Abstract The concept of resting state functional magnetic resonance imaging (fMRI) is built onto an original finding in 1995 that brain hemispheres present synchronous signal fluctuations with distinct patterns. fMRI measurements rely on blood oxygenation changes that indirectly mirror neural activity. Therefore, the origin of functional connectivity patterns, resting state networks (RSNs), has been a widely debated research question and numerous contributing factors have been identified. According to current understanding the fluctuations reflect maintenance of the system integrity in addition to spontaneous thought and action processes in the resting state. A popular method to study the functional connectivity in resting state fMRI is spatial independent component analysis (ICA) that decomposes signal sources into statistically independent components. The dichotomy of functional specialization versus functional integration has a correspondence in fMRI studies where RSNs play the integrative viewpoint of brain function. Although canonical large-scale RSNs are broadly distributed they also express modularity that can be accomplished by ICA with a high number of estimated components. The characteristics of high ICA dimensionality are broadly investigated in the thesis. An enduring issue in resting state research has been the confounding noise sources like motion and cardiorespiratory processes which may hamper the analysis. In this thesis the ability of ICA to separate these noise sources from the default mode network, a major RSN, is studied. Additionally, the suitability of ICA for full frequency spectrum analysis, a relatively rare setting in biosignal analysis, is investigated. The results of the thesis support the viewpoint of ICA as a robust analysis method for functional connectivity analysis. Cardiorespiratory and motion induced noise did not confound the functional connectivity analyses with ICA. High dimensional ICA provided better signal source separation, revealed the modular structure of the RSNs and pinpointed the specific aberrations in the autism spectrum disorder population. ICA was also found applicable for fully explorative analysis in both the spatial and temporal domains and indicated functional connectivity changes induced by transcranial bright light stimulation. / Tiivistelmä Konsepti lepotilan tutkimisesta toiminnallisella magneettikuvauksella (engl. functional magnetic resonance imaging, fMRI) on rakentunut vuonna 1995 tehdylle löydökselle aivopuoliskojen välillä synkronisesta signaalivaihtelusta. Mittaukset perustuvat veren hapetuksen muutoksiin, jotka epäsuorasti heijastelevat hermostollista toimintaa. Tämän takia toiminnallisen kytkennällisyyden muodot, lepotilaverkostot, ovat olleet laajasti väitelty tutkimusaihe ja monia verkostoihin vaikuttavia tekijöitä onkin tunnistettu. Nykykäsityksen mukaan signaalivaihtelut lepotilassa heijastelevat järjestelmän yhtenäisyyden ylläpitoa spontaanin ajattelun ja toiminnan lisäksi. Suosittu menetelmä toiminnallisen kytkennällisyyden tutkimiseen lepotilan fMRI:ssä on spatiaalinen itsenäisten komponenttien analyysi (engl. independent component analysis, ICA), joka hajottaa signaalilähteet tilastollisesti itsenäisiin komponentteihin. Aivotoiminnan mallintamisessa kahtiajaolla toiminnalliseen erikoistumiseen ja toiminnalliseen integraatioon on vastaavuus fMRI-tutkimukseen, jossa lepotilaverkostot vastaavat toiminnallisen integraation näkökulmasta. Vaikka kanoniset lepotilaverkostot ovat laaja-alaisia, ne ovat toisaalta modulaarisia, jota voidaan tutkia tutkimalla korkean komponenttimäärän ICA-hajotelmaa. Korkea- dimensioisen ICA-hajotelman ominaisuuksia tutkitaan laajasti tässä väitöskirjassa. Kestoaihe lepotilatutkimuksessa on ollut analyysiä hankaloittavien kohinalähteiden kuten liikkeen ja kardiorespiratoristen prosessien vaikutus. Väitöskirjassa tutkitaan ICA:n kykyä erotella kohinalähteitä ’default mode’ -verkostosta, joka on merkittävin lepotilaverkosto. Lisäksi tutkitaan ICA:n soveltuvuutta täyden taajuuskaistan analysointiin, joka on verrattain harvinaista biosignaalien analyysissä. Väitöskirjan tulokset tukevat näkemystä ICA:n suorituskyvystä toiminnallisen kytkennällisyyden analyysissä. Kardiorespiratorinen ja liikkeestä lähtöisin oleva kohina ei häirinnyt merkittävästi ICA-tuloksia. Korkeadimensioinen ICA tarjosi paremman erottelun signaalilähteille, paljasti lepotilaverkostojen modulaarisen rakenteen ja määritti erityisen poikkeaman autismin kirjon oireyhtymän populaatiossa. ICA:n havaittiin olevan soveltuva täyseksploratiiviselle analyysille ajassa ja avaruudessa; tulos viittaa toiminnallisen kytkennällisyyden muutoksiin kallon läpäisevän kirkasvalostimulaation aikaansaamana.

Le but de ce travail de thèse a été de caractériser in vivo chez l’Homme, la connectivité cérébrale sur son versant fonctionnel, par le biais de deux techniques: la stéréoélectroencéphalographie (SEEG) et l’IRM fonctionnelle (IRMf) de repos. Ces travaux se sont intégrés dans le cadre du bilan préchirurgical des épilepsies du lobe temporal pharmacorésistantes, dont le but est de déterminer la zone épileptogène à réséquer pour traiter ces patients. Alors que plusieurs études en électrophysiologie ont montré que durant les crises, il existait une synchronisation anormalement élevée entre les structures impliquées dans les processus épileptogènes, aucune donnée de connectivité n’était disponible en période intercritique. Pourtant, la période intercritique est le siège d’anomalies interictales enregistrées en EEG, de profonds remaniements morphologiques, et est associée à des troubles cognitifs. Nous apportons avec ce travail, grâce au recueil et au traitement de signaux SEEG et IRMf enregistrés durant la période intercritique, de nouvelles connaissances i) sur l’organisation de la connectivité fonctionnelle basale (CFB) chez les sujets sains ; ii) sur les altérations de la CFB chez des groupes de patients mais également au niveau individuel ; iii) sur le rapport entre ces anomalies de CFB et les troubles cognitifs observés chez ces patients ; iv) enfin, sur les différences et les similitudes de la CFB évaluée par SEEG et IRMf chez les mêmes sujets, ouvrant ainsi de nouvelles perspectives dans la compréhension des relations entre le signal BOLD et le signal EEG. / The aim of this thesis was to characterize the Human brain functional connectivity in vivo based on signals recorded using stereoelectroencephalography (SEEG) and resting-state functional MRI (fMRI). This work was conducted during the presurgical assessment of drug resistant temporal lobe epilepsy, which aims at determining the epileptogenic zone to be removed to treat these patients. While several electrophysiological studies have shown high synchronization between structures involved in the epileptogenic process during seizure, no similar connectivity data was available during inter-critical period. However, the interictal period is characterized by spikes recorded on EEG, morphological alterations and cognitive impairment. By analyzing fMRI and SEEG signals recorded during the interictal period, this work provides new insights into, i) basal functional connectivity (BFC) organization in healthy subjects, ii) BFC alterations in patients groups but also at the individual level, iii) the relationship between these BFC abnormalities and cognitive impairment observed in these patients; iv) the differences and similarities of BFC evaluated by SEEG and fMRI in the same subjects, thus opening up new perspectives in better understanding of relationships between BOLD and SEEG signal coupling.

Hintergrund. Exzessives Sorgen ist durch anhaltende, sich wiederholende negative Gedanken gekennzeichnet, die als aufdringlich und unkontrollierbar empfunden werden. Chronisches Sorgen kann zu einer schwer beeinträchtigenden mentalen Aktivität werden und es wird angenommen, dass es zur Entstehung, Aufrechterhaltung und Verschlechterung einer Reihe von somatischen Gesundheitsproblemen und psychischen Störungen beiträgt. Theoretische Modelle und empirische Befunde deuten darauf hin, dass exzessives Sorgen mit einer gestörten Bottom-up-Salienzverarbeitung, einer unzureichenden Top-down-Aufmerksamkeitssteuerung, Defiziten in der Emotionsregulation und abnormalen selbstreferenziellen mentalen Funktionen verbunden sind. Neuroimaging-Studien zu exzessivem Sorgen zeigen Veränderungen funktioneller Aktivierung und Konnektivität in limbischen und paralimbischen Hirnstrukturen, welche die Reaktivität auf emotionale Stimuli unterstützen, in präfrontalen Strukturen, die in Top-down-Prozesse involviert sind, welche der Aufmerksamkeitssteuerung und Emotionsregulation zugrunde liegen, und in medialen kortikalen Mittellinienstrukturen, die an selbstreferenziellen mentalen Aktivitäten beteiligt sind. Im Hinblick auf das Vorhandensein, die genaue Lokalisation der beteiligten Hirnareale und die Richtung der Effekte präsentieren diese Studien jedoch weitgehend heterogene Ergebnisse. Die hohe Variabilität der Befunde erschwert es, ein kohärentes Verständnis der neurobiologischen Mechanismen exzessiven Sorgens zu erlangen. Um dieses Verständnis zu erweitern und künftige Richtungen für die weitere Forschung auf diesem Gebiet aufzuzeigen, verfolgte die vorliegende Dissertationsschrift drei Ziele: (i) die emotionsbezogene, aufgabenbasierte fMRT-Literatur zu exzessivem Sorgen auf quantitative, datengesteuerte Weise zusammenzufassen, um konsistente funktionelle Störungen über Studien hinweg zu identifizieren; (ii) zu bestimmen, mit welchen psychologischen Prozessen die identifizierten Hirnregionen assoziiert sind, und in welchen funktionellen Hirnnetzwerken sie wirken; und (iii) Anomalien in der grundlegenden Hirnorganisation zu untersuchen, die mit exzessivem Sorgen assoziiert sind. Methoden. Eine State-of-the-Art koordinatenbasierte Meta-Analyse wurde unter Anwendung des Activation Likelihood Estimation (ALE) Algorithmus durchgeführt, um die Übereinstimmung zwischen 16 Neuroimaging-Experimenten zu bestimmen, die Veränderungen in der funktionellen Aktivität des Gehirns während der Verarbeitung emotionaler Inhalte zwischen Personen mit hoher und normaler Sorgenneigung berichten. Die identifizierten Regionen wurden mithilfe von Metadaten der funktionellen Magnetresonanztomographie (fMRT) hinsichtlich ihrer psychologischen Funktionen charakterisiert (Verhaltens-Charakterisierung). Zusätzlich wurde meta-analytic-connectivity modeling (MACM) verwendet, um ihre globalen funktionellen Konnektivitätsmuster zu bestimmen und so zugehörige Gehirnnetzwerke zu identifizieren. Schließlich wurde fMRT im Ruhezustand (resting-state) verwendet, um die funktionellen Konnektivitätsmuster zwischen 21 Personen mit hoher und 21 Personen mit normaler Sorgenneigung ohne einer aufgabenbezogenen Gehirnaktivierung zu vergleichen. Dispositionelle Sorgen wurden mit dem Penn State Worry Questionnaire als verlässliches Selbstauskunftsmaß für schwere Sorgen erhoben. Saatregion-basierte Analysen mit den meta-analytisch abgeleiteten Hirnregionen als Saatregionen und eine datengesteuerte Multi-Voxel-Pattern-Analyse (MVPA) wurden durchgeführt, um funktionelle Konnektivitätsunterschiede zwischen den beiden Gruppen zu detektieren. Darüber hinaus wurden gruppenüber-greifende Korrelationen zwischen dem aktuellen Sorgenausmaß (State-Sorgen) und den funktionellen Konnektivitätsmustern der Saat-Regionen sowie den aus der MVPA abgeleiteten Komponenten-Werten analysiert. Ergebnisse. Die Meta-Analyse ergab konvergente Anomalien bei Individuen mit hoher im Vergleich mit normaler Sorgenneigung, hauptsächlich in einem linkshemisphärischen Cluster, welcher Teile des mittleren frontalen Gyrus, des inferioren frontalen Gyrus und der anterioren Insula umfasst. Die Verhaltens-Charakterisierung zeigte, dass der identifizierte Cluster mit der Sprachverarbeitung und dem Gedächtnis assoziiert ist. Darüber hinaus ergaben die meta-analytischen Konnektivitätskartierungen starke funktionelle Verbindungen zwischen den beobachteten konvergenten Regionen und frontalen, temporalen und parietalen Hirnregionen, die sich mit Teilen von zwei verhaltensrelevanten Hirnnetzwerken überschneiden, nämlich dem Salienznetzwerk (SN) und dem Default-Netzwerk (DN). Die resting-state funktionellen Konnektivitätsanalysen zeigten keine Unterschiede zwischen Individuen mit hoher und normaler Sorgenneigung und auch keine Korrelationen zwischen den resting-state funktionellen Konnektivitätsmustern und State-Sorgen, weder mit dem auf Saatregionen basierenden Ansatz noch mit dem MVPA-Ansatz. Schlussfolgerungen. Die Ergebnisse dieser Dissertationsschrift deuten darauf hin, dass exzessives Sorgen mit einer gestörten Funktion in Hirnarealen zusammenhängt, die mit bottom-up und top-down Aufmerksamkeitssteuerung sowie Emotionserzeugung und Emotionsregulation in Verbindung gebracht werden. Die Verhaltensanalyse deckte Assoziationen zwischen dem identifizierten Cluster und der Sprachverarbeitung auf, die mit dem übermäßigen inneren Sprechen bei zu Sorgen neigenden Personen zusammenhängen könnten. Diese Assoziation ist bisher eher unbeachtet geblieben und sollte weiter erforscht werden. Darüber hinaus stellen die identifizierten Hirnregionen Schlüsselknoten in interagierenden neuronalen Netzwerken dar, die endogen und exogen orientierte Kognition unterstützen und das dynamische Zusammenspiel zwischen diesen Prozessen steuern. Ihre veränderte netzwerkübergreifende Dynamik könnte die Ursache für die Unfähigkeit von zu schweren Sorgen neigen-den Personen sein, sich von intern orientierten Kognitionen zu lösen, wenn adaptives Reagieren einen externen Fokus der Aufmerksamkeit erfordern würde. Die Nullergebnisse der Ruhezustandsanalysen könnten auf das Studiendesign zurückzuführen sein oder durch Charakteristika des Sorgens selbst verursacht werden, werden aber nicht als Beleg dafür interpretiert, dass Anomalien in der intrinsischen Konnektivität des Gehirns in Verbindung mit exzessivem Sorgen nicht vorhanden sind. Die Ergebnisse dieser Arbeit können zukünftige Forschungen anleiten, die z.B. untersuchen könnten, ob und wie sich die dynamischen zeitlichen Interaktionen innerhalb und zwischen den hier identifizierten Netzwerken in Abhängigkeit vom Schweregrad des Sorgens unterscheiden. Die ALE-Ergebnisse liefern eine A-priori-Auswahl von Hirnregionen für solche Studien. Ein besseres Verständnis der Veränderungen in den Gehirnnetzwerken, die exzessivem Sorgen zugrunde liegen, und der psychologischen Funktionen, die dadurch beeinträchtigt werden, wird Ansatzpunkte für die Verbesserung therapeutischer Interventionen liefern.:Contents TABLES VIII FIGURES IX ABBREVIATIONS X ABSTRACT 1 1 THEORETICAL BACKGROUND 6 1.1 WORRY 6 1.1.1 DEFINITION, NATURE AND FUNCTION OF WORRY 6 1.1.2 THE WORRY CONTINUUM – NORMAL VERSUS MALADAPTIVE WORRY 7 1.1.3 THE DELETERIOUS EFFECTS OF EXCESSIVE WORRY 8 1.1.4 THEORETICAL MODELS OF EXCESSIVE WORRY 11 1.2 FUNCTIONAL BRAIN NETWORKS AND EXCESSIVE WORRY 18 1.2.1 A SYSTEMS NEUROSCIENCE VIEW OF EXCESSIVE WORRY 18 1.2.2 EMPIRICAL EVIDENCE: FMRI STUDIES ON EXCESSIVE WORRY 20 1.3 RESEARCH QUESTION 32 2 STUDY I: A QUANTITATIVE META-ANALYSIS OF FMRI STUDIES INVESTIGATING EMOTIONAL PROCESSING IN EXCESSIVE WORRIERS: APPLICATION OF ACTIVATION LIKELIHOOD ESTIMATION ANALYSIS 35 2.1 ABSTRACT 36 2.2 INTRODUCTION 37 2.3 METHODS 40 2.3.1 LITERATURE SEARCH AND STUDY SELECTION 40 2.3.2 ACTIVATION LIKELIHOOD ESTIMATION 46 2.3.3 META-ANALYTIC CONNECTIVITY MODELING 47 2.3.4 ANALYSIS OF BEHAVIORAL DOMAIN PROFILES 47 2.4 RESULTS 48 2.4.1 SIGNIFICANT ALE CLUSTERS 48 2.4.2 FUNCTIONAL CONNECTIVITY OF THE DERIVED ALE-CLUSTER – MACM-ANALYSIS 51 2.4.3 FUNCTIONAL CHARACTERIZATION OF THE DERIVED ALE-CLUSTER – BEHAVIORAL ANALYSIS 54 2.5 DISCUSSION 55 2.6 CONCLUSION 59 2.7 SUPPLEMENTARY MATERIAL STUDY I: LISTING OF ALE CLUSTERS SIGNIFICANT AT P < 0.001 UNCORRECTED, CLUSTER SIZE > 100MM3 60 3 STUDY II: HIGH AND LOW WORRIERS DO NOT DIFFER IN UNSTIMULATED RESTING-STATE BRAIN CONNECTIVITY 61 3.1 ABSTRACT 62 3.2 INTRODUCTION 63 3.3 MATERIALS AND METHODS 65 3.3.1 PARTICIPANTS AND PROCEDURE 65 3.3.2 FMRI DATA ACQUISITION 66 3.3.3 SELF-REPORT ASSESSMENTS AND STATE WORRY ASSESSMENT 66 3.3.4 STATISTICAL ANALYSES 67 3.4 RESULTS 69 3.4.1 SELF-REPORT MEASURES 69 3.4.2 FMRI RESULTS 72 3.5 DISCUSSION 72 3.6 CONCLUSION 75 3.7 SUPPLEMENTARY MATERIAL STUDY II: STATE WORRY ASSESSMENT 75 4 GENERAL DISCUSSION 76 4.1 CONVERGENT ABERRANT FUNCTION IN THE MFG-IFG-INSULA-CLUSTER 76 4.2 META-ANALYTIC FUNCTIONAL CHARACTERIZATION AND CONNECTIVITY MAPPING OF THE MFG-IFG-INSULA CLUSTER 82 4.3 NO RESTING-STATE FUNCTIONAL CONNECTIVITY DIFFERENCES BETWEEN HW AND LW 84 4.4 STRENGTHS AND LIMITATIONS 87 4.5 FUTURE DIRECTIONS 90 4.6 CONCLUSION 91 REFERENCES 92 APPENDIX: DECLARATION OF HONOUR / EIGENSTÄNDIGKEITSERKLÄRUNG 131 / Background. Excessive worry is characterized by persistent, repetitive negative thoughts that are perceived as intrusive and uncontrollable. Chronic worrying can become a severely debilitating mental activity and is thought to contribute to the development, maintenance and deterioration of a range of somatic health problems and mental disorders. Theoretical accounts and empirical findings suggest that excessive worry is associated with impaired bottom-up salience-processing, insufficient top-down attentional control, deficits in emotion regulation and abnormal self-referential mental functions. Neuroimaging studies of excessive worry indicate functional activation and connectivity alterations in limbic and paralimbic brain structures that support reactivity to emotional stimuli, in prefrontal structures implicated in top-down processes underlying attentional control and emotion regulation, and in cortical midline structures involved in self-referential mental activity. However, with regard to the presence, the exact localization of the brain areas involved and the directionality of the effects, these studies have presented largely heterogenous results. The high variability of findings makes it difficult to achieve a coherent understanding of the neurobiological mechanisms of excessive worry. To extend this understanding and provide future directions for continued research in this area, the aim of this thesis was threefold: (i) to synthesize the emotional task-based fMRI literature on excessive worry in a quantitative, data-driven manner for the purpose of identifying consistent functional perturbations across studies; (ii) to determine the psychological processes with which the identified brain regions are associated and the functional brain networks in which they operate; and (iii) to examine abnormalities in basic brain organization associated with excessive worry. Methods. A state-of-the-art coordinate-based meta-analysis was conducted applying the activation likelihood estimation (ALE) algorithm to determine concordance among 16 neuroimaging experiments reporting alterations in brain functional activity during emotional processing between individuals experiencing high versus normal levels of worry. The identified regions were behaviorally characterized using functional magnetic resonance imaging (fMRI) metadata. Additionally, meta-analytic-connectivity modeling (MACM) was used to determine their global functional connectivity (FC) patterns and thus identify related brain networks. Finally, resting-state fMRI was used to compare FC patterns between 21 high and 21 low worriers in the absence of task-related brain activation. Dispositional worry was assessed using the Penn State Worry Questionnaire as a reliable self-report measure of severe worry. Seed-based analyses with the meta-analytically derived brain regions as seeds and a data-driven multi-voxel pattern analysis (MVPA) were performed to detect FC differences between the two groups. In addition, cross-group correlations between state worry levels and the FC patterns of the seed regions as well as the MVPA-derived component scores were analyzed. Results. The meta-analysis revealed convergent aberrations in high compared to normal worriers mainly in a left-hemispheric cluster comprising parts of the middle frontal gyrus, inferior frontal gyrus and anterior insula. Behavioral characterization indicated the identified cluster to be associated with language processing and memory. Furthermore, meta-analytic connectivity mapping yielded strong functional connections between the observed convergent regions and frontal, temporal, and parietal brain regions that overlap with parts of two behaviorally relevant brain networks, specifically the salience network (SN) and the default network (DN). The resting-state FC (rsFC) analyses revealed no differences between high and normal worriers and also no correlations between rsFC patterns and state worry, neither using the seed-based nor the MVPA approach. Conclusions. The results of this thesis indicate that excessive worry is related to disturbed functioning in brain areas that have been related to bottom-up and top-down attentional control as well as emotion generation and regulation. Behavioral analysis uncovered associations between the identified cluster and language processing that might be related to the exaggerated inner speech processes in worry prone individuals. This association has so far remained rather unnoticed and requires further exploration. Moreover, the identified brain regions constitute key nodes within interacting neural networks that support internally and externally oriented cognition and control the dynamic interplay among these processes. Their altered cross-network dynamics may underlie the inability of worry-prone individuals to disengage from internally oriented cognitions when adaptive responding would require an external focus of attention. The null-findings of the resting-state analyses might be due to the study design or caused by characteristics of worry itself, but are not interpreted as evidence that abnormalities in the brain's intrinsic connectivity associated with excessive worrying are absent. The results of this thesis may guide future research that could, for example, investigate whether and how the dynamic temporal interactions within and between the networks identified here differ depending on the severity of worry. The ALE results provide an a priori selection of brain regions for such studies. Increasing our understanding of the aberrations in brain networks that underlie excessive worry and the psychological functions that are impaired as a result will provide targets for improving therapeutic interventions.:Contents TABLES VIII FIGURES IX ABBREVIATIONS X ABSTRACT 1 1 THEORETICAL BACKGROUND 6 1.1 WORRY 6 1.1.1 DEFINITION, NATURE AND FUNCTION OF WORRY 6 1.1.2 THE WORRY CONTINUUM – NORMAL VERSUS MALADAPTIVE WORRY 7 1.1.3 THE DELETERIOUS EFFECTS OF EXCESSIVE WORRY 8 1.1.4 THEORETICAL MODELS OF EXCESSIVE WORRY 11 1.2 FUNCTIONAL BRAIN NETWORKS AND EXCESSIVE WORRY 18 1.2.1 A SYSTEMS NEUROSCIENCE VIEW OF EXCESSIVE WORRY 18 1.2.2 EMPIRICAL EVIDENCE: FMRI STUDIES ON EXCESSIVE WORRY 20 1.3 RESEARCH QUESTION 32 2 STUDY I: A QUANTITATIVE META-ANALYSIS OF FMRI STUDIES INVESTIGATING EMOTIONAL PROCESSING IN EXCESSIVE WORRIERS: APPLICATION OF ACTIVATION LIKELIHOOD ESTIMATION ANALYSIS 35 2.1 ABSTRACT 36 2.2 INTRODUCTION 37 2.3 METHODS 40 2.3.1 LITERATURE SEARCH AND STUDY SELECTION 40 2.3.2 ACTIVATION LIKELIHOOD ESTIMATION 46 2.3.3 META-ANALYTIC CONNECTIVITY MODELING 47 2.3.4 ANALYSIS OF BEHAVIORAL DOMAIN PROFILES 47 2.4 RESULTS 48 2.4.1 SIGNIFICANT ALE CLUSTERS 48 2.4.2 FUNCTIONAL CONNECTIVITY OF THE DERIVED ALE-CLUSTER – MACM-ANALYSIS 51 2.4.3 FUNCTIONAL CHARACTERIZATION OF THE DERIVED ALE-CLUSTER – BEHAVIORAL ANALYSIS 54 2.5 DISCUSSION 55 2.6 CONCLUSION 59 2.7 SUPPLEMENTARY MATERIAL STUDY I: LISTING OF ALE CLUSTERS SIGNIFICANT AT P < 0.001 UNCORRECTED, CLUSTER SIZE > 100MM3 60 3 STUDY II: HIGH AND LOW WORRIERS DO NOT DIFFER IN UNSTIMULATED RESTING-STATE BRAIN CONNECTIVITY 61 3.1 ABSTRACT 62 3.2 INTRODUCTION 63 3.3 MATERIALS AND METHODS 65 3.3.1 PARTICIPANTS AND PROCEDURE 65 3.3.2 FMRI DATA ACQUISITION 66 3.3.3 SELF-REPORT ASSESSMENTS AND STATE WORRY ASSESSMENT 66 3.3.4 STATISTICAL ANALYSES 67 3.4 RESULTS 69 3.4.1 SELF-REPORT MEASURES 69 3.4.2 FMRI RESULTS 72 3.5 DISCUSSION 72 3.6 CONCLUSION 75 3.7 SUPPLEMENTARY MATERIAL STUDY II: STATE WORRY ASSESSMENT 75 4 GENERAL DISCUSSION 76 4.1 CONVERGENT ABERRANT FUNCTION IN THE MFG-IFG-INSULA-CLUSTER 76 4.2 META-ANALYTIC FUNCTIONAL CHARACTERIZATION AND CONNECTIVITY MAPPING OF THE MFG-IFG-INSULA CLUSTER 82 4.3 NO RESTING-STATE FUNCTIONAL CONNECTIVITY DIFFERENCES BETWEEN HW AND LW 84 4.4 STRENGTHS AND LIMITATIONS 87 4.5 FUTURE DIRECTIONS 90 4.6 CONCLUSION 91 REFERENCES 92 APPENDIX: DECLARATION OF HONOUR / EIGENSTÄNDIGKEITSERKLÄRUNG 131

Environ 25% de la population mondiale est atteinte de troubles psychiatriques qui sont typiquement associés à des problèmes comportementaux, fonctionnels et/ou cognitifs et dont les corrélats neurophysiologiques sont encore très mal compris. Non seulement ces dysfonctionnements réduisent la qualité de vie des individus touchés, mais ils peuvent aussi devenir un fardeau pour les proches et peser lourd dans l’économie d’une société. Cibler les mécanismes responsables du fonctionnement atypique du cerveau en identifiant des biomarqueurs plus robustes permettrait le développement de traitements plus efficaces. Ainsi, le premier objectif de cette thèse est de contribuer à une meilleure caractérisation des changements dynamiques cérébraux impliqués dans les troubles mentaux, plus précisément dans la schizophrénie et les troubles d’humeur. Pour ce faire, les premiers chapitres de cette thèse présentent, en intégral, deux revues de littératures systématiques que nous avons menées sur les altérations de connectivité cérébrale, au repos, chez les patients schizophrènes, dépressifs et bipolaires. Ces revues révèlent que, malgré des avancées scientifiques considérables dans l’étude de l’altération de la connectivité cérébrale fonctionnelle, la dimension temporelle des mécanismes cérébraux à l’origine de l’atteinte de l’intégration de l’information dans ces maladies, particulièrement de la schizophrénie, est encore mal comprise. Par conséquent, le deuxième objectif de cette thèse est de caractériser les changements cérébraux associés à la schizophrénie dans le domaine temporel. Nous présentons deux études dans lesquelles nous testons l’hypothèse que la « disconnectivité temporelle » serait un biomarqueur important en schizophrénie. Ces études explorent les déficits d’intégration temporelle en schizophrénie, en quantifiant les changements de la dynamique neuronale dite invariante d’échelle à partir des données magnétoencéphalographiques (MEG) enregistrés au repos chez des patients et des sujets contrôles. En particulier, nous utilisons (1) la LRTCs (long-range temporal correlation, ou corrélation temporelle à longue-distance) calculée à partir des oscillations neuronales et (2) des analyses multifractales pour caractériser des modifications de l’activité cérébrale arythmique. Par ailleurs, nous développons des modèles de classification (en apprentissage-machine supervisé) pour mieux cerner les attributs corticaux et sous-corticaux permettant une distinction robuste entre les patients et les sujets sains. Vu que ces études se basent sur des données MEG spontanées enregistrées au repos soit avec les yeux ouvert, ou les yeux fermées, nous nous sommes par la suite intéressés à la possibilité de trouver un marqueur qui combinerait ces enregistrements. La troisième étude originale explore donc l’utilité des modulations de l’amplitude spectrale entre yeux ouverts et fermées comme prédicteur de schizophrénie. Les résultats de ces études démontrent des changements cérébraux importants chez les patients schizophrènes au niveau de la dynamique d’invariance d’échelle. Elles suggèrent une dégradation du traitement temporel de l’information chez les patients, qui pourrait être liée à leurs symptômes cognitifs et comportementaux. L’approche multimodale de cette thèse, combinant la magétoencéphalographie, analyses non-linéaires et apprentissage machine, permet de mieux caractériser l’organisation spatio-temporelle du signal cérébrale au repos chez les patients atteints de schizophrénie et chez des individus sains. Les résultats fournissent de nouvelles preuves supportant l’hypothèse d’une « disconnectivité temporelle » en schizophrénie, et étendent les recherches antérieures, en explorant la contribution des structures cérébrales profondes et en employant des mesures non-linéaires avancées encore sous-exploitées dans ce domaine. L’ensemble des résultats de cette thèse apporte une contribution significative à la quête de nouveaux biomarqueurs de la schizophrénie et démontre l’importance d’élucider les altérations des propriétés temporelles de l’activité cérébrales intrinsèque en psychiatrie. Les études présentées offrent également un cadre méthodologique pouvant être étendu à d’autres psychopathologie, telles que la dépression. / Psychiatric disorders affect nearly a quarter of the world’s population. These typically bring about debilitating behavioural, functional and/or cognitive problems, for which the underlying neural mechanisms are poorly understood. These symptoms can significantly reduce the quality of life of affected individuals, impact those close to them, and bring on an economic burden on society. Hence, targeting the baseline neurophysiology associated with psychopathologies, by identifying more robust biomarkers, would improve the development of effective treatments. The first goal of this thesis is thus to contribute to a better characterization of neural dynamic alterations in mental health illnesses, specifically in schizophrenia and mood disorders. Accordingly, the first chapter of this thesis presents two systematic literature reviews, which investigate the resting-state changes in brain connectivity in schizophrenia, depression and bipolar disorder patients. Great strides have been made in neuroimaging research in identifying alterations in functional connectivity. However, these two reviews reveal a gap in the knowledge about the temporal basis of the neural mechanisms involved in the disruption of information integration in these pathologies, particularly in schizophrenia. Therefore, the second goal of this thesis is to characterize the baseline temporal neural alterations of schizophrenia. We present two studies for which we hypothesize that the resting temporal dysconnectivity could serve as a key biomarker in schizophrenia. These studies explore temporal integration deficits in schizophrenia by quantifying neural alterations of scale-free dynamics using resting-state magnetoencephalography (MEG) data. Specifically, we use (1) long-range temporal correlation (LRTC) analysis on oscillatory activity and (2) multifractal analysis on arrhythmic brain activity. In addition, we develop classification models (based on supervised machine-learning) to detect the cortical and sub-cortical features that allow for a robust division of patients and healthy controls. Given that these studies are based on MEG spontaneous brain activity, recorded at rest with either eyes-open or eyes-closed, we then explored the possibility of finding a distinctive feature that would combine both types of resting-state recordings. Thus, the third study investigates whether alterations in spectral amplitude between eyes-open and eyes-closed conditions can be used as a possible marker for schizophrenia. Overall, the three studies show changes in the scale-free dynamics of schizophrenia patients at rest that suggest a deterioration of the temporal processing of information in patients, which might relate to their cognitive and behavioural symptoms. The multimodal approach of this thesis, combining MEG, non-linear analyses and machine-learning, improves the characterization of the resting spatiotemporal neural organization of schizophrenia patients and healthy controls. Our findings provide new evidence for the temporal dysconnectivity hypothesis in schizophrenia. The results extend on previous studies by characterizing scale-free properties of deep brain structures and applying advanced non-linear metrics that are underused in the field of psychiatry. The results of this thesis contribute significantly to the identification of novel biomarkers in schizophrenia and show the importance of clarifying the temporal properties of altered intrinsic neural dynamics. Moreover, the presented studies offer a methodological framework that can be extended to other psychopathologies, such as depression.

Introduction : L’altération des systèmes de récompense et de douleur serait impliquée dans le développement et le maintien de la douleur chronique. En contexte expérimental, l’arrêt d’une stimulation douloureuse et désagréable peut déclencher une sensation de plaisir (pleasant pain relief, PPR), mais contrairement aux mécanismes endogène d’inhibition de douleur (ICPM) la variabilité interindividuelle du PPR semble ne jamais avoir été étudiée. Objectifs : Étude 1) Investiguer la variabilité interindividuelle du PPR et de l’ICPM chez des sous-groupes de sujets sains ayant des réponses de douleur dynamiques similaires (analyses de trajectoires) durant un test à l’eau froide (CPT) Étude 2) Mesurer la connectivité au repos de régions du système de récompense en IRMf en lien avec le PPR. Méthode : 1) Une thermode Peltier et un CPT (10 ℃) ont été administrés séquentiellement (N=122). Le PPR a été mesuré pendant quatre minutes après l’arrêt du CPT et, 2) suite à l’application d’un gel froid en imagerie fonctionnelle (N=38). Résultats : 1) Quatre trajectoires ont été identifiées selon les réponses de douleur durant le CPT. Le PPR était corrélé aux trajectoires de douleur, mais pas l’efficacité d’ICPM. 2) Les connexions au repos, significativement corrélées au PPR, étaient les suivantes : noyau accumbens (NAcc) gauche-cortex cingulaire postérieur, NAcc gauche-cortex cingulaire antérieur dorsal, amygdale gauche-cortex préfrontal dorsolatéral, NAcc droit-cervelet crus II gauche et droit et cortex préfrontal ventromédian-cervelet crus II droit. Discussion : La sensibilisation et l’aspect désagréable moyen de douleur durant une stimulation nociceptive tonique ainsi que la connectivité au repos entre des régions modulant le plaisir (récompense) et la cognition affectent le degré de PPR. / Introduction: Alteration of reward and pain systems is related to the development of chronic pain. In experimental settings, the cessation of a painful and unpleasant stimulation has been shown to elicit a pleasant pain relief (PPR), but contrary to inhibitory conditioned pain modulation mechanisms (ICPM) inter-individual variability of PPR appears to have never been studied. Objectives: Study 1) Investigate inter-individual variability of PPR and ICPM in subgroups of healthy subjects with similar dynamic pain responses (trajectory analyses) during a cold pressor test (CPT) Study 2) Measure resting state functional connectivity with regions of the reward system related to PPR. Method: 1) Peltier thermode and CPT (10℃) were administered sequentially (N=122). The PPR was measured for four minutes after CPT offset 2) PPR was measured (N=38) following the application of a cold gel during functional imaging. Results: 1) Four trajectories were identified based on pain responses during CPT. PPR was correlated with pain trajectories, but not the efficacy of ICPM. 2) Resting state connections, significantly correlated with PPR, were the following: left nucleus accumbens (NAcc)-posterior cingulate cortex, left NAcc-dorsal anterior cingulate cortex, left amygdala-dorsolateral prefrontal cortex, right NAcc-left and right cerebellum crus II and ventromedial prefrontal cortex-right cerebellum crus II. Discussion: Sensitization and average pain unpleasantness during tonic nociceptive stimulation, as well as connectivity between regions modulating pleasure (reward) and cognition, affect the degree of PPR.

[pt] O agonista do receptor do peptídeo semelhante ao glucagon 1 (GLP-1), melhora o controle glicêmico, reduz o apetite e o peso corporal, sendo usado para o tratamento de diabetes tipo 2 (DM2). Também se mostrou associado a alterações nas respostas cerebrais, relacionadas a estímulos alimentares. Entretanto, seus efeitos na conectividade funcional intrínseca do cérebro não são conhecidos. Com objetivo de melhor entender o papel do GLP-1 na conectividade intrínseca do cérebro em pacientes DM2, dados de ressonância magnética funcional (RMf) de redes do estado de repouso relevantes para o comportamento alimentar foram analisados em dois estudos. Em ambos, todas as imagens foram adquiridas após um jejum noturno (8-12 horas). O estudo 1 teve como meta investigar o efeito agudo do bloqueio de GLP-1 na conectividade funcional. Foram adquiridas imagens de RMf durante o estado de repouso, em dois dias separados, de 20 pacientes DM2 sem complicações e 20 controles saudáveis, primeiro sob infusão de solução salina e, posteriormente, sob a infusão de antagonista do receptor de GLP-1. Já o estudo 2 teve como objetivo investigar, em pacientes DM2, se haveria diferenças na conectividade intrínseca, quando comparados os tratamentos com agonista do GLP1 liraglutida e com insulina glargina. Os mesmos pacientes DM2, participantes do estudo 1, foram tratados, em ordem aleatória, por 12 semanas com liraglutida e por 12 semanas com insulina glargina. Os dados de RMf em estado de repouso foram coletados antes do início do tratamento, após 10 dias e após 12 semanas. As análises de neuroimagem foram corrigidas para múltiplas comparações com o Family-wise error, as correlações foram feitas com coeficiente de correlação de Pearson. Os resultados do estudo 1 mostraram que, durante a infusão da solução salina, pacientes DM2 apresentaram maior conectividade comparados a controles na ínsula esquerda e opérculo, relacionada à maior perda de peso, mediada pelo agonista de GLP-1 após 10 dias e 12 semanas. Além disso, a conectividade foi maior em pacientes DM2 versus controles no polo frontal, córtex frontal medial, no giro cingulado anterior e no giro paracingulado, a qual se correlacionou com menor perda de peso, mediada por agonista de GLP-1, após 10 dias (todos P(FWE) menor que 0,05). Não houve efeito da infusão do antagonista do receptor de GLP-1 ou do tratamento com agonista de GLP-1, na conectividade (todos P(FWE) maior que 0,05). Em conclusão, a conectividade basal em estado de repouso mostrou estar relacionada à mudança de peso, mediada pelo agonista do GLP-1, com maior conectividade frontal correlacionando com menos perda de peso durante o tratamento com agonista do GLP-1, enquanto maior conectividade na ínsula esquerda, correlacionou com maior perda de peso, mediada pelo GLP-1, indicando relação entre a conectividade intrínseca dessas redes e o efeito de perda de peso do tratamento com GLP-1. / [en] The glucagon-like peptide 1 (GLP-1) receptor agonist is used for the treatment of type 2 diabetes (DM2) as it improves glycemic control, reduces appetite and body weight. It is also related to altered brain responses to food stimuli, but its effects on intrinsic brain connectivity are unknown. With the goal of better understanding GLP-1 s role in the intrinsic brain connectivity of DM2 patients, functional resonance imaging (fMRI) data of resting-state networks relevant for eating behavior was analyzed in two studies. In both, all images were acquired after an overnight fast (8-12 hours). Study 1 aimed to investigate the acute effect of GLP1 blockade on functional connectivity. On two separate days, fMRI data was acquired from 20 DM2 patients and 20 healthy controls, first under saline infusion and thereafter under GLP-1 antagonist infusion. Study 2 aimed to investigate, in DM2 patients, if there were any between treatment differences in intrinsic connectivity when comparing GLP-1 receptor agonist liraglutide with insulin glargine. The same DM2 participants in study 1 were thus treated in random order for 12 weeks with liraglutide and insulin glargine, fMRI data was collected at the start of treatment, after 10 days and after 12 weeks. Study 1 results showed that, during saline infusion, DM2 patients had greater connectivity compared to controls in the left insula and operculum, which related to greater GLP-1 mediated weightloss after 10 days and 12 weeks. Also, connectivity was greater in DM2 patients versus controls in the frontal pole, frontal medial cortex, anterior cingulate and paracingulate giry, which related to less GLP-1 mediated weight-loss after 10 days (all P(FWE) less than 0.05). There was no effect on connectivity for GLP-1 antagonist, and no long-term differences between treatments (all P(FWE) less than 0.05). In conclusion, baseline resting-state connectivity was shown to be related to GLP-1 mediated weightchange, with greater frontal connectivity relating to less weight loss during GLP-1 treatment, while higher left insula connectivity correlated to greater weight loss during GLP-1 treatment, indicating a relationship between baseline intrinsic connectivity in these regions and weight loss during GLP-1 treatment.

Bien que la majeure partie du génome humain soit présente en deux copies (une copie héritée de chaque parent), certains segments peuvent être délétés (une copie) ou dupliqués (trois copies). La recherche a montré que plusieurs variations du nombre de copies (CNVs) augmentent le risque de troubles neurodéveloppementaux (e.g. autisme, TDAH, schizophrénie). Or, on connait peu les effets des CNVs sur le développement et le fonctionnement cérébral. L’électroencéphalographie (EEG) au repos s’avère être une méthode adaptée pour étudier les perturbations de l’activité neuronale chez les porteurs de CNVs. L’objectif de ce projet était de déterminer s’il existe des signatures EEG à l’état de repos qui sont caractéristiques des enfants porteurs de CNVs pathogéniques. L’activité cérébrale au repos de 109 porteurs de CNVs (66 délétions, 43 duplications) âgés de 3 à 17 ans a été enregistrée en EEG durant 4 minutes. Pour mieux prendre en compte les variations développementales, les indices EEG (puissance spectrale et connectivité fonctionnelle) ont été corrigés avec un modèle normatif estimé à partir de 256 contrôles du Heatlhy Brain Network. Les résultats ont montré une puissance bêta et gamma accrue dans les régions postérieures ainsi qu’une sous-connectivité globale à des échelles temporelles distinctes chez les porteurs de CNVs. Les porteurs d’une délétion et d’une duplication pouvaient être différenciés par leur connectivité dans les fréquences bas-alpha: la connectivité des porteurs d’une duplication était plus perturbée que celle des porteurs d’une délétion. Les perturbations distinctives en connectivité se sont avérées plus proéminentes à l’adolescence. Les résultats suggèrent que les porteurs de CNVs présentent des altérations électrophysiologiques par rapport aux témoins neurotypiques, indépendamment de la région génomique affectée. / Although most of the human genome is present in two copies (one copy inherited from each parent), some segments can be deleted (one copy) or duplicated (three copies). Research has shown that many copy number variations (CNVs) increase the risk of neurodevelopmental disorders (e.g. autism, ADHD, schizophrenia). However, little is known about the effects of CNVs on brain development and function. Resting-state electroencephalography (EEG) is a suitable method to study the disturbances of neuronal functioning in CNVs. We aimed to determine whether there are resting-state EEG signatures that are characteristic of children with pathogenic CNVs. Resting-state brain activity of 109 CNVs carriers (66 deletions, 43 duplications) aged 3 to 17 years was recorded in EEG for 4 minutes. To better account for developmental variations, EEG indices (power spectral density and functional connectivity) were corrected with a normative model estimated from 256 Heatlhy Brain Network controls. Results showed increased beta and gamma power in posterior regions as well as a global under-connectivity at distinct frequency bands in CNVs carriers. Deletion and duplication carriers can be differentiated by their connectivity in low alpha frequencies: the connectivity of the duplication carriers was more disrupted than that of the deletion carriers. The distinctive connectivity perturbations were found to be most prominent during adolescence. The results suggest that CNVs carriers show electrophysiological alterations compared to neurotypical controls, regardless of the gene dosage effect and of their affected genomic region. Moreover, a specific signature of the molecular alterations associated with deletions was found.

A significant subset of women plagued with breast cancer continue to experience chronic neuropathic pain (CNP) long after undergoing cancer treatment. Medical interventions such as pharmacotherapy and/or surgery have been most widely used to abate painful symptoms with limited efficacy. Other alternatives are required given a heavy reliance on pharmaceuticals can lead to tolerance, dependence and severe side effects. Possibilities include cognitive behavioural therapy (CBT), physical therapy, and mindfulness interventions to supplement pharmacotherapies. Mindfulness practice in particular has been offered to a variety of chronic pain groups including breast cancer patients, however evidence is lacking to support its effectiveness in CNP for breast cancer survivors (BCS). The purpose of the present study was to explore the benefits a mindfulness-based stress reduction program (MBSR) may have on altering the underlying neuronal correlates linked with pain-related symptoms associated with CNP in BCS. The primary objective was to investigate how mindfulness training might possibly mediate the brain’s capacity for emotional reactivity, white matter integrity, and activation of the default mode network (DMN) and how these changes may correlate with levels of pain severity and pain interference, improving overall quality of life. To achieve these results, several brain imaging techniques were used in order to observe the correlation between the subjective experience of pain and the objective manifestation of brain changes that may be potentiated by MBSR training. A total of 23 participants were placed in either an 8 week MBSR intervention group (n=13) or a waitlist control group (n =10). All women were scanned with MRI before and after the 8 week intervention regardless of group allotment. The following neuroimaging modalities were used for each scanning session: resting state fMRI (rsfMRI) to monitor changes to functional connectivity in the default mode network (DMN); Diffusion Tensor Imaging (DTI) to assess the structural integrity of white matter tracts; and the Emotional Stroop Task (EST) to examine emotional reactivity in response to pain related stimuli. Exploratory results from this pilot study indicate that improvements to functional connectivity were apparent in the MBSR group relative to control, indicative of more efficient communication in areas related to attention, self-awareness, emotion regulation and pain. Improvements were also noted as increased cerebral white matter health and reduced emotional reactivity to pain related stimuli in the group of MBSR trained participants relative to control. Additionally, these functional and structural changes correlated with the self-reported pain measures in the MBSR group, suggesting that the MBSR group demonstrated improvements to ratings of pain severity and pain interference whereas the opposite occurred with the control group. The results have been interpreted as improvements to patients’ perception of pain and quality of life post MBSR training, however, were not limited to the subjective experience of pain. The inclusion of neuroimaging modalities provides objective and empirical support for MBSR training as it highlights the underlying brain mechanisms that were altered as part of MBSR treatment. Ultimately, the evidence suggests that MBSR could be incorporated as part of the treatment protocol for women experiencing CNP post breast cancer treatment.

Numerous studies demonstrate that moment-to-moment neural variability is behaviorally relevant and beneficial for tasks and behaviors requiring cognitive flexibility. However, it remains unclear whether the positive effect of neural variability also holds for cognitive persistence. Moreover, different brain variability measures have been used in previous studies, yet comparisons between them are lacking. In the current study, we examined the association between resting-state BOLD signal variability and two metacontrol policies (i.e., persistence vs. flexibility). Brain variability was estimated from resting-state fMRI (rsfMRI) data using two different approaches (i.e., Standard Deviation (SD), and Mean Square Successive Difference (MSSD)) and metacontrol biases were assessed by three metacontrol-sensitive tasks. Results showed that brain variability measured by SD and MSSD was highly positively related. Critically, higher variability measured by MSSD in the attention network, parietal and frontal network, frontal and ACC network, parietal and motor network, and higher variability measured by SD in the parietal and motor network, parietal and frontal network were associated with reduced persistence (or greater flexibility) of metacontrol (i.e., larger Stroop effect or worse RAT performance). These results show that the beneficial effect of brain signal variability on cognitive control depends on the metacontrol states involved. Our study highlights the importance of temporal variability of rsfMRI activity in understanding the neural underpinnings of cognitive control.

While task-based functional magnetic resonance imaging (fMRI) has helped us understand the functional role of many regions in the human brain, many diseases and complex behaviors defy explanation. Alternatively, if no task is performed, the fMRI signal between distant, anatomically connected, brain regions is similar over time. These correlations in “resting state” fMRI have been strongly linked to behavior and disease. Previous work primarily calculated correlation in entire fMRI runs of six minutes or more, making understanding the neural underpinnings of these fluctuations difficult. Recently, coordinated dynamic activity on shorter time scales has been observed in resting state fMRI: correlation calculated in comparatively short sliding windows and quasi-periodic (periodic but not constantly active) spatiotemporal patterns. However, little relevance to behavior or underlying neural activity has been demonstrated. This dissertation addresses this problem, first by using 12.3 second windows to demonstrate a behavior-fMRI relationship previously only observed in entire fMRI runs. Second, simultaneous recording of fMRI and electrical signals from the brains of anesthetized rats is used to demonstrate that both types of dynamic activity have strong correlates in electrophysiology. Very slow neural signals correspond to the quasi-periodic patterns, supporting the idea that low-frequency activity organizes large scale information transfer in the brain. This work both validates the use of dynamic analysis of resting state fMRI, and provides a starting point for the investigation of the systemic basis of many neuropsychiatric diseases.