résolution sur des sujets sains. La démonstration de son intérêt clinique potentiel est apportée en observant des différences structurelles subtiles entre deux groupes de sujets sains et en identifiant les sous-structures du striatum. Enfin, notre méthode est appliquée dans un contexte clinique pour identifier les altérations de connectivité structurelle du complexe hippocampo-amygdalien, impliqué dans l’ELT, en lien avec la pathologie. Nos travaux ont ainsi permis d’identifier les altérations globales et diffuses du réseau structurel liées à l’ELT, et plus particulièrement ont mis en évidence des disparités importantes selon la latéralité de la pathologie. Nous avons également présenté une nouvelle méthode d’extraction du connectome structurel augmentant considérablement sa précision anatomique et défini les outils nécessaires à l’analyse haute-résolution du connectome structurel. L’intérêt de cette méthode a été démontré par le gain de précision anatomique obtenu pour l’étude de l’architecture cérébrale du sujet sain ou pour une meilleure identification de réseaux pathologiques, ouvrant ainsi de nombreuses perspectives sur la caractérisation de l’architecture cérébrale et son lien sur le fonctionnement du cerveau. / Epilepsy is a frequent disease affecting 0.5 to 1% of the general population, characterized by recurrent seizures responsible for severe medical and psychosocial handicaps. The causes of epilepsy may be antenatal or acquired brain lesions, genetic history or metabolic disorders. Temporal lobe epilepsy (TLE) is the most common medically intractable epilepsy in adults, often associated with hippocampal sclerosis. Although TLE has been perceived for a long time as a hippocampal disorder, many studies show that the disease actually affects brain regions beyond the hippocampus and temporal lobe suggesting diffuse alteration of the brain structural network. However, these alterations are still unknown. Recent advances in diffusion weighted imaging and processing allow for the acquisition of brain anatomical images and the modeling of white matter fibers. Brain network architecture can then be represented mathematically by means of a graph, called “structural connectome”, defining the strength of the structural links (white matter fibers) across brain regions.The purpose of this thesis is to identify structural network alterations associated with epilepsy, in particular TLE. The secondary objective is to develop new methods for extracting the structural connectome in order to increase the anatomical accuracy and better localize network alterations.Therefore, we first review the state of the art of the methods used for extracting and analyzing the structural connectome and establish their limitations. We then introduce a new method to extract the structural connectome with increased anatomical accuracy, which we called “high-resolution structural connectome”. The purpose is to provide a framework to analyze brain connectivity at high-resolution and to define the necessary tools for individual and group analysis, keeping in mind processing time and memory and disk usages.Then, we analyze the structural connectome of TLE patients with hippocampal sclerosis to reveal underlying pathological network, we also highlight pathological network discrepancies between left and right sided lesions. Inter- and intra-subject stability and repeatability of the high-resolution structural connectome are assessed with a cohort of healthy subjects. We demonstrate potential clinical interest by observing subtle structural differences between two groups of healthy subjects and by delineating the sub-fields of the striatum. Finally, our method is applied to the pathological case of TLE and aims at uncover structural connectivity alterations of the hippocampo-amygdalian complex, known to be involved in TLE.In conclusion, we extend our current knowledge on TLE by showing that this is a network disease involving widespread brain regions, whose pattern largely depends on lesion laterality. We also introduce a new method for extracting the structural connectome at high-resolution, considerably increasing the anatomical accuracy. The interest of this method is demonstrated on healthy subjects to better characterize the healthy brain and on the diseased brain to localize more precisely the brain regions associated with the pathology.

Le concept de réseaux - l'idée que deux ou plusieurs nœuds distribués peuvent interagir pour produire un phénomène - a longtemps été utilisé dans la recherche et le traitement de l'épilepsie. En effet, même dans les épilepsies considérées comme «focales», les perspectives cliniques et théoriques soulignent l'importance des questions suivantes, à savoir : 1) comment pouvons-nous localiser, partitionner et caractériser les réseaux impliqués dans l'épilepsie et 2) dans quelle mesure ces réseaux interagissent avec le réseau cérébral à grande échelle? Récemment, la notion de réseaux pathologiques dans l'épilepsie a été renforcée par l’apport de la neuroimagerie, avec en particulier le paradigme 'd'état de repos' qui reconnaît l'information inhérente à l'activité spontanée du cerveau, en plus de celle liée aux événements transitoires exogènes et paroxystiques endogènes.En tirant partie de ces techniques, ce travail fournit de nouveaux concepts sur 1) les relations multimodales et le couplage entre l’hémodynamique et la connectivité fonctionnelle électro physiologique aussi bien dans les cortex épileptiques que non affectés, 2) les processus pathologiques affectant l’homéostasie ionique et les dysfonctionnements neuronaux dans les réseaux épileptiques, 3) les interactions au niveau de groupe entre les réseaux épileptiques et les propriétés topologiques du cerveau, et 4) comment les interactions entre la pathologie épileptique et des propriétés uniques du réseau cérébral peuvent contribuer à produire des effets cliniques au niveau du réseau. / The concept of networks – the idea that two or more distributed nodes may interact to produce a phenomenon – has long been of utility in research into and the treatment of epilepsy. Indeed, even in epilepsies deemed ‘focal’, clinical and theoretical insights underline the importance of the questions 1) how can we localize, partition and characterize networks involved in epilepsy, and 2) to what extent do such networks interact with the brain network at large? Recently, the notion of pathological network effects in epilepsy has been reinvigorated with input from neuroimaging, especially a ‘resting-state’ paradigm that recognizes the systemic information inherent in the ongoing activity of the brain in addition to that provided when it is disturbed by transient exogenous events and endogenous paroxysms. By leveraging these techniques, this work provides novel insights into 1) the multimodal relationships and coupling between haemodynamic- and electrophysiologically-defined functional connectivity, both in epileptic and unaffected cortices 2) pathological processes affecting ionic homeostasis and neural dysfunction in epileptic networks 3) group-level interactions between epileptic networks and brain network topological properties and 4) how interactions between epileptic pathology and unique brain network properties may contribute to produce to clinical effects at the network level. This work opens up new perspectives on the understanding of network effects in epilepsy, sources of variance in their analysis, the biological processes occurring in parallel and contributing to them and their role in an individualized understanding of pathology.

Les troubles du spectre autistique (TSA) sont des troubles neurodéveloppementaux fortement héritables. En parallèle, la théorie de l'hypoconnectivité fronto – postérieure semble être au coeur de la physiopathologie des TSA. Afin de comprendre la contribution des facteurs de risque familiaux de ce trouble, nous avons conduit conjointement une étude de connectivité anatomique et fonctionnelle chez des parents non atteints de sujets atteints de TSA. Nous avons réalisé une étude de tractographie en cerveau entier afin de comparer les valeurs de l'anisotropie fractionnelle généralisée le long des principaux faisceaux de substance blanche chez 85 sujets adultes : 39 parents non atteints, 18 sujets atteints de TSA comparés à 28 sujets contrôles. Après avoir corrigé pour les tests multiples, nous avons mis en évidence une diminution significative de l'anisotropie fractionnelle généralisée le long du faisceau fronto- occipital inférieur bilatéral chez les parents non atteints, les sujets atteints de TSA en comparaison aux sujets contrôles. Afin de comprendre l'implication fonctionnelle de la dysconnectivité anatomique fronto – occipitale retrouvée en tractographie, nous avons calculé la connectivité fonctionnelle entre les régions fronto – occipitales selon les extrémités du faisceau frontal - occipital inférieur bilatéral. En comparaison à 28 sujets contrôles, nous avons ainsi observé une augmentation significative de la connectivité fonctionnelle fronto - occipitale chez 38 parents non atteints et chez 13 sujets atteints de TSA. Une étude de connectivité fonctionnelle en cerveau entier serait une perspective prometteuse quant à l'interprétation de l'augmentation de la connectivité fronto – occipitale observée. Les anomalies fronto – occipitales montrées chez les parents non atteints, les sujets atteints de TSA pourraient correspondre à la mise en évidence d'un endophénotype dans les TSA. / Autism Spectrum Disorder (ASD) are neurodevelopmental disorders highly heritable.In parallel, the underconnectivity theory of ASD assumes that fronto-posterior brain disconnectivity is at the core of its pathophysiology. Our goal was to assess long-range structural and functional connectivity in unaffected parents of subjects with ASD to better understand the contributions of familial factors to heightened risk of ASD. We performed a diffusion weighted imaging (DWI) based whole brain tractography to compare generalized fractional anisotropy (gFA) in the main deep long white matter tracts in 85 adults: 39 unaffected parents, 18 probands compared to 28 controls. After corrections for multiple comparisons, we identified a significant decrease in gFA in the bilateral inferior frontal occipital fasciculus (IFOF) in both probands with ASD and unaffected parents when compared to controls. To understand the functional implication of fronto – occipital anatomical disconnectivity, we assessed the functional connectivity between the regions linked by IFOF exhibiting significant alterations in gFA. We also showed that both probands and unaffected parents exhibited a significantly increased functional connectivity between the frontal and occipital regions linked by the IFOF. In order to better understand and extend this interesting results, to evaluate the global functional connectivity of our sample in order to be able to interpret the increase of fronto-occipital functional connectivity would be an important perspective. These findings highlight an altered fronto-occipital connectivity in subjects with ASD and unaffected parents suggesting that fronto-occipital disconnectivity may be an endophenotype of ASD.

Since the early eighties researchers have been studying the use of technology that supports collaboration amongst co-workers and group members. This field of computer science became known as Computer Supported Cooperative Work (CSCW). With the advent of wireless and mobile Internet communication technologies research in the CSCW field has been focused on providing “access, anytime and anywhere”. The main contribution of this study is to introduce and analyze the technology required to support casually connected collaboration. Firstly, we define casually connected collaboration as having “access, anytime and anywhere” to collaborators and resources without having explicit control or knowledge over the environment and its technical abilities. In order to distinguish between connected, mobile, and casually connected collaboration we introduce a conceptual model of collaboration that extrapolates the term “access, anytime and anywhere”. We then aim to prove the soundness of our model by using it to classify some well known collaboration scenarios. Furthermore, by evaluating the functional and non-functional requirements for a casually connected collaboration solution, we argue that current commercial and CSCW research implementations do not sufficiently meet these demands. We then present Nomad: a Peer-to-Peer framework specifically designed to overcome the challenges encountered in casually connected collaboration. We study the technology requirements and highlight the implementation details that enabled us to successfully conform to the requirements set by casually connected collaboration. Finally, we pave the road for future work by investigating new features introduced into the Microsoft .NET Framework version 4.0, Visual Studio 2010 and language enhancements made to C# version 4.0. / Dissertation (MSc)--University of Pretoria, 2009. / Computer Science / unrestricted

This dissertation aims to challenge the existing built thresholds that are in place between South Africa and its neighbouring countries. In the words of Heidegger, “ A boundary is not that at which something stops, but as the Greeks recognised, the boundary is that from which something begins it’s prescencing.”¹ Boundaries and edges need to be created. They need to be made strong. They need to dominate and control. But borders must not be barriers. This study sets out to investigate the design of a border complex as a transition zone between two countries. It intends to encourage international connectivity, and encourage fair trade and travel, by addressing the functional requirements expectant of a border post, and ensuring the necessary control. Emphasis is also placed on movement through a site where many challenging physical constraints call for innovative design solutions, and where a harsh climate needs to be addressed. / Dissertation (MArch (Prof))--University of Pretoria, 2007. / Architecture / unrestricted

This work concerns the investigation of the neuronal mechanisms at the basis of language acquisition and processing, and the complex interactions of language and attention processes in the human brain. In particular, this research was motivated by two sets of existing neurophysiological data which cannot be reconciled on the basis of current psycholinguistic accounts: on the one hand, the N400, a robust index of lexico-semantic processing which emerges at around 400ms after stimulus onset in attention demanding tasks and is larger for senseless materials (meaningless pseudowords) than for matched meaningful stimuli (words); on the other, the more recent results on the Mismatch Negativity (MMN, latency 100-250ms), an early automatic brain response elicited under distraction which is larger to words than to pseudowords. We asked what the mechanisms underlying these differential neurophysiological responses may be, and whether attention and language processes could interact so as to produce the observed brain responses, having opposite magnitude and different latencies. We also asked questions about the functional nature and anatomical characteristics of the cortical representation of linguistic elements. These questions were addressed by combining neurocomputational techniques and neuroimaging (magneto-encephalography, MEG) experimental methods. Firstly, a neurobiologically realistic neural-network model composed of neuron-like elements (graded response units) was implemented, which closely replicates the neuroanatomical and connectivity features of the main areas of the left perisylvian cortex involved in spoken language processing (i.e., the areas controlling speech output – left inferior-prefrontal cortex, including Broca’s area – and the main sensory input – auditory – areas, located in the left superior-temporal lobe, including Wernicke’s area). Secondly, the model was used to simulate early word acquisition processes by means of a Hebbian correlation learning rule (which reflects known synaptic plasticity mechanisms of the neocortex). The network was “taught” to associate pairs of auditory and articulatory activation patterns, simulating activity due to perception and production of the same speech sound: as a result, neuronal word representations distributed over the different cortical areas of the model emerged. Thirdly, the network was stimulated, in its “auditory cortex”, with either one of the words it had learned, or new, unfamiliar pseudoword patterns, while the availability of attentional resources was modulated by changing the level of non-specific, global cortical inhibition. In this way, the model was able to replicate both the MMN and N400 brain responses by means of a single set of neuroscientifically grounded principles, providing the first mechanistic account, at the cortical-circuit level, for these data. Finally, in order to verify the neurophysiological validity of the model, its crucial predictions were tested in a novel MEG experiment investigating how attention processes modulate event-related brain responses to speech stimuli. Neurophysiological responses to the same words and pseudowords were recorded while the same subjects were asked to attend to the spoken input or ignore it. The experimental results confirmed the model’s predictions; in particular, profound variability of magnetic brain responses to pseudowords but relative stability of activation to words as a function of attention emerged. While the results of the simulations demonstrated that distributed cortical representations for words can spontaneously emerge in the cortex as a result of neuroanatomical structure and synaptic plasticity, the experimental results confirm the validity of the model and provide evidence in support of the existence of such memory circuits in the brain. This work is a first step towards a mechanistic account of cognition in which the basic atoms of cognitive processing (e.g., words, objects, faces) are represented in the brain as discrete and distributed action-perception networks that behave as closed, independent systems.

Conservation is plagued by the issue of prioritization - what to conserve and where to conserve it - which relies on identification and assessment of risks. In this body of work, I identify some of the risks related to climate change impacts on biodiversity, as well as potential solutions. Climate changes are underway across nearly all terrestrial areas and will continue in response to greenhouse gas emissions over centuries. Other extinction drivers, such as habitat loss due to urbanization, commonly operate over localized areas. Urbanization contributes, at most, less than 2% of the total range loss for terrestrial species at risk when averaged within an ecodistrict (Chapter 2). Documented impacts of climate change, to date, include: extinction, population loss, reduction in range area, and decreased abundance for multiple taxonomic groups. Examining species’ and populations’ physiological limits provides insight into the mechanistic basis, as well as geography, of climate change impacts (Chapter 3). Climate changes, and the ecological impacts of climate changes, are scale-dependent. Thus, the biotic implications are more accurately assessed through comparisons of local impacts for populations. Under a scenario of climate change, equatorward margins may be strongly limited by climatic conditions and not by biotic interactions. Yet, geographic responses at poleward margins do not appear directly linked to changes in breeding season temperature (Chapter 4). New ideas on how regions with attenuated climate change (climate refugia) may be used to lower species climate-related extinction risk while simultaneously improving habitat connectivity should be considered in the context of potential future consequences (i.e. range disjunction, alternative biological responses) (Chapter 5). Contemporary climate refugia are identifiable along multiple climatic dimensions, and are similar in size to current protected areas (Chapter 6). Determining how, when, and where species distributions are displaced by climate change as well as methods of reducing climatic displacement involves integrating knowledge from distribution shift rates for populations, occurrence of climate refugia, and dispersal barriers. Such assessments, in the Yellowstone to Yukon region, identify dramatically different pathways for connectivity than assessments that are not informed by considerations of species richness and mobility (Chapter 7).

La fragmentation des habitats est aujourd’hui considérée comme l’une des plus sérieuses menaces sur la biodiversité, et une des causes majeures de la crise actuelle d’extinction des espèces. Dans les paysages fragmentés, le mouvement est un processus clé pour la survie des espèces faunistiques et floristiques, celui-ci étant facilité par une connectivité importante entre les éléments du paysage. Parmi les éléments qui concourent à augmenter la connectivité du paysage, les corridors écologiques sont particulièrement intéressants et efficaces. Le principal objectif de cette thèse est d’évaluer l’intérêt des images THRS pour identifier et caractériser les éléments du paysage agricole potentiellement constitutifs des corridors écologiques. Pour cela, nous avons évalué le potentiel des images RSO et de l’utilisation combinée des données optiques et radar THRS pour caractériser des continuités écologiques à échelle fine, tant d’un point de vue structurel que fonctionnel. Les résultats montrent que les images radar et plus particulièrement un indicateur polarimétrique qui en est dérivé, l’entropie de Shannon, permettent de quantifier le degré d’ouverture de la canopée des haies, indicateur qui a pu être évalué d’un point de vue écologique. De plus, ils ont permis de mettre en évidence les indicateurs, les méthodes de classification et les dates clefs pour la cartographie des zones humides. Une évaluation de la synergie d’images optiques et radar à permis de montrer que les données radar permettent de cartographie des cultures. Nous avons également montré l’apport des séries temporelles d’images radar pour l’amélioration de modèles biologiques visant à expliquer la distribution de la biodiversité dans les paysages agricoles. Enfin, nous avons pu développer de nouvelles métriques de connectivité en intégrant des informations dérivées de données de télédétection qui se sont avérées pertinentes d’un point de vue écologique. / Habitat fragmentation is now considered as one of the most serious threats to biodiversity, and a major cause of the current crisis of species extinction. In fragmented landscapes, movement is a key process for the survival of flora and fauna species. Movement is facilitated by a significant connectivity between landscape elements. Among elements that contribute to increase landscape connectivity, ecological corridors are particularly useful and effective. The main objective of this pHD is to evaluate the interest of the VHSR images to identify and characterize potential agricultural landscape elements of the ecological networks. We evaluated the potential of SAR images and the combined use of VHSR optical and radar data to characterize ecological continuities at a fine scale, both structural and functional points of view. Results show that radar images and especially a polarimetric indicator, the Shannon entropy, allow quantifying the canopy cover of hedges. This indicator was assessed from an ecologically point of view. In addition, we highlight the indicators, classification methods and key dates to map wetland vegetation types. An evaluation of the synergy of radar and optical images shows the usefulness of radar data for crops mapping. We also showed the contribution of radar time series for improving biological models that explain the biodiversity distribution in agricultural landscapes. Finally, we developed new connectivity metrics, which relevant from an ecological point of view, by integrating information derived from remote sensing data

Le cortex insulaire est le cinquième lobe du cerveau en charge de l'intégration de nombreuses fonctions cognitives, sous-tendues par une organisation cytoarchitectonique etune connectivité aussi riche que complexe. Ce travail vise à évaluer la connectivité fonctionnelle insulaire du cerveau humain par le biais de stimulation électrique intra-cérébrale et de potentiels évoqués cortico-corticaux (PECC) réalisés chez des patients explorés en stéréoélectroencéphalographie (SEEG) pour une épilepsie partielle réfractaire. Nous avons développé un protocole automatisé permettant destimuler successivement l’ensemble des bipoles d’enregistrement intracérébraux (deux plots contigus d’une même électrode) disponibles chez les patients explorés en SEEG. Deux sériesde 20 stimulations monophasiques d’une durée unitaire de 1 ms et d’une intentisté de 1 mA, étaient délivrés à une fréquence de 0,2 Hz au niveau de chaque bipole (105 en moyenne,produisant un total d’environ 11.000 PECC par patient). Un premier travail a consisté dans lamise au point d’une méthode fiable d’analyse statistique objective des PECC significatifs, encomplement de l’analyse visuelle, sur un échantillon de 33017 enregistrements chez trois patients. L’analyse a porté sur les quatre fenêtres temporelles post-stimulation suivantes: 10-100 ms, 100-300 ms, 300-500 ms, 500-1000 ms. La seconde partie de notre thèse a appliquéces méthodes à l’étude des connections intra-insulaires sur un échantillon de10 patients présentant au moins deux éléctrodes intra-insulaires. La dernière partie de notre travail s’est intéressé aux efférences insulaires sur un échantillon de 11 patients. L’étude des PECC apporte des éléments de connectivité fonctionnelle derésolution spatiale et temporelle inégalée, complémentaires de ceux découlant des techniquesde neuroimagerie. La gestion complexe du volume de données à gérer pour chaque patientpeut être résolu par des procédures d’analyse statistiques automatisée de sensibilité etspécificité satisfaisante. Le pattern des connections intra- et extra-insulaires révélé par cetteapproche permet une meilleure compréhension de la physiologie de l’insula chez l’Homme etdes modalités de propagations des décharges épileptiques impliquant ce lobe. / The insular cortex is the fifth lobe of the brain and is in charge of the integration of many cognitive functions, underpinned by a rich cytoarchitectonic organization and a complex connectivity. Our work aims to evaluate the insular functional connectivity of the human brain using intracerebral electrical stimulation and recording of cortico-cortical evoked potentials (CCEPs) in patients investigated with stereoelectroencephalography (SEEG) for refractory partial epilepsy. We first developed an automated protocol to stimulate successively all intracerebral recorded bipoles (two contiguous leads of the same electrode) available in patients undergoing SEEG. Two sets of 20 monophasic stimulation of 1 ms duration and 1mA intensity were delivered at a frequency of 0.2 Hz at each bipole (105 on average, producing a total of about 11,000 recordings per patient). We then develop a reliable and objective statistical method to detect significant CCEPs as a complement to visual analysis, and validate this approach on a sample of 33017 recordings in three patients. The analysis was performed over four distinct post-stimulus epochs: 10-100 ms, 100-300 ms, 300-500 ms, 500-1000 ms. In the second part of our thesis, we applied these methods to the study of intrainsular connections on a sample of 10 patients with at least two intra-insular electrodes. The last part of our work used the same approach to investigate insular efferents in a sample of 11 patients. The study of CCEPs provides novel and important findings regarding the human brain functional connectivity, with unmatched spatial and temporal resolutions as compared to neuroimaging techniques. The complex management of large volume of data in each patient can be solved by automated statistical analysis procedures with satisfactory sensitivity and specificity. The pattern of connections within and outside the insula revealed by this approach provides a better understanding of the physiology of the Human insula as well as of the propagation of epileptic discharges involving this lobe.