Une interface cerveau-ordinateur permet d'interagir avec un système, comme un système d'écriture, uniquement par l'activité cérébrale. Un des phénomènes neurophysiologiques permettant cette interaction est le potentiel évoqué cognitif P300, lequel correspond à une modification du signal 300 ms après la présentation d'une information attendue. Cette petite réaction cérébrale est difficile à observer par électroencéphalographie car le signal est bruité. Dans cette thèse, de nouvelles techniques basées sur la théorie des ondelettes sont développées pour améliorer la détection des P300 en utilisant des mesures de similarité entre les canaux électroencéphalographiques. Une technique présentée dans cette thèse débruite les signaux en considérant simultanément la phase des signaux. Nous avons également étendu cette approche pour étudier la localisation du P300 dans le but de sélectionner automatiquement la fenêtre temporelle à étudier et faciliter la détection / Brain-Computer Interfaces (BCI) are control and communication systems which were initially developed for people with disabilities. The idea behind BCI is to translate the brain activity into commands for a computer application or other devices, such as a spelling system. The most popular technique to record brain signals is the electroencephalography (EEG), from which Event-Related Potentials (ERPs) can be detected and used in BCI systems. Despite the BCI popularity, it is generally difficult to work with brain signals, because the recordings contains also noise and artifacts, and because the brain components amplitudes are very small compared to the whole ongoing EEG activity. This thesis presents new techniques based on wavelet theory to improve BCI systems using signals' similarity. The first one denoises the signals in the wavelet domain simultaneously. The second one combines the information provided by the signals to localize the ERP in time by removing useless information

Un des enjeux majeurs de la Psychologie Cognitive est de décrire les grandes fonctions mentales, notamment chez l’humain. Du point de vue neuroscientifique, il s’agit de modéliser l’activité cérébrale pour en extraire les éléments et mécanismes spatio-temporels susceptibles d’être mis en correspondance avec les opérations cognitives. Le travail de cette thèse a consisté à définir et mettre en œuvre des stratégies originales permettant de confronter les modèles cognitifs existants à des données issues d’enregistrements neurophysiologiques chez l’humain. Dans une première étude nous avons démontré que la distinction entre les organisations classiques de la dénomination de dessin sériel-parallèle, doit être adressée au niveau des essais uniques et non sur la moyenne des signaux. Nous avons conçu et mené l’analyse des signaux SEEG de 15 patients pour montrer que l’organisation temporelle de la dénomination de dessin n’est pas, au sens strict, parallèle. Dans une deuxième étude nous avons combiné trois techniques d’enregistrements : SEEG, EEG et MEG pour clarifier l’organisation spatiale des sources d’activité neuronales. Nous avons établi la faisabilité de l’enregistrement sur un patient qui exécute une tâche de perception visuelle. Au delà des corrélations entre les signaux moyens des trois techniques, cette analyse a révélé des corrélations au niveau des essais uniques. À travers deux approches expérimentales, cette thèse propose de nombreux développements méthodologiques et conceptuels originaux et pertinents. Ces contributions ouvrent de nouvelles perspectives à partir desquelles les signaux neurophysiologiques pourront informer les théories des Neurosciences Cognitives. / A major issue in Cognitive Psychology is to describe human cognitive functions. From the Neuroscientific perceptive, measurements of brain activity are collected and processed in order to grasp, at their best resolution, the relevant spatio-temporal features of the signal that can be linked with cognitive operations. The work of this thesis consisted in designing and implementing strategies in order to overcome spatial and temporal limitations of signal processing procedures used to address cognitive issues. In a first study we demonstrated that the distinction between picture naming classical temporal organizations serial-parallel, should be addressed at the level of single trials and not on the averaged signals. We designed and conducted the analysis of SEEG signals from 5 patients to show that the temporal organization of picture naming involves a parallel processing architecture to a limited degree only. In a second study, we combined SEEG, EEG and MEG into a simultaneous trimodal recording session. A patient was presented with a visual stimulation paradigm while the three types of signals were simultaneously recorded. Averaged activities at the sensor level were shown to be consistent across the three techniques. More importantly a fine-grained coupling between the amplitudes of the three recording techniques is detected at the level of single evoked responses. This thesis proposes various relevant methodological and conceptual developments. It opens up several perspectives in which neurophysiological signals shall better inform Cognitive Neuroscientific theories.

The study of neural and behavioural mechanisms of perceptual decision making is often done by experimental tasks involving the categorization of sensory stimuli. Among the key perceptual tasks that decision neuroscience researchers use are motion discrimination paradigms that include tracking and specifying the net direction of a single dot or a group of moving dots. These motion discrimination paradigms, such as the random-dot motion task, allow the measurement of the participant's perceptual decision making abilities in multiple task difficulty levels by varying the amount of noise in the sensory stimuli. Computational models of perceptual decision making, such as the drift-diffusion model, are widely used to analyze the behavioural measurements from these motion discrimination experiments. However, the standard drift-diffusion model can only analyze the average measures like reaction times or the proportion of correct decisions to explain the behavioural data. In the past decade, an emerging computational modeling approach was introduced to analyze the choice behaviour based on precise noise patterns in the sensory stimuli. These computational models that use spatiotemporal stimulus details have shown promise in the single-trial analysis of motion discrimination behaviour. In this thesis, I further develop the advanced computational models of perceptual decision making that use spatiotemporal dynamics of motion stimuli to provide detailed explanations of perceptual choice behaviour. First, I demonstrate the usefulness of equipping an extended Bayesian Model, equivalent to the extended drift-diffusion model, with trial-wise stimulus information leading to a significantly better explanation of behavioural data from a single-dot tracking experiment. Second, I show that the extended drift-diffusion model constrained by spatiotemporal stimulus details can explain the consistent biased choice behaviour in response to stochastic motion stimuli. Based on this model-based analysis, I provide evidence that the source of the observed biased choice behaviour is the presence of subtle motion information in the sensory stimuli. These results further emphasize the effectiveness of using spatiotemporal details of stochastic stimuli in detailed model-based analyses of the experimental data and provide computational interpretations of the data related to underlying mechanisms of perceptual decision making.

Perceptual decision making can be described as a process of accumulating evidence to a bound which has been formalized within drift-diffusion models (DDMs). Recently, an equivalent Bayesian model has been proposed. In contrast to standard DDMs, this Bayesian model directly links information in the stimulus to the decision process. Here, we extend this Bayesian model further and allow inter-trial variability of two parameters following the extended version of the DDM. We derive parameter distributions for the Bayesian model and show that they lead to predictions that are qualitatively equivalent to those made by the extended drift-diffusion model (eDDM). Further, we demonstrate the usefulness of the extended Bayesian model (eBM) for the analysis of concrete behavioral data. Specifically, using Bayesian model selection, we find evidence that including additional inter-trial parameter variability provides for a better model, when the model is constrained by trial-wise stimulus features. This result is remarkable because it was derived using just 200 trials per condition, which is typically thought to be insufficient for identifying variability parameters in DDMs. In sum, we present a Bayesian analysis, which provides for a novel and promising analysis of perceptual decision making experiments.

Perceptual decision making can be described as a process of accumulating evidence to a bound which has been formalized within drift-diffusion models (DDMs). Recently, an equivalent Bayesian model has been proposed. In contrast to standard DDMs, this Bayesian model directly links information in the stimulus to the decision process. Here, we extend this Bayesian model further and allow inter-trial variability of two parameters following the extended version of the DDM. We derive parameter distributions for the Bayesian model and show that they lead to predictions that are qualitatively equivalent to those made by the extended drift-diffusion model (eDDM). Further, we demonstrate the usefulness of the extended Bayesian model (eBM) for the analysis of concrete behavioral data. Specifically, using Bayesian model selection, we find evidence that including additional inter-trial parameter variability provides for a better model, when the model is constrained by trial-wise stimulus features. This result is remarkable because it was derived using just 200 trials per condition, which is typically thought to be insufficient for identifying variability parameters in DDMs. In sum, we present a Bayesian analysis, which provides for a novel and promising analysis of perceptual decision making experiments.

Cette thèse propose de nouvelles méthodes d'analyse d'enregistrements cérébraux intra-crâniens (potentiels de champs locaux), qui pallie les lacunes de la méthode temps-fréquence standard d'analyse des perturbations spectrales événementielles : le calcul d'une moyenne sur les enregistrements et l'emploi de l'activité dans la période pré-stimulus. La première méthode proposée repose sur la détection de sous-ensembles d'électrodes dont l'activité présente des synchronisations cooccurrentes en un même point du plan temps-fréquence, à l'aide de modèles bayésiens de mélange gaussiens. Les sous-ensembles d'électrodes pertinents sont validés par une mesure de stabilité calculée entre les résultats obtenus sur les différents enregistrements. Pour la seconde méthode proposée, le constat qu'un bruit blanc dans le domaine temporel se transforme en bruit ricien dans le domaine de l'amplitude d'une transformée temps-fréquence a permis de mettre au point une segmentation du signal de chaque enregistrement dans chaque bande de fréquence en deux niveaux possibles, haut ou bas, à l'aide de modèles bayésiens de mélange ricien à deux composantes. À partir de ces deux niveaux, une analyse statistique permet de détecter des régions temps-fréquence plus ou moins actives. Pour développer le modèle bayésien de mélange ricien, de nouveaux algorithmes d'inférence bayésienne variationnelle ont été créés pour les distributions de Rice et de mélange ricien. Les performances des nouvelles méthodes ont été évaluées sur des données artificielles et sur des données expérimentales enregistrées sur des singes. Il ressort que les nouvelles méthodes génèrent moins de faux-positifs et sont plus robustes à l'absence de données dans la période pré-stimulus / This thesis promotes new methods to analyze intracranial cerebral signals (local field potentials), which overcome limitations of the standard time-frequency method of event-related spectral perturbations analysis: averaging over the trials and relying on the activity in the pre-stimulus period. The first proposed method is based on the detection of sub-networks of electrodes whose activity presents cooccurring synchronisations at a same point of the time-frequency plan, using bayesian gaussian mixture models. The relevant sub-networks are validated with a stability measure computed over the results obtained from different trials. For the second proposed method, the fact that a white noise in the temporal domain is transformed into a rician noise in the amplitude domain of a time-frequency transform made possible the development of a segmentation of the signal in each frequency band of each trial into two possible levels, a high one and a low one, using bayesian rician mixture models with two components. From these two levels, a statistical analysis can detect time-frequency regions more or less active. To develop the bayesian rician mixture model, new algorithms of variational bayesian inference have been created for the Rice distribution and the rician mixture distribution. Performances of the new methods have been evaluated on artificial data and experimental data recorded on monkeys. It appears that the new methods generate less false positive results and are more robust to a lack of data in the pre-stimulus period

The goal of a Brain-Computer Interface (BCI) consists of the development of a unidirectional interface between a human and a computer to allow control of a device only via brain signals. While the BCI systems of almost all other groups require the user to be trained over several weeks or even months, the group of Prof. Dr. Klaus-Robert Müller in Berlin and Potsdam, which I belong to, was one of the first research groups in this field which used machine learning techniques on a large scale. The adaptivity of the processing system to the individual brain patterns of the subject confers huge advantages for the user. Thus BCI research is considered a hot topic in machine learning and computer science. It requires interdisciplinary cooperation between disparate fields such as neuroscience, since only by combining machine learning and signal processing techniques based on neurophysiological knowledge will the largest progress be made.<br><br> In this work I particularly deal with my part of this project, which lies mainly in the area of computer science. I have considered the following three main points:<br><br> <b>Establishing a performance measure based on information theory:</b> I have critically illuminated the assumptions of Shannon's information transfer rate for application in a BCI context. By establishing suitable coding strategies I was able to show that this theoretical measure approximates quite well to what is practically achieveable.<br> <b>Transfer and development of suitable signal processing and machine learning techniques:</b> One substantial component of my work was to develop several machine learning and signal processing algorithms to improve the efficiency of a BCI. Based on the neurophysiological knowledge that several independent EEG features can be observed for some mental states, I have developed a method for combining different and maybe independent features which improved performance. In some cases the performance of the combination algorithm outperforms the best single performance by more than 50 %. Furthermore, I have theoretically and practically addressed via the development of suitable algorithms the question of the optimal number of classes which should be used for a BCI. It transpired that with BCI performances reported so far, three or four different mental states are optimal. For another extension I have combined ideas from signal processing with those of machine learning since a high gain can be achieved if the temporal filtering, i.e., the choice of frequency bands, is automatically adapted to each subject individually.<br> <b>Implementation of the Berlin brain computer interface and realization of suitable experiments:</b> Finally a further substantial component of my work was to realize an online BCI system which includes the developed methods, but is also flexible enough to allow the simple realization of new algorithms and ideas. So far, bitrates of up to 40 bits per minute have been achieved with this system by absolutely untrained users which, compared to results of other groups, is highly successful. / Ein Brain-Computer Interface (BCI) ist eine unidirektionale Schnittstelle zwischen Mensch und Computer, bei der ein Mensch in der Lage ist, ein Gerät einzig und allein Kraft seiner Gehirnsignale zu steuern. In den BCI Systemen fast aller Forschergruppen wird der Mensch in Experimenten über Wochen oder sogar Monaten trainiert, geeignete Signale zu produzieren, die vordefinierten allgemeinen Gehirnmustern entsprechen. Die BCI Gruppe in Berlin und Potsdam, der ich angehöre, war in diesem Feld eine der ersten, die erkannt hat, dass eine Anpassung des Verarbeitungssystems an den Menschen mit Hilfe der Techniken des Maschinellen Lernens große Vorteile mit sich bringt. In unserer Gruppe und mittlerweile auch in vielen anderen Gruppen wird BCI somit als aktuelles Forschungsthema im Maschinellen Lernen und folglich in der Informatik mit interdisziplinärer Natur in Neurowissenschaften und anderen Feldern verstanden, da durch die geeignete Kombination von Techniken des Maschinellen Lernens und der Signalverarbeitung basierend auf neurophysiologischem Wissen der größte Erfolg erzielt werden konnte.<br><br> In dieser Arbeit gehe ich auf meinem Anteil an diesem Projekt ein, der vor allem im Informatikbereich der BCI Forschung liegt. Im Detail beschäftige ich mich mit den folgenden drei Punkten:<br><br> <b>Diskussion eines informationstheoretischen Maßes für die Güte eines BCI's:</b> Ich habe kritisch die Annahmen von Shannon's Informationsübertragungsrate für die Anwendung im BCI Kontext beleuchtet. Durch Ermittlung von geeigneten Kodierungsstrategien konnte ich zeigen, dass dieses theoretische Maß den praktisch erreichbaren Wert ziemlich gut annähert.<br> <b>Transfer und Entwicklung von geeigneten Techniken aus dem Bereich der Signalverarbeitung und des Maschinellen Lernens:</b> Eine substantielle Komponente meiner Arbeit war die Entwicklung von Techniken des Machinellen Lernens und der Signalverarbeitung, um die Effizienz eines BCI's zu erhöhen. Basierend auf dem neurophysiologischem Wissen, dass verschiedene unabhängige Merkmale in Gehirnsignalen für verschiedene mentale Zustände beobachtbar sind, habe ich eine Methode zur Kombination von verschiedenen und unter Umständen unabhängigen Merkmalen entwickelt, die sehr erfolgreich die Fähigkeiten eines BCI's verbessert. Besonders in einigen Fällen übertraf die Leistung des entwickelten Kombinationsalgorithmus die beste Leistung auf den einzelnen Merkmalen mit mehr als 50 %. Weiterhin habe ich theoretisch und praktisch durch Einführung geeigneter Algorithmen die Frage untersucht, wie viele Klassen man für ein BCI nutzen kann und sollte. Auch hier wurde ein relevantes Resultat erzielt, nämlich dass für BCI Güten, die bis heute berichtet sind, die Benutzung von 3 oder 4 verschiedenen mentalen Zuständen in der Regel optimal im Sinne von erreichbarer Leistung sind. Für eine andere Erweiterung wurden Ideen aus der Signalverarbeitung mit denen des Maschinellen Lernens kombiniert, da ein hoher Erfolg erzielt werden kann, wenn der temporale Filter, d.h. die Wahl des benutzten Frequenzbandes, automatisch und individuell für jeden Menschen angepasst wird.<br> <b>Implementation des Berlin Brain-Computer Interfaces und Realisierung von geeigneten Experimenten:</b> Eine weitere wichtige Komponente meiner Arbeit war eine Realisierung eines online BCI Systems, welches die entwickelten Methoden umfasst, aber auch so flexibel ist, dass neue Algorithmen und Ideen einfach zu verwirklichen sind. Bis jetzt wurden mit diesem System Bitraten von bis zu 40 Bits pro Minute von absolut untrainierten Personen in ihren ersten BCI Experimenten erzielt. Dieses Resultat übertrifft die bisher berichteten Ergebnisse aller anderer BCI Gruppen deutlich. <br> <hr> Bemerkung:<br> Der Autor wurde mit dem <i>Michelson-Preis</i> 2005/2006 für die beste Promotion des Jahrgangs der Mathematisch-Naturwissenschaftlichen Fakultät der Universität Potsdam ausgezeichnet.

Η σύγχρονη μελέτη της κεντρικής και αυτόνομης δραστηριότητας θεωρούνται χρήσιμες στην ανάδειξη της σχέσης των δύο συστημάτων και της φυγόκεντρου ανατροφοδότησης των περιφερικών αλλαγών. Στην εργασία αυτή διερευνώνται οι πιθανές σχέσεις μεταξύ κεντρικών και αυτόνομων απαντήσεων σε σωματαισθητικούς ερεθισμούς. Η ηλεκτροδερμική δραστηριότητα (ΗΔΑ) είναι δείκτης της αυτόνομης δραστηριότητας. Προκλητά δυναμικά και ΗΔΑ καταγράφονταν ταυτόχρονα έπειτα από μια σειρά ερεθισμών του μέσου νεύρου ,με τρείς διαφορετικές εντάσεις, σε έξι φυσιολογικά άτομα. Η χαμηλότερη ένταση ερεθισμού ρυθμίζονταν έτσι ώστε να μη γίνεται αντιληπτή από το άτομα. Τα επάρματα P40, P100, N200 και P300 των προκλητών δυναμικών εξήχθησαν από το καταγεγραμμένο σήμα. Μελετήσαμε το πλάτος και τον λανθάνοντα χρόνο και διερευνήσαμε την πιθανή συσχέτιση των μεγεθών αυτών της ΗΔΑ. Χρησιμοποιήσαμε την συνηθισμένη μέθοδο της μεσοποίησης (average), καθώς και την προσέγγιση των μοναδιαίων απαντήσεων, εξάγωντας τις απαντήσεις με την χρήση χωρικού φίλτρου. Το συμπαθητικό νευρικό σύστημα φαινόταν να απαντά ακόμα και σε ερεθισμούς μη αντιληπτούς από το υποκείμενο. Η φλοιική επεξεργασία των αντιληπτών και μη ερεθισμάτων όπως αποτυπώνονταν με τα επάρματα των προκλητών δυναμικών, φαίνεται να είναι όμοια και στις δύο περιπτώσεις κατά τη διάρκεια των πρώτων 40ms έπειτα από τον ερεθισμό, ενώ τα μακρά κύματα ήταν απόνατ στους μη αντιληπτούς ερεθισμούς. Τα πλάτη της ΗΔΑ και του P300 με την προσέγγιση των μοναδιαίων απαντήσεων παρουσίαζαν μια σημαντική θετική συσχέτιση. / Concurrent studies of central and autonomic activity are considered useful in elucidating the relationship between the two systems and indicating the centripetal feedback of peripheral changes. The SSR (sympathetic skin response) is one index of autonomic arousal. In our study we examine the possible relationships between central and autonomic responses to somatosensory stimuli. EPs (evoked potentials) and SSRs were simultaneously recorded during a series of electrical stimuli of median nerve in six normal adults using three different intensities of stimuli. The weakest of them was unperceived by subject. The P40, P100, N200 and P300 waves of EPs were extracted and their latencies and amplitudes were analysed in order to find correlations with those of the SSRs. We used the conventional method of signal averaging and the single trial (ST) approach, as the EP waves were subtracted by spatial filtering. Interestingly, sympathetic nervous system seemed to react even to stimuli unperceived by the subject. The cortical processing of consciously perceived and unperceived somatosensory stimuli as it was expressed by the evoked potentials seems to be identical during the first 40ms after the stimulus onset while later waves were absent for unperceived stimuli. SSR and P300 amplitude at the ST level had a positive correlation.