La perception visuelle des visages est une fonction importante du cerveau humain, essentielle pour les interactions sociales. L’étude des bases neurales de la perception des visages a débuté il y a plusieurs décennies et les découvertes servent de modèle pour la compréhension de la perception visuelle en général. L’imagerie structurelle des patients présentant un déficit de reconnaissance des visages à la suite d’une lésion cérébrale a montré l’importance d’un vaste territoire au sein du cortex ventral occipito-temporal (VOTC), du lobe occipital jusqu’au lobe temporal antérieur (LTA), avec une prédominance droite. L’imagerie par résonnance magnétique fonctionnelle (IRMf) a montré l’existence de zones cérébrales circonscrites qui répondent plus fortement aux visages qu’aux autres objets visuels (organisation en « cluster ») principalement dans le VOTC postérieur. Cependant l’IRMf a été limitée dans sa capacité à retrouver de telles régions dans le lobe temporal antérieur à cause d’artefacts méthodologiques. Les études d’électro-encéphalographie intracrânienne (iEEG) réalisées chez les patients épileptiques sont une opportunité unique d’enregistrer l’activité neuronale directe avec un très haut rapport signal/bruit. Les études iEEG ont enregistré des réponses sélectives aux visages largement distribuées dans le VOTC, sans organisation en « clusters ». Malgré des années de recherches, plusieurs questions cruciales restent sans réponse : (1) quelle est l’organisation spatiale des régions sélectives aux visages (organisation distribuée vs. en « clusters ») ? ; (2) quelles sont les bases neurales de la perception des visages dans le LTA ? ; (3) quelles sont les régions critiques pour la perception des visages ? Afin de répondre à ces questions, nous avons utilisé les enregistrements et les stimulations électriques intracérébraux. Dans une 1ère étude (Jonas et al., sous presse), nous avons combiné les enregistrements iEEG avec la stimulation visuelle périodique rapide (FPVS). La méthode FPVS est basée sur le principe suivant : présenter des stimuli visuels à une fréquence fixe va générer une réponse EEG périodique à la même fréquence. Nous avons utilisé cette approche pour réaliser une cartographie complète des réponses sélectives aux visages dans le VOTC (28 participants). Nous leur avons montré des séquences d’images d’objets présentées à une fréquence fixe et rapide (6 Hz), avec un visage présenté tous les 5 objets (1,2 Hz). Les réponses sélectives aux visages ont été identifiées objectivement (à la fréquence de stimulation) et quantifiées dans tout le VOTC. Bien que ces réponses aient été enregistrées de manière largement distribuée, nous avons identifié plusieurs régions dans les lesquelles les réponses les plus fortes se regroupent spatialement (en « clusters »). De plus, nous avons enregistré la plus forte réponse dans le gyrus fusiforme droit. Enfin, nous avons enregistré des réponses sélectives aux visages dans 3 régions distinctes du LTA. Dans 3 autres études, nous rapportons de très rares cas de stimulations électriques de régions sélectives aux visages, testant leur rôle critique dans la perception des visages. Nous rapportons un cas de déficit transitoire de la perception des visages après stimulation du gyrus occipital inférieur droit, la région sélective au visage la plus postérieure (Jonas et al., 2012, 2014) et un cas similaire après stimulation du LTA (Jonas et al., 2015). Dans l’ensemble, ces études montrent que : (1) les régions impliquées dans la perception des visages sont largement distribuées le long du VOTC et certaines sont marquées par un regroupement spatial de leurs réponses les plus fortes ; (2) plusieurs régions distinctes sont sélectives aux visages dans le LTA; (3) des régions spécifiques dans le VOTC postérieur et le LTA sont critiques pour la perception des visages. Ces études montrent l’intérêt des enregistrements intracérébraux pour la compréhension des mécanismes de perception visuelle / Visual perception of faces is a primary function of the human brain, critical for social interactions. The neural basis of face perception in humans has been investigated extensively for decades as a primary research goal, whose findings may serve as a rich model for understanding perceptual recognition. Structural imaging of individuals with face recognition impairment following brain damage point to a large territory of the human ventral occipito-temporal cortex (VOTC), from the occipital lobe to the anterior temporal lobe (ATL), with a right hemispheric advantage. Functional magnetic resonance imaging studies (fMRI) have reported face-selective responses (larger responses to faces than other visual objects) in circumscribed regions (clustered organization) of the posterior VOTC. However, they failed to report genuine responses in the ATL because of methodological artefacts. Intracranial electroencephalographic (iEEG) recordings performed in epileptic patients offer a unique opportunity to measure direct local neural activity with a very high signal-to-noise ratio. In contrast to fMRI studies, iEEG studies recorded face-selective responses in widely distributed regions of the VOTC without any evidence of a clustered organization. Despite decades of research, several outstanding questions are still unanswered: (1) what is the spatial organization of brain regions supporting face perception (clustered vs. distributed)?; (2) what are the neural basis of face perception in the ATL?; (3) which are the critical regions for face perception? To address these gaps in knowledge, we used human iEEG recordings and electrical intracerebral stimulations. In a first study (Jonas et al., in press), we combined iEEG recordings with the Fast Periodic Visual Stimulation (FPVS), a powerful approach providing objective and high signal-to-noise brain responses. FPVS is based on the simple principle: presenting visual stimuli at a fixed rate generates a periodic EEG response at exactly the same frequency. We use this approach to report a comprehensive map of face-selective responses across the VOTC in a large group of participants (N=28). They were presented with natural images of objects at a rapid fixed rate (6 images per second: 6 Hz), with face stimuli interleaved as every 5th stimulus (i.e., 1.2 Hz). Face-selective responses were objectively (i.e., exactly at the face stimulation frequency) identified and quantified throughout the whole VOTC. Although face-selective responses were widely distributed, specific regions displayed a clustered spatial organization of their most face-selective responses. Among these regions, the right fusiform gyrus showed the largest face-selective response. In addition, we recorded face-selective responses in 3 distinct regions of the ATL. In 3 others studies, we reported very rare cases of intracerebral electrical stimulation of face-selective brain regions, testing the critical role of these regions in face perception. We reported a case of transient inability to recognize faces following the stimulation of the right inferior occipital gyrus, the most posterior face-selective region (Jonas et al., 2012, 2014) and one similar case following the stimulation of the right ATL (Jonas et al., 2015). Overall, these studies show that: (1) face-selective responses are widely distributed but some specific regions displayed a clustered spatial organization of their most face-selective responses; (2) several distinct regions are face-selective in the ATL; (3) specific brain regions in the posterior VOTC and in the ATL are critical for face perception. These finding also illustrate the diagnostic value of intracerebral electrophysiological recordings in understanding visual recognition processes

Visuell-räumliche Aufmerksamkeit spielt eine Schlüsselrolle in der Wortverarbeitung und der Augenbewegungskontrolle beim Lesen. Dennoch bleibt die räumliche und zeitliche Dynamik von Aufmerksamkeitsbereitstellung innerhalb einer Fixation ungeklärt. Die vorliegende Dissertation nutzt drei Ansätze in der Ko-Registrierung von Augenbewegungen und EEG um diese Forschungslücke zu untersuchen und direkte Einblicke in die Aufmerksamkeitsverteilung zwischen Fovea und Parafovea, ihre Anpassung an Verarbeitungsansprüche und Sakkaden sowie ihre Auswirkungen auf Wortverarbeitung zu liefern. In Studie 1 wurde das Probe Paradigma als Maß zur Abbildung räumlicher Aufmerksamkeit ohne Augenbewegungen implementiert. Die vergrößerte, Probe-bezogene N1 zwischen Wortpräsentationen deutet auf eine Bereitstellung und Umverteilung zusätzlicher Ressourcen zur simultanen fovealen und parafovealen Wortverarbeitung hin. Studie 2 untersuchte den Preview Benefit und seine Modulation durch Sakkaden und Verarbeitungsschwierigkeit und zielte darauf ab, das elektrophysiologische Korrelat als indirektes Maß für parafoveale Aufmerksamkeit in ERPs und FRPs zu etablieren. Es zeigte sich ein Effekt des orthographischen Previews auf die N1. Interaktionen mit Lesemodus und Verarbeitungsschwierigkeit implizieren, dass dem Preview Effekt Aufmerksamkeitseffekte zugrunde liegen. In Studie 3 wurden zwei Datensätze hinsichtlich der Lateralisierung von Oszillationen im Alpha-Band untersucht um die in Studie 2 generierten Hypothesen zu unterstützen. Alpha war im sakkadischen Lesen stärker rechts-lateralisiert als in RSVP und eine stärkere Lateralisierung sagte eine kürzere Fixationsdauer vorher, was die Rolle von parafovealer Aufmerksamkeit und ihre Verbindung zu Sakkaden betont. Trotz der Einschränkungen der drei Ansätze stellt die Kombination von Augenbewegungen, ERPs, FRPs und EEG-Oszillationen geeignete Maße für Aufmerksamkeitsprozesse in der Wortverarbeitung dar, die traditionelle Forschungsmethoden ergänzen. / Visuo-spatial attention is a key contributor to word processing and oculomotor control in reading. Yet, the spatial and temporal dynamics of attention allocation within a fixation remain unclear. The present dissertation uses three approaches in the co-registration of eye movements and EEG to investigate this gap and provide direct, online insights into attention distribution across fovea and parafovea, its adaptation to processing load and saccadic behavior, as well as its effects on word processing. In Study 1, the probe paradigm was implemented as a mapping tool of spatial attention adaptation in the absence of eye movements, yielding ERPs. The enhanced probe-related N1 in-between trials indicates a recruitment and redistribution of additional resources to achieve simultaneous foveal and parafoveal word-recognition. Study 2 tested the preview benefit and its modulation by saccades and foveal load, thereby aiming at establishing its electrophysiological correlate as an indirect index of parafoveal attention allocation in ERPs and FRPs. There was an effect of orthographic preview on the N1. Interactions with reading mode and foveal load indicate underlying attention effects. In Study 3, two datasets were reanalyzed with regard to the lateralization of oscillatory activity in the alpha-band in order to directly support the assumptions generated in Study 2. Alpha was more strongly right-lateralized in saccadic reading compared to RSVP, and moment-to-moment lateralization predicted shorter subsequent fixation duration, emphasizing the role of parafoveal attention allocation and its relation to saccades. Despite the limitations of the three approaches at this point, the combination of eye movements, ERPs, FRPs, and EEG oscillations provides suitable online markers of attention processes in word recognition that complement traditional research methods.

Teil I dieser Arbeit untersucht Effekte einer semantischen Eigenschaft von syntaktisch transitiven Prädikaten auf die Satzverarbeitung: den Grad der Affiziertheit, den solch ein Prädikat für eines seiner Argumente impliziert, also das Ausmaß in dem ein Teilnehmer eine Zustandsveränderung während des ausgedrückten Ereignisses erfährt. Drei Experimente (zu Akzeptabilitätsurteilen, Lesezeiten & ereigniskorrelierten Potentialen [EKP]) überprüften den Einfluss von Affiziertheit auf die lexikalisch-semantische Verarbeitung des Prädikats und die Integration folgender Argument-Nominalphrasen (NPn) beim Satzlesen. Um den Effekt von Affiziertheit auf Verarbeitungsprozesse beim Argumentlinking zu maximieren, machten die Experimente Gebrauch von deverbalen, eventiven, deutschen -ung-Nominalisierungen, die von Verben mit verschiedenem Affiziertheitsgrad abgeleitet sind. Hierbei wurde der Einfluss der Affiziertheit ausgenutzt, folgende Objekts- oder Subjektsgenitive als Argument zu lizensieren. Während sich keine klaren Effekte bei der Verarbeitung der Nominalisierungen ergaben, zeigten die Ergebnisse zur Integration der Genitiv-NPn eine spezifische Interaktion zwischen dem Grad der Affiziertheit und der Akzeptabilität, einen Objekts- bzw. Subjektsgenitiv zu verwirklichen. Dieser Befund wurde von übereinstimmenden Lesezeiten- und EKP-Interaktionseffekten begleitet. Die EKP-Ergebnisse legen eine zentrale Rolle von zwei späten Positivierungen bei der Integration der Genitivargumente nahe. Diese Befunde werden im Kontext theoretischer Modelle von abgestufter Affiziertheit und von Satzverarbeitungsmodellen diskutiert, mit Fokus auf der möglichen Rolle der Prototypikalität von Argumenten und Unterschieden zwischen prädikatsinduzierter und argumentinhärenter semantischer Argumentprominenz. Teil II stellt zwei Softwarepakete für die statistische Umgebung R vor, die Schnittstellen für die zur Analyse der EKP-Daten verwendeten WFMM Software (Morris & Carroll, 2006) bieten. / Part I of this thesis investigates the effects of a semantic feature of syntactically transitive predicates on online sentence processing: the degree of affectedness such a predicate implies for one of its arguments, taken to indicate the extent to which an event participant undergoes a change of state during the event expressed. Three experiments (using acceptability judgements, reading times & event-related potentials [ERPs]) probed the effects of affectedness on the processing of the predicate and on the integration of following argument noun phrases (NPs), thus assessing the impact of affectedness on lexical-semantic processing and its influence on processes related to online argument linking. Maximising the effects of affectedness on linking-related processes, the experiments used German -ung nominalisations derived from verbs implying different degrees of affectedness, exploiting a paradigm involving the linking of either object or subject genitive argument NPs to deverbal, eventive nominalisations. While no clear effects of affectedness emerged for the processing of the nominalisations, results related to the integration of genitive NPs revealed a specific interaction between the degree of affectedness and the acceptability of realising either object or subject genitives. This pattern was accompanied by consistent reading time and ERP interaction effects. ERP results suggest a prominent role of two late positivities with different spatial foci in the integration of the genitive arguments. These findings are discussed in the context of theoretical models of graded affectedness and sentence processing models, considering possible roles of argument prototypicality and differences between predicate-induced and argument-inherent semantic argument prominence. Part II introduces two software packages for the statistical platform R, offering interfaces for the WFMM software (Morris & Carroll, 2006) used for the analysis of the ERP data in part I.

Lesen zielt darauf ab, Bedeutung aus geschriebenem Text zu extrahieren. Interessanterweise unterscheiden sich Wörter beträchtlich hinsichtlich der Menge mit ihnen assoziierter Bedeutung, und es wurde kürzlich gezeigt, dass eine hohe Bedeutungshaltigkeit lexikalische und semantische Aufgaben erleichtert. Die vorliegende Dissertation kombiniert ereigniskorrelierte Potentiale (EKPs) und konnektionistische Modellierung, um einige offene Fragen zur Rolle der Bedeutungshaltigkeit bei der Wortverarbeitung anzugehen. Hierbei wurden EKPs verwendet, um den Zeitverlauf unabhängiger Einflüsse der Anzahl semantischer Merkmale und Assoziationen beim Wortlesen zu bestimmen sowie Einflüsse von Bedeutungshaltigkeit auf implizites Wortlernen zu untersuchen. Um die zugrundeliegenden Mechanismen besser zu verstehen, wurden die Ergebnisse anschließend mittels eines semantischen Netzwerk-Modells simuliert. Es zeigten sich keine Einflüsse der Anzahl der Assoziationen, aber eine schnelle Aktivierung semantischer Merkmale, die das EKP bereits ab 190 ms beeinflussten - nur 20 bis 30 ms nach und zeitlich überlappend mit der Aktivierung orthographischer Repräsentationen, die durch N1-Lexikalitätseffekte angezeigt wurden. Im weiteren Verlauf ging eine hohe Merkmalsanzahl mit größeren N400-Amplituden einher. Zudem verstärkten semantische Merkmale Wiederholungseinflüsse auf die Akkuratheit lexikalischer Entscheidungen und N400-Amplituden, was einen ersten Hinweis auf Einflüsse von Bedeutungshaltigkeit auf implizites Wortlernen darstellt. Diese Ergebnisse stehen im Einklang mit merkmalsbasierten semantischen Netzwerk-Modellen. Simulationen legen nahe, dass semantische Aktivierung lexikalische Entscheidungen erleichtert, während Netzwerk-Fehler in engem Zusammenhang mit N400-Amplituden stehen. Da Netzwerk-Fehler psychologisch als implizite Vorhersagefehler interpretiert werden, deuten diese Ergebnisse darauf hin, dass N400-Amplituden implizite Vorhersagefehler im semantischen System widerspiegeln. / Language ultimately aims to convey meaning. Importantly, the amount of associated semantic information varies considerably between words. Recent evidence suggests that the richness of semantic representations facilitates performance in lexical and semantic tasks, but much remains to be learned about semantic richness effects. The present dissertation combined event-related brain potentials (ERPs) and connectionist modeling to address several unresolved issues concerning the role of semantic richness in word processing. Specifically, ERPs were employed to investigate the time course of independent influences of the number of semantic features and associates during word reading (study 1) and influences of semantic richness on implicit word learning (study 2). Aiming at advancing a mechanistic understanding of the obtained results, both studies were subsequently simulated using a network model of semantic cognition (study 3). Results showed no influences of the number of associates, but fast access to semantic features, with influences of feature-based semantic richness starting at about 190 ms - a mere 20 to 30 ms after and temporally overlapping with the activation of orthographic representations as reflected by N1 lexicality effects. Later on, a high number of semantic features induced larger N400 amplitudes. Furthermore, the number of semantic features enhanced repetition priming effects on lexical decision accuracy and N400 amplitudes, providing initial evidence for influences of semantic richness on implicit word learning. These results are in line with feature-based network models of semantic cognition. Simulations with such a model suggest that semantic activation can facilitate lexical decisions, while network error closely corresponds to N400 amplitudes. In psychological terms, network error has been conceptualized as implicit prediction error. Thus, these results are taken to suggest that N400 amplitudes reflect implicit prediction error in semantic memory.

INTRODUÇÃO: O transtorno fonológico é uma alteração na fala que ocasiona um grau variável de inteligibilidade de fala. A gravidade deste transtorno pode ser medida por meio do Percentage of Consonants Correct-Revised, que verifica o número de consoantes corretas em uma amostra de fala em relação ao número total de consoantes desta amostra. A literatura relata sete subtipos do transtorno fonológico, dentre eles a relacionada à otite média com efusão. Os Potenciais Evocados Auditivos também são utilizados para avaliar indivíduos com alteração de linguagem, pois esta população pode apresentar um déficit na via auditiva central. OBJETIVO: caracterizar os resultados dos potenciais evocados auditivos de longa latência N1, P2, N2 e P300 obtidos em crianças com transtorno fonológico, e verificar a evolução dos resultados destes potenciais frente à terapia fonoaudiológica, correlacionando esta evolução ao histórico de otite e a gravidade deste transtorno. MÉTODOS: Participaram da pesquisa 66 crianças, 25 sem transtorno fonológico (grupo controle) e 41 com transtorno fonológico (grupo estudo). As crianças do grupo estudo foram divididas em dois subgrupos: 22 formaram o subgrupo estudo A e 19 formaram o subgrupo estudo B. Todas as crianças foram submetidas à avaliação audiológica básica e aos potencias evocados auditivos de longa latência. Após a avaliação audiológica completa, as crianças do subgrupo estudo A foram submetidas a 12 sessões de terapia fonoaudiológica e reavaliadas audiologicamente após este período e, as crianças do subgrupo estudo B, crianças que aguardavam terapia em fila de espera, foram reavaliadas após três meses da avaliação inicial. RESULTADOS: os resultados demonstraram, na análise dos dados quantitativos, que o grupo estudo, antes da terapia fonoaudiológica, apresentou diferença estatisticamente significante para as latências dos componentes P2 e P300 e para a amplitude do P300, quando comparado com o grupo controle. Na comparação das latências não foram observadas diferenças estatisticamente significantes entre a primeira e segunda avaliações audiológicas, tanto para o subgrupo estudo A como para o subgrupo estudo B. Verificou-se diferença estatisticamente significante para as amplitudes do P300 e do P2/N2 na comparação entre a primeira e segunda avaliações audiológicas para os subgrupos estudo A e B, respectivamente. Na análise dos dados qualitativos, verificou-se que o P300 apresentou maior porcentagem de resultados alterados no grupo estudo quando comparado ao grupo controle, sendo que o tipo de alteração mais freqüentemente encontrado foi o aumento de latência. Após terapia fonoaudiológica, observou-se maior ocorrência de melhora nos resultados para todos os componentes estudados. Os resultados demonstraram, também, que não existiu associação entre a evolução dos resultados dos potenciais evocados auditivos de longa latência e o histórico de otite, bem como correlação com o Percentage of Consonants Correct-Revised. CONCLUSÕES: crianças com transtorno fonológico apresentam alterações no P300, sugerindo comprometimento da via auditiva central, provavelmente decorrente de alteração no processamento auditivo, apresentando melhora nos resultados de todos os componentes dos potenciais evocados auditivos de longa latência frente à terapia fonoaudiológica. Não existe associação entre a evolução dos resultados e histórico de otite, bem como correlação entre a evolução dos resultados e Percentage of Consonants Correct-Revised. / INTRODUCTION: Phonological disorder is a speech disorder that causes a varying degree of speech intelligibility. The severity of this disorder may be measured by the Percentage of Consonants Corrects-Revised, which verifies the number of correct consonants in a speech sample in relation to the total number of consonants in the sample. Literature reports seven subtypes of phonological disorder, amongst them the one related to otitis media with effusion. Auditory evoked potentials are also used to assess individuals with language disorder since this population may present deficit in the central auditory pathway. AIM: to characterize the long latency auditory evoked potentials results N1, P2, N2 and P300 of children with phonological disorder and to verify the improvement of such potentials results with the speech therapy, correlating this improvement to the background of otitis and the severity of this disorder. METHODS: Sixty six children took part in this study, 25 without phonological disorder (control group) and 41 with phonological disorder (study group). Children of the study group were divided into two subgroups: study subgroup A, composed by 22 children and study subgroup B composed by 19 children. All children underwent a basic audiological evaluation and long latency auditory evoked potentials. After the complete audiological assessment, children from study subgroup A underwent 12 sessions of speech therapy and were audiologically re-assessed after this period; children from study subgroup B were re-assessed three months after the initial assessment. RESULTS: the analysis of quantitative data revealed that the study group presented significant statistical difference, before the speech therapy, for the latencies of components P2 and P300 and for the amplitude of P300 when compared to the control group. Comparing the latencies, no significant statistical differences were observed between the first and the second audiological evaluations, either for the study subgroup A or B. A significant statistical difference was verified for the amplitudes of P300 and P2/N2 in the comparison between the first and the second audiological evaluations for subgroups A and B respectively. The analysis of qualitative data revealed that the P300 presented higher percentage of altered results in the study group when compared to the control group, and the most frequent type of alteration found was increased latency. After the speech therapy, the results of all components analyzed improved. Results also showed that there was no association between the improvement of long latency auditory evoked potentials results with the background of otitis, as well as with the Percentage of Consonants Corrects-Revised. CONCLUSIONS: children with phonological disorder present altered P300 suggesting involvement of the central auditory pathway probably due to alterations in the auditory processing, presenting improvement in all components of long latency auditory evoked potentials results after speech therapy. There is no association between the improvement of results and the background of otitis, as well as no correlation between the improvement of results and the Percentage of Consonants Corrects-Revised.

L'étude par neuro-imagerie de l'activité de plusieurs cerveaux en interaction (hyperscanning) permet d'étendre notre compréhension des neurosciences sociales. Nous proposons un cadre pour l'hyperscanning utilisant les interfaces cerveau-ordinateur multi-utilisateur qui inclut différents paradigmes sociaux tels que la coopération ou la compétition. Les travaux de cette thèse comportent trois contributions interdépendantes. Notre première contribution est le développement d'une plateforme expérimentale sous la forme d'un jeu vidéo multijoueur, nommé Brain Invaders 2, contrôlé par la classification de potentiels évoqués visuels enregistrés par électroencéphalographie (EEG). Cette plateforme est validée par deux protocoles expérimentaux comprenant dix-neuf et vingt-deux paires de sujets et utilise différentes approches de classification adaptative par géométrie riemannienne. Ces approches sont théoriquement et expérimentalement comparées et nous montrons la supériorité de la fusion des classifieurs indépendants sur la classification d'un hypercerveau durant la seconde contribution. L'analyse de coïncidence des signaux entre les individus est une approche classique pour l'hyperscanning, elle est pourtant difficile quand les signaux EEG concernés sont transitoires avec une grande variabilité (intra- et inter-sujet) spatio-temporelle et avec un faible rapport signal-à-bruit. En troisième contribution, nous proposons un nouveau modèle composite de séparation aveugle de sources physiologiquement plausibles permettant de compenser cette variabilité. Une solution par diagonalisation conjointe approchée est proposée avec une implémentation d'un algorithme de type Jacobi. A partir des données de Brain Invaders 2, nous montrons que cette solution permet d'extraire simultanément des sources d'artéfacts, des sources d'EEG évoquées et des sources d'EEG continues avec plus de robustesse et de précision que les modèles existants. / The study of several brains interacting (hyperscanning) with neuroimagery allows to extend our understanding of social neurosciences. We propose a framework for hyperscanning using multi-user Brain-Computer Interfaces (BCI) that includes several social paradigms such as cooperation or competition. This dissertation includes three interdependent contribution. The first contribution is the development of an experimental platform consisting of a multi-player video game, namely Brain Invaders 2, controlled by classification of visual event related potentials (ERP) recorded by electroencephalography (EEG). The plateform is validated through two experimental protocols including nineteen and twenty two pairs of subjects while using different adaptive classification approaches using Riemannian geometry. Those approaches are theoretically and experimentally compared during the second contribution ; we demonstrates the superiority in term of accuracy of merging independent classifications over the classification of the hyperbrain during the second contribution. Analysis of inter-brain synchronizations is a common approach for hyperscanning, however it is challenging for transient EEG waves with an great spatio-temporal variability (intra- and inter-subject) and with low signal-to-noise ratio such as ERP. Therefore, as third contribution, we propose a new blind source separation model, namely composite model, to extract simultaneously evoked EEG sources and ongoing EEG sources that allows to compensate this variability. A solution using approximate joint diagonalization is given and implemented with a fast Jacobi-like algorithm. We demonstrate on Brain Invaders 2 data that our solution extracts simultaneously evoked and ongoing EEG sources and performs better in term of accuracy and robustness compared to the existing models.

INTRODUÇÃO: O Acidente Vascular Cerebral Isquêmico (AVCI) é o evento mais freqüente dentre os AVCs, sendo caracterizado pela interrupção da irrigação sanguínea ao cérebro, a qual pode acarretar em lesão celular e alterações nas funções neurológicas. As manifestações clínicas desta doença podem incluir alterações das funções motoras, sensitivas, cognitivas, perceptivas, da linguagem entre outras. Sendo assim, é extremamente importante que sejam identificadas possíveis alterações nas vias auditivas, periférica e central, as quais podem prejudicar a qualidade de vida destes indivíduos. OBJETIVO: caracterizar os achados das avaliações comportamentais, eletroacústicas e eletrofisiológicas da audição em indivíduos destros com lesão isquêmica do hemisfério cerebral direito, bem como compará-los aos obtidos em indivíduos normais da mesma faixa etária. MÉTODOS: foram realizadas audiometria tonal, logoaudiometria, medidas de imitância acústica, potencial evocado auditivo de tronco encefálico (PEATE), potencial evocado auditivo de média latência (PEAML) e potencial cognitivo (P300) em 17 indivíduos com lesão do hemisfério direito (grupo pesquisa) e 25 normais (grupo controle), com idades entre 20 e 70 anos. RESULTADOS: Na análise dos dados qualitativos não foram encontradas alterações na avaliação comportamental da audição para os dois grupos. Ambos os grupos apresentaram alterações nos resultados do PEATE e do PEAML, sendo que houve diferença estatisticamente significante entre os grupos, para esses dois potenciais, nas quais o grupo pesquisa apresentou maior ocorrência de alterações. No PEATE a alteração mais freqüente foi do tipo tronco encefálico baixo, sendo que, entre os grupos, houve diferença estatisticamente significante, na qual o grupo pesquisa mostrou maior ocorrência dessa alteração. No PEAML a alteração predominante foi do tipo ambas (efeito orelha e efeito eletrodo ocorrendo concomitantemente) para o grupo pesquisa, e do tipo efeito eletrodo para o grupo controle. Na análise dos dados quantitativos (realizada apenas para os potenciais evocados auditivos) verificou-se, no PEATE, que ocorreu diferença estatisticamente significante entre os grupos com relação às latências das ondas III e V e interpicos I-III e I-V. Para o PEAML, a diferença estatisticamente significante entre os grupos ocorreu apenas para a latência da onda Na na posição C3/A1. Para o P300, ocorreu diferença entre os grupos com relação à latência da onda P300, sendo que o grupo pesquisa apresentou tempo médio de latência maior; além disso, houve uma tendência estatisticamente significante entre as orelhas direita e esquerda dentro do grupo pesquisa, mostrando aumento de latência da onda P300 na orelha direita. CONCLUSÃO: Indivíduos destros com lesão de hemisfério direito apresentaram limiares auditivos dentro da normalidade na avaliação comportamental da audição, entretanto, apresentaram resultados indicativos de déficit no sistema nervoso auditivo central, nas avaliações eletrofisiológicas da audição. Foram observados comprometimentos em tronco encefálico baixo, bem como nas regiões subcorticais e corticais. Dificuldades auditivas não foram percebidas pelos indivíduos, sugerindo que, provavelmente tal sinal possa estar relacionado à uma heminegligência auditiva. Tornam-se necessários mais estudos que avaliem a via auditiva central destes indivíduos para uma melhor caracterização dos achados eletrofisiológicos / INTRODUCTION: The ischemic cerebral stroke (ICS) is the most frequent event among cerebral strokes. It is characterized by the interruption of blood supply to the brain, which can lead to cell damage and alterations in neurological functions. The clinical manifestations of this disease may include alterations in motor, sensory, cognitive, perceptual and language functions among others. Therefore, the identification of possible alterations in both peripheral and central auditory pathways that may impair the quality of life of these individuals is extremely important. OBJECTIVE: To characterize the findings of behavioral, electrophysiological and electroacoustic hearing evaluations in right-handed individuals with right hemisphere ischemic lesion, and to compare such data to those obtained in normal individuals with the same age. METHODS: Pure tone audiometry, speech audiometry, acoustic immittance measures, brainstem auditory evoked potential (BAEP), Auditory Middle-Latency Response (AMLR) and cognitive potential (P300) were carried out in 17 subjects with right hemisphere lesions (research group) and 25 normal individuals (control group), aged between 20 and 70 years. RESULTS: No alterations were found on the qualitative data analysis of the hearing behavioral assessment of both groups. Both groups showed alterations in the BAEP and AMLR results, with statistically significant differences between groups for both potentials and a higher occurrence of alterations in the research group. The lower brainstem was the most frequent alteration type in the BAEP, and a statistically significant difference between groups was observed, with higher occurrence of such alteration in the research group. With regards the AMLR, the alteration predominantly observed was the Both type one (ear effect and electrode effect occurring concurrently) for the research group, and the electrode effect type one for the control group. In the analysis of quantitative data (performed only for the auditory evoked potentials), a statistically significant difference between groups was observed with respect to the BAEP latencies of waves III, V and interpeaks I-III and I-V. Regarding the AMLR measures, a statistically significant difference between groups was observed only for the Na wave latency in the C3/A1 position. For the P300, a difference between groups was observed, with higher mean latencies for the research group. In addition, there was a trend to statistically significant difference between right and left ears in the research group, which showed increased latency of P300 wave in the right ear. CONCLUSION: Right-handed individuals with right hemisphere lesion showed hearing thresholds within normal limits in the behavioral hearing assessment. However, they presented results indicative of central auditory nervous system deficits on the electrophysiological assessment of hearing. Alterations were observed in lower brainstem and in sub-cortical and cortical regions. Hearing difficulties were not perceived by these individuals, suggesting that this signal can probably be related to an auditory hemineglect. Further studies that evaluate the central auditory pathway of individuals with ICS are needed to better characterize the electrophysiological findings

L’objectif de cette thèse est d’étudier l’activation du système affectif médiée par le lexique orthographique au cours de la reconnaissance visuelle des mots. Plus précisément, nous avons étudié l’influence du voisinage orthographique émotionnel négatif et la sensibilité de l’amorçage orthographique à la valence négative de voisins plus fréquents dans une tâche de décision lexicale (TDL) combinée à un paradigme d’amorçage. Le recueil de mesures comportementales et électrophysiologiques (potentiels évoqués) nous a également permis d’évaluer la précocité de l’activation des composantes émotionnelles des voisins. Des mots neutres (e.g., FUSEAU, TOISON) ayant un seul voisin orthographique plus fréquent neutre (e.g., museau) ou négatif (e.g., poison) ont ainsi été présentés dans la TDL. Ils étaient précédés de leur voisin ou d’une amorce contrôle non alphabétique présenté pendant 66 ou 166 ms. Dans un premier temps, l’état émotionnel des participants a été contrôlé (Expériences 1-4). Dans un second temps, il a été manipulé a priori par une induction d’humeur triste (Expériences 5 et 7) ou déterminé a posteriori en considérant le niveau d’épuisement professionnel des participants (Expériences 7-8). Le traitement des mots fréquents neutres ou négatifs a été examiné en complément (Expérience 6). Les résultats montrent un effet inhibiteur du voisinage orthographique émotionnel négatif sur les temps de reconnaissance des mots cibles ainsi qu’un effet inhibiteur d’amorçage orthographique, accru par la durée de présentation des amorces. Trois composantes (P150, N200 et N400) constituent les corrélats électrophysiologiques de l’effet d’amorçage orthographique, sensibles à la valence négative des voisins et à la durée de présentation des amorces. Enfin, l’état émotionnel des individus modifie l’effet d’amorçage orthographique. Les résultats sont interprétés dans un modèle de type Activation Interactive de reconnaissance visuelle des mots adapté aux traitements affectifs. / The aim of this thesis was to study the activation of the affective system mediated by the orthographic lexicon during visual word recognition. More precisely, we have investigated the influence of the negative emotional orthographic neighbourhood and the sensitivity of orthographic priming to the negative valence of higher-frequency neighbours in the lexical decision task (LDT) combined with a priming paradigm. The recording of behavioural and electrophysiological (event-related brain potentials) measures provides also evidences on the early activation of affective components of the neighbours. Neutral words (e.g., FUSEAU [spindle], TOISON [fleece]) with one higher-frequency neighbour, that was either neutral (e.g., museau [muzzle]) or negative (e.g., poison), were presented in the LDT. They were preceded either by their neighbour or by a nonalphabetic control prime, presented 66 or 166 ms. Firstly, the emotional state of participants was controlled (Experiments 1-4). Secondly, it was manipulated a priori by a sad mood induction (Experiments 5 and 7) or determined a posteriori by considering the burnout level of participants (Experiments 7-8). The processing of negative or neutral frequent words have been also examined (Experiment 6). The results showed an inhibitory effect of negative emotional orthographic neighbourhood on target recognition time and an inhibitory effect of orthographic priming, increased by prime duration. Three components (P150, N200, and N400) were the electrophysiological correlates of orthographic priming effect, also depending on the negative valence of higher-frequency neighbours and prime duration. Finally, the emotional state of individuals modified the orthographic priming effect. The results are interpreted in an Interactive Activation model extended to affective processing.

Les troubles des tics, comme le syndrome de Gilles de la Tourette et le trouble de tics chroniques, sont des conditions neuropsychiatriques impliquant des tics moteurs et/ou phoniques. En plus de nombreuses comorbidités, les patients qui en sont atteints ont aussi des difficultés neuropsychologiques, notamment au niveau de l’inhibition et des fonctions motrices. La thérapie cognitivo-comportementale permet d’améliorer les tics et la condition générale de ces patients. Nous avons donc enregistré, durant une tâche de compatibilité stimulus-réponse, les potentiels évoqués cognitifs et les potentiels de latéralisation motrice (lateralized readiness potential; LRP) chez 20 patients atteints de trouble des tic avant et après une thérapie cognitivo-comportementale, et chez 20 participants contrôles. Chez les patients atteints de trouble des tics, nos résultats ont révélé une apparition plus tardive de l’amorce du LRP moyenné par rapport au stimulus, une amplitude plus élevée du LRP moyenné par rapport à la réponse, et une suractivation frontale liée aux processus d’inhibition. Suite à la thérapie, le retard au niveau de la latence de l’amorce du LRP moyenné par rapport à la réponse est comblé et l’amplitude du LRP moyenné par rapport à la réponse est normalisée, mais pas la suractivation frontale liée à l’inhibition. Cela suggère donc que la thérapie induit une modification des processus prémoteurs de sélection et de préparation de la réponse, ainsi que des processus d’exécution motrice, mais n’altère pas la suractivation frontale reliée aux fonctions inhibitrices. Étant donnés ces résultats, nous suggérons que la thérapie cognitivo-comportementale induit une modification du fonctionnement des aires motrices du cerveau. / Gilles de la Tourette syndrome and chronic tic disorder are two neuropsychiatric condition involving motor and/or phonic tics. In the DSM, these two conditions are characterized as “tic disorders”. Patients with such diagnoses face numerous comorbidities, and also show multiple neuropsychological impairments, especially concerning inhibition and motor processing. Tic symptoms, as well as general condition, can be improved with a cognitive-behavioural therapy (CBT). To this end, we recorded, during a stimulus-response compatibility task, event-related potentials and lateralized readiness potentials in 20 patients with tic disorder and 20 healthy controls. Patients and controls were paired on age, sex and hand dominance. Our result revealed a delay in stimulus-locked lateralized readiness potential (LRP) onset latency, higher response-locked LRP peak amplitude and a frontal overactivation related to stimulus inhibition in GTS and CTD patients. Following CBT, stimulus-locked LRP onset latency and response-locked LRP peak amplitude were normalized, but the frontal overactivation related to inhibition processing remained unchanged. These results suggest that CBT induces a modification of pre-motor processes such as response selection and preparation, as well as motor processes like response execution, but does not affect cortical activation related to stimulus inhibition. Since we found a reduction in tic symptoms, as well as a normalization of stimulus-locked LRP onset latency and response-locked LRP peak amplitude, and because LRPs are partly generated by the supplementary motor area, we suggest that CBT leads to a modification of this structure’s functioning.

Les tics affectent 1% des individus et sont associés avec une diminution de la qualité de vie. L’importante hétérogénéité phénoménologique retrouvée chez ceux-ci représente un obstacle majeur pour l’évaluation et le traitement de ces symptômes, et explique potentiellement la présence de données neurobiologiques contradictoires. Certaines variables rarement contrôlées, comme la complexité des tics et la demande motrice des tâches pourraient expliquer l’hétérogénéité de ces résultats. Une meilleure compréhension des processus cognitifs affectés pourrait être atteinte par l’étude de la chronométrie des événements cérébraux. Ainsi, notre objectif était d’évaluer l’impact de la complexité des tics et du type de réponse sur les potentiels électrocorticaux liés à l’inhibition, à l’attention et à la mémoire de patients tics. Nous avons comparé 12 patients présentant des tics simples avec 12 patients atteints de tics complexes, qui furent appariés à 15 participants contrôles sains. Deux tâches oddball furent accomplies, dont l’une exigeait une réponse motrice et l’autre une réponse non-motrice (compter le nombre de stimuli). Durant ces tâches, nous avons enregistré des composantes électrocorticales indexant des processus d’attention (P200), d’inhibition (N200) et de mémoire (P300). Pour la tâche non-motrice, nos résultats révélèrent une N200 plus ample chez les deux groupes de patients tics et une P300 réduite seulement chez ceux avec des tics simples. Quant à la tâche motrice, les deux groupes de patients tics présentaient une P300 réduite. Selon nos résultats, la complexité des tics et la demande motrice des tâches peuvent affecter les mécanismes neuronaux sous-tendant les différentes étapes du traitement de l’information. / Tic disorders affect up to 1% of individuals worldwide and are associated with social stigma. A major obstacle to these disorders’ evaluation and treatment is the large heterogeneity of phenotypes, which could explain the important number of conflicting neurobiological data. Certain variables, like tic complexity and tasks’ cognitive demand, could explain these heterogeneous results. We propose that a better understanding of cognitive impairments often seen in tic disorders could be obtained by studying the chronometry of electrocortical activity underlying inhibition, attention, memory and motor processes. Our aim was to evaluate the impact of tic complexity as well as task response demand on tic patients’ electrocortical profiles. Thus, we compared 12 patients exhibiting simple tics to 12 patients showing complex tics, and matched them to 15 healthy control participants. All participants performed two Oddball tasks in which one required a motor response (button press) and the other a non-motor response (counting). During both tasks, electrophysiological components were recorded, serving as indices of processes underlying attention (P200), inhibition (N200) and memory (P300). Our results revealed that for the non-motor task, both groups of tics patients had an enhanced N200, while only those with simple tics showed a reduced P300. For the motor task, both groups of tics patients exhibited a decreased P300. According to our results, both tic complexity and motor demand can affect neural mechanisms of the processing stream. Conflicting results found in previous studies might have been confounded by tic complexity as well as task motor requirements.