Dissertations / Theses on the topic 'Neuroimaging'

The application of machine learning algorithms to analyze and determine disease related patterns in neuroimaging has emerged to be of extreme interest in Computer-Aided Diagnosis (CAD). This study is a small step towards categorizing Alzheimer's disease, Neurode-generative diseases, Psychiatric diseases and Cerebrovascular Small Vessel diseases using CAD. In this study, the SPECT neuroimages are pre-processed using powerful data reduction techniques such as Singular Value Decomposition (SVD), Independent Component Analysis (ICA) and Automated Anatomical Labeling (AAL). Each of the pre-processing methods is used in three machine learning algorithms namely: Artificial Neural Networks (ANNs), Support Vector Machines (SVMs) and k-Nearest Neighbors (k-nn) to recognize disease patterns and classify the diseases. While neurodegenerative diseases and psychiatric diseases overlap with a mix of diseases and resulted in fairly moderate classification, the classification between Alzheimer's disease and Cerebrovascular Small Vessel diseases yielded good results with an accuracy of up to 73.7%.

Chronic pain is a debilitating symptom of a wide range of conditions. These conditions are both highly prevalent and create adverse consequences for individuals and society. Whilst understanding of chronic pain conditions has improved, in a number of cases the mechanisms of chronic pain are not fully understood and no cure is available. It is appreciated that chronic pain is not only unpleasant in itself, but can also lead to a reorganisation of the nervous system resulting in further suffering. These factors present a justification for further investigation into the mechanisms and effects of chronic pain to enable progress towards more effective treatments. Neuroimaging techniques have helped our understanding the mechanisms and effects of chronic pain. Techniques have been developed to examine the structure, chemistry and activity of the brain. This thesis describes investigations that used neuroimaging to examine the effects of chronic pain on the human brain. A distinction has been drawn between chronic widespread pain (CWP) and chronic localised pain (CLP). Historically, the latter was seen as a condition of the peripheral components of the pain system. More recently, however, an understanding has been gained that central mechanisms may also be a factor in these conditions. The purpose of my investigations was to examine differences and similarities in the effects of two CWP and CLP conditions on the human brain. Fibromyalgia (FM) and Knee Osteoarthritis (OA) were chosen as representatives of these classes of condition. The effects on neurochemistry, brain structure and coordinated brain activity in these conditions were compared using magnetic resonance spectroscopy (MRS), voxel-based morphometry (VBM) and resting state functional connectivity (rs-FC). Using MRS I observed a reduction in N-Acetylaspartic acid (NAA) in the thalamus of OA patients when compared to FM. Using VBM I observed that grey matter volume (GMV) was reduced in the left brainstem and posterior cingulate cortex in FM patients when compared to OA. GMV was reduced in the left precentral, middle frontal and supramarginal gyri in OA when compared to FM. Using rs-FC I observed an increase in functional connectivity in the default mode network of FM patients when compared to OA. I observed increased functional connectivity within the default mode network (DMN) in both pain conditions compared to healthy controls. I also observed increased functional connectivity between the precuneus and regions in both the DMN and executive attention networks. Consideration is given to these findings in the context of previous relevant research. The implications of the results are related the patients’ own experience of their condition and links to clinical measures are also discussed. The findings provide further evidence for the neural basis of elements of patients’ experience of their condition and further understanding of the differences between the wider presentations of these conditions. The findings are drawn together to demonstrate where the effects of CWP and CLP overlap and where their effects contrast. Consideration is given to the mechanisms at work in these conditions that suggest differing effects on the components of the pain system and also to demonstrate where prolonged abnormal peripheral input may be a factor driving adaptation in CLP.

Overweight and stress interact in complex ways. Excess weight promotes chronic low-grade inflammatory states that can mobilise the hypothalamic-pituitary-adrenal (HPA) axis. HPA axis activation resulting from frequent stress situations can modify energy uptake and expenditure. Separately, both conditions have been linked to changes in brain integrity and executive performance. The organism adapts to situations of caloric surplus through boosting immune, neuroendocrine and cardiometabolic systems to restore energy homeostasis. The allostatic load model establishes that the cumulative effects of adapting to challenging scenarios may result in adverse health situations in the future. There is sufficient evidence to consider that a state of overweight is inherently linked to a higher chronic physiological stress, or allostatic load. Our hypothesis was that, independently of the effects of visceral adiposity, the aggregated effects of the biological alterations related to overweight would be enough detrimental to brain structure and executive functioning. Lean-to-obese volunteers aged 21 to 40 years were recruited from primary health care centres belonging to the Consorci Sanitari de Terrassa. Subjects underwent a medical and neuropsychological examination, as well as a magnetic resonance imaging acquisition at the Hospital Clínic de Barcelona. The allostatic load index consisted of the sum of several biomarkers representing physiological stress. Overweight subjects had a greater allostatic load than healthy weight participants. The allostatic load escalation was negatively correlated with the morphology of cortical areas and tracts known to be ascribed to circuits involved in cognitive control, reward-processing and the integration of visceral-sensory signalling. Finally, the intensification in this index correlated with worse cognitive flexibility.
El sobrepès i l'estrès interactuen de formes complexes. L'excés de pes promou estats inflamatoris crònics de baix grau que poden mobilitzar l'eix hipotalàmic-pituitari-adrenal (HPA). L'activació de l'eix HPA resultant de situacions d'estrès freqüents pot modificar la captació i la despesa d'energia. Les dues condicions s'han vinculat per separat a canvis en la integritat cerebral i l'acompliment executiu. L'organisme s'adapta a situacions de superàvit calòric a través de impulsar sistemes immunes, neuroendocrins i cardiometabòlics per restaurar l'homeòstasi energètica. El model de càrrega alostàtica estableix que els efectes acumulatius de l'adaptació a escenaris desafiadors poden resultar en situacions adverses per a la salut en el futur. Hi ha evidència suficient per a considerar que un estat de sobrepès està inherentment vinculat a un major estrès fisiològic crònic, o càrrega alostàtica. La nostra hipòtesi va ser que, independentment dels efectes de l'adipositat visceral, els efectes agregats de les alteracions biològiques relacionades amb l'excés de pes resultarien suficientment perjudicials per a la estructura cerebral i el funcionament executiu. Es van reclutar voluntaris amb normopès i sobrepès amb edats compreses entre els 21 i els 40 anys de centres d'atenció primària de salut pertanyents al Consorci Sanitari de Terrassa. Els subjectes es van sotmetre a un examen mèdic i neuropsicològic, així com a l'adquisició d'imatges per ressonància magnètica a l'Hospital Clínic de Barcelona. L'índex de càrrega alostàtica va consistir en la suma de diversos biomarcadors representant estrès fisiològic. Els subjectes amb sobrepès van presentar major càrrega alostàtica que els participants de pes saludable. L'escalada de càrrega alostàtica es va correlacionar negativament amb la morfologia d'àrees corticals i tractes coneguts per estar adscrits a circuits implicats en el control cognitiu, el processament de recompenses i la integració de la senyalització visceral-sensorial. Finalment, la intensificació en l'esmentat índex va correlacionar amb una pitjor flexibilitat cognitiva.
El sobrepeso y el estrés interactúan de formas complejas. El exceso de peso promueve estados inflamatorios crónicos de bajo grado que pueden movilizar el eje hipotalámico- pituitario-adrenal (HPA). La activación del eje HPA resultante de situaciones de estrés frecuentes puede modificar la captación y el gasto de energía. Ambas condiciones se han vinculado por separado a cambios en la integridad cerebral y el desempeño ejecutivo. El organismo se adapta a situaciones de superávit calórico a través de varias modificaciones fisiológicas. Esto incluye impulsar sistemas inmunes, neuroendocrinos y cardiometabólicos para restaurar la homeostasis energética. El modelo de carga alostática establece que los efectos acumulativos de la adaptación a escenarios desafiantes pueden resultar en situaciones adversas para la salud en el futuro. Existe evidencia suficiente para considerar que un estado de sobrepeso está inherentemente vinculado a un mayor estrés fisiológico crónico, o carga alostática. Nuestra hipótesis fue que, independientemente de los efectos de la adiposidad visceral, los efectos agregados de las alteraciones biológicas relacionadas con el sobrepeso resultarían suficientemente perjudiciales para la estructura cerebral y el funcionamiento ejecutivo. Se reclutaron voluntarios con normopeso y sobrepeso con edades comprendidas entre los 21 y los 40 años de centros de atención primaria de salud pertenecientes al Consorci Sanitari de Terrassa. Los sujetos se sometieron a un examen médico y neuropsicológico, así como a la adquisición de imágenes por resonancia magnética en el Hospital Clínic de Barcelona. El índice de carga alostática consistió en la suma de varios biomarcadores que representan estrés fisiológico. Los sujetos con sobrepeso presentaron mayor carga alostática que los participantes de peso saludable. La escalada de carga alostática se correlacionó negativamente con la morfología de áreas corticales y tractos conocidos por estar adscritos a circuitos implicados en el control cognitivo, el procesamiento de recompensas y la integración de la señalización visceral-sensorial. Finalmente, la intensificación en dicho índice correlacionó con una peor flexibilidad cognitiva.

Nuclear imaging enables quantitative measurements of biological processes in vivo and has revolutionised biomedical research, drug development and clinical practice. Despite the advances made in this field, the ability to image fundamental aspects of neurological diseases remains a challenge. This is partly due to the limited availability of radiotracers for imaging excitatory neurotransmission and detection of inflammation as well as an array of other biochemical processes central to the operational function of the brain. The aim of this research was to expand the arsenal of radiotracers available for neuroimaging in order to study key pathological processes involved in neurological diseases. With the aim to target neuronal Voltage Gated Sodium Channels (VGSCs), Vascular Cell Adhesion Molecule – 1 (VCAM-1) and N-methyl-D-Aspartate Receptors (NMDARs), radiotracers have been synthesised and evaluated. Abnormal expression of these receptors has been implicated in a number of pathological conditions including epilepsy, multiple sclerosis and neurodegeneration. The radiotracers were characterised and evaluated via in vivo imaging (MRI and SPECT/CT) and ex-vivo studies (phosphorimaging, biodistribution and metabolite analysis) in order to determine if they hold significant potential as tools to study neuronal pathways as well as for diagnostic imaging and treatment monitoring. Iodinated analogues of the iminodihydroquinoline WIN17317-3, and the 1-benzazepin-2-one BNZA have been evaluated as neuronal VGSC tracer candidates in healthy mice. Whilst the WIN17317-3 analogue suffered from poor brain uptake and was rapidly metabolised in vivo, the BNZA analogue exhibited excellent in vivo stability and its promising uptake in the brain warrants further investigations. Even though N-(1-Napthyl)-N’-(3-[123I]-iodophenyl)-N’-methylguanidine ([123I]CNS-1261) has demonstrated favourable pharmacokinetics for brain imaging in clinical studies, [125I]CNS-1261 was not successful in discriminating NMDAR expression between naïve rats and those induced with status epilepticus using lithium and pilocarpine. Promisingly, a multi modal contrast agent comprising micron sized particles of iron oxide conjugated to I-125 radiolabelled antibodies, highlighted the up-regulation of VCAM-1 in rat models of cerebral inflammation and in the lithium pilocarpine model of status epilepticus. This versatile imaging agent presents an exciting opportunity to identify an early biomarker for epileptogenesis.

 Background and Objectives: ‘Neurodevelopmental disorders’ is often synonymously used with childhood developmental disorders such as autism spectrum disorder (ASD), however, increasingly new lines of evidence from genetics and epidemiology suggests having schizophrenia and bipolar disorder to be included as well. For example, there is a strong tendency for schizophrenia and bipolar disorder to occur in people with ASD and shared aetiological factors such as prenatal infection and maternal vitamin D deficiency during pregnancy have all been linked with increased risks in all three conditions. To investigate into this, I have turned to brain imaging, a technique which has opened up a new horizon for neurobiologists. Typically, neuroimaging studies focus on one disorder, matching patients with healthy volunteers and compare their brain structures volumetric differences. On the other hand, such studies are limited by various factors including small ample size, low power, no psychiatric control group, and sample or design heterogeneity. Methods: To summarize all the data into a more meaningful biological representation, Anatomical Likelihood Estimation (ALE), a cutting edge meta-analytic approach was applied. The rationale behind ALE is that it identifies brain differences most consistently reported across studies, while filtering away differences that are least documented. In this thesis, a novel application of ALE known as “dual disorder ALE” is introduced, which serves to estimate the extent of brain regional differences implicated in either disorder – in other words, a method to quantify which areas of the brain are more likely to be affected by ASD, schizophrenia or bipolar disorder. Findings: The analysis is separated into two parts. First, dual disorder ALE technique was applied to investigate the relationship between ASD and first-episode schizophrenia. Data from 25 MRI studies was extracted comprising 660 participants (308 ASD, 352 schizophrenia) and 801 healthy controls. In ASD and FE schizophrenia, there were similar brain differences near the limbic-striato-thalamic circuitry, and distinctive brain differences including amygdala, caudate, frontal and medial gyrus for schizophrenia and putamen for ASD. In the second part comparing bipolar disorder and schizophrenia, data from 651 schizophrenic patients, 540 bipolar patients, and 1438 healthy controls was used, and matched one-to-one by pairing up bipolar disorder studies with corresponding schizophrenia studies to minimize confounders. The ALE result indicated that there are substantial overlaps across the two disorders, with schizophrenia having more extensive brain differences than bipolar disorder. Conclusions: Both parts of the analysis suggest that there are similar aetiological pressures affecting neurodevelopmental disorders including ASD, schizophrenia and bipolar disorder.
published_or_final_version
Psychiatry
Master
Master of Philosophy

Lacunar stroke accounts for one quarter of all ischaemic stroke and in the long term carries a greater risk of death and disability than was previously realised. Much of our current knowledge originated from neuropathological studies in the 1950s and 1960s. In the last thirty years, brain computed tomography (CT) and magnetic resonance imaging (MRI) have revolutionised our understanding of lacunar stroke and associated features of cerebral small vessel disease (SVD), namely white matter lesions (WML), enlarged perivascular spaces (EPVS) and brain microbleeds (BMB). The purpose of the projects which led to the writing of this thesis was to improve understanding of imaging characteristics of cerebral SVD. We aimed to assess (i) clinical and imaging features which might explain misclassification of lacunar infarcts as cortical infarcts and vice versa, (ii) the proportion of symptomatic lacunar infarcts progressing to lacunar cavities and associations of cavitation, (iii) completeness of reporting of lacunar lesions in the lacunar stroke literature, (iv) definitions and detection of lacunar lesions amongst SVD researchers, (v) the relationship between WML and carotid stenosis, (vi) clinical and imaging associations of EPVS and, (vii) observer variability in the assessment of EPVS and BMB, in order to develop visual rating scales. Section one describes neuroimaging of lacunar stroke. To investigate features which might explain clinical stroke subtype misclassification (‘clinical-imaging dissociation’), I used data from a stroke study. The main factor associated with clinical-imaging dissociation was diabetes, and in patients with acute lacunar infarction, proximity of the lacunar infarct to the cortex, age, diabetes and left hemisphere location. To investigate the proportion of symptomatic lacunar infarcts progressing to cavities, I used data from two stroke studies. A fifth of patients with acute lacunar ischaemic stroke showed definite cavitation on follow-up imaging at a median of 227 days; cavitation was associated with increasing time to follow-up. To assess completeness of reporting of lacunar lesions in the lacunar stroke literature, I reviewed 50 articles from three journals with a stroke focus. There was marked variation in terminology and descriptions of imaging definitions of lacunar lesions. To assess lacunar lesion definitions and detection amongst SVD researchers, I used an online survey consisting of case-based and non-case-based questions. There was marked variation in definitions and descriptions. Cavitated lesions were detected with the highest degree of confidence. Section two describes neuroimaging of associated features of cerebral SVD. Using data from two stroke studies, I examined the relationship between WML and ipsilateral carotid artery stenosis. There was no association between carotid stenosis and WML. I tested the association of EPVS with WML and lacunar stroke subtype using data from a stroke study. Total EPVS were associated with age and deep WML; basal ganglia (BG) EPVS were associated with age, centrum semiovale (CS) EPVS, cerebral atrophy and lacunar stroke subtype. Quantification of observer variability in EPVS rating was assessed on 60 MRI scans selected from a stroke study and an ageing cohort. Intrarater agreement was good and interrater agreement was moderate. Main reasons for interrater disagreement included the visualisation of very small EPVS and the presence of concomitant WML and lacunar lesions. Observer variability in BMB rating was quantified using MRI scans from a stroke study. Interrater agreement was moderate but improved following modification of the pilot rating scale (BOMBS; Brain Observer MicroBleed Scale), which had its main effect by differentiating ‘certain’ BMB from ‘uncertain’ BMB and BMB ‘mimics’. In conclusion, neuroimaging, particularly MRI, is a valuable tool for the investigation of lacunar stroke and associated features of cerebral SVD. With recent technological advances in both CT and MRI, neuroimaging will remain central to future SVD studies, hopefully leading to a much improved understanding of this important disease.

Developments in technology have enabled scientists to study brain function in an unprecedented way. Functional neuroimaging is the use of neuroimaging technologies to capture information about the state of a brain, with the goal of studying the relationship between mental functions and brain activity. One such technology is functional magnetic resonance imaging (fMRI), which produces a signal that can be used to create cross-sectional images of the brain. These images can be used to measure brain activity in different sections of the brain. In fMRI recording there is a tradeoff between spatial and temporal resolution. My first contribution in this thesis is to present a novel algorithm for generating cross-sectional images. This is a signal processing problem with high dimensionality, but few measurements. My algorithm uses ideas from sparse modelling because variations in functional MR images are sparse over time in the wavelet domain. It will enable high resolution images to be generated using fewer measurements. Sequences of functional MR images are recorded while subjects perform different tasks. The second contribution of this thesis is a machine learning technique to predict different tasks from the captured fMRI sequences. Existing methods perform poorly at this prediction task due to the curse of dimensionality. I overcome this problem by designing a novel sparse modelling method based on the assumption that the active brain region in response to a target task is sparse in the whole brain area. The final contribution to this thesis is the design of different assessment criteria for selecting the most relevant voxels to interpret the neural activity. The conventional selection method uses the assessment of predictive performance, resulting in many false positive selections due to the small number of samples. To overcome this problem, I introduce the concept of stability. My method selects the relevant voxels using the assessments of both predictive performance and stability, which significantly reduces the selection error while maintaining the predictive performance.

Neuroimaging techniques can give novel insights into the nature of human cognition. We do not wish only to label patterns of activity as potentially associated with a cognitive process, but also to probe this in detail, so as to better examine how it may inform mechanistic theories of cognition. A possible approach towards this goal is to extend EEG 'brain-computer interface' (BCI) tools - where motor movement intent is classified from brain activity - to also investigate visual cognition experiments. We hypothesised that, building on BCI techniques, information from visual object tasks could be classified from EEG data. This could allow novel experimental designs to probe visual information processing in the brain. This can be tested and falsified by application of machine learning algorithms to EEG data from a visual experiment, and quantified by scoring the accuracy at which trials can be correctly classified. Further, we hypothesise that ICA can be used for source-separation of EEG data to produce putative activity patterns associated with visual process mechanisms. Detailed profiling of these ICA sources could be informative to the nature of visual cognition in a way that is not accessible through other means. While ICA has been used previously in removing 'noise' from EEG data, profiling the relation of common ICA sources to cognitive processing appears less well explored. This can be tested and falsified by using ICA sources as training data for the machine learning, and quantified by scoring the accuracy at which trials can be correctly classified using this data, while also comparing this with the equivalent EEG data. We find that machine learning techniques can classify the presence or absence of visual stimuli at 85% accuracy (0.65 AUC) using a single optimised channel of EEG data, and this improves to 87% (0.7 AUC) using data from an equivalent single ICA source. We identify data from this ICA source at time period around 75-125 ms post-stimuli presentation as greatly more informative in decoding the trial label. The most informative ICA source is located in the central occipital region and typically has prominent 10-12Hz synchrony and a -5 μV ERP dip at around 100ms. This appears to be the best predictor of trial identity in our experiment. With these findings, we then explore further experimental designs to investigate ongoing visual attention and perception, attempting online classification of vision using these techniques and IC sources. We discuss how these relate to standard EEG landmarks such as the N170 and P300, and compare their use. With this thesis, we explore this methodology of quantifying EEG neuroimaging data with machine learning separation and classification and discuss how this can be used to investigate visual cognition. We hope the greater information from EEG analyses with predictive power of each ICA source quantified by machine learning separation and classification and discuss how this can be used to investigate visual cognition. We hope the greater information from EEG analyses with predictive power of each ICA source quantified by machine learning might give insight and constraints for macro level models of visual cognition.

Mestrado em Engenharia de Computadores e Telemática,
Brain atlases have been used as reference to classify and tag topological information either structural or functional from brain images. Using atlases, the resulting analysis allows the extraction of semantic information from the existing image data. However the process of classifying and tagging brain images using an atlas is often tedious and mostly dependent on human observation and validation. At the same time, even when available is often difficult to use, namely when using typical query retrieve services in modern imaging repositories (e.g. DICOM based PACS). In this work we propose NEArBy, a solution that provides query and retrieve services based on brain atlas semantics that can be easily integrated in existing DICOM based imaging repositories. Using a web interface, NEArBy supports not only typical DICOM query retrieve searches but also query tokens matching the brain atlas dictionary. To automate the semantic tagging of the brain images we rely on external methods to identify relevant spatial features that are later labelled using standard brain atlas. Being DICOM a tag based standard, atlases related tags are then privately embedded into DICOM files as NEArBy XML descriptors. These descriptors encode the mapping between feature type, spatial location in the atlas and the respective atlas tag. XML encoded tags are also suitable for indexation by a medical imaging Q/R tool such as Dicoogle allowing queries based both on standard DICOM tags and specifically on atlases related tokens included by NEArBy middleware. NEArBy provides a way to perform queries over a medical imaging repository using technical and atlas based topological information. We illustrate the NEArBy potential usage over a set of functional magnetic resonance imaging (fMRI) datasets using the web user interface to formulate the queries with atlas related criteria and access the retrieved results. Several experiments were successfully performed demonstrating the effectiveness in retrieving subjects with activations in similar areas and in specific locations.
Os atlas cerebrais têm vindo a ser utilizados como referência na classificação e identificação de informação topológica, tanto estrutural como funcional, de imagens do cérebro. Com recurso ao atlas, a análise que daí surge permite a extração de informação semântica a partir dos dados existentes na imagem. Contudo, o processo de classificação e catalogação de imagens cerebrais com recurso ao atlas é frequentemente entediante e maioritariamente depende da observação e validação humana. Simultaneamente, mesmo quando disponível, é frequentemente de difícil utilização, nomeadamente quando se faz uso de serviços de consulta e recuperação de informação em repositórios de imagens médicas atuais. (e.g. PACS com base DICOM). Neste trabalho, propomos a NEArBy, a solução que disponibiliza serviços de consulta e recuperação de informação com base na semântica de atlas cerebrais, facilmente integrado em repositórios de imagens médicas DICOM existentes. Recorrendo a uma interface web, a NEArBy suporta, não apenas as típicas buscas de consulta e recuperação, mas também chaves de consulta correspondendo ao dicionário de atlas cerebral. Para automatizar a catalogação semântica das imagens cerebrais, recorremos a métodos externos na identificação de características espaciais relevantes que são posteriormente rotulados usando um atlas cerebral standard. Sendo o DICOM um standard baseado em tags, estas relacionadas com o atlas são, assim, discretamente embebidas em ficheiros DICOM como descritores XML NEArBy. Estes descritores codificam o mapeamento entre o tipo de característica, localização espacial no atlas e a respetiva tag do atlas. As tags codificadas do XLM são também adequadas para a indexação através de uma ferramenta de imagens médicas Q/R, como Dicoogle, permitindo consultas com base, ambos em tags standard de DICOM e em chaves relacionadas com o atlas incluídas no middleware NEArBy. NEArBy permite fazer consultas num repositório de imagens médicas, recorrendo a informação técnica e topológica com base em atlas. Ilustramos a potencial utilização da NEArBy num conjunto de imagens por ressonância magnética funcional (IRMf), utilizando a interface web do utilizador para formular as consultas em critérios relacionados com o atlas e aceder aos resultados daí recuperados. Foram levadas a cabo várias experiências com sucesso, demonstrando a eficácia na recuperação de sujeitos com ativações em áreas similares e em locais específicos.

The aim of this study was to shed light on some of the neurophysiological mechanisms behind visual perception and specifically look into feedback processes that may be taking place during visual processing and also inhibition processes of the visual cortex. The oblique effect is a preference of the visual system for cardinal orientations rather than oblique ones. A recent MEG study (Koelewijn, et al. 2011) finds in V1 an initial inverse oblique effect (80 msec from stim onset) which however later (120msec from stimulus onset) showed a trend towards the classical oblique effect and feedback processes here are suggested taking place from the extrastriate cortex. We look into this using fMRI and interestingly we do manage to find an inverse oblique effect, which indicates that the initial MEG “inverse” effect in V1 is detectable with fMRI even though fMRI does not have the temporal resolution of MEG. Unfortunately in this fMRI study the extrastiate region was not localized. In the 3rd experimental chapter (following up on a study by Edden et al.) we look into the relation of behavioural thresholds and gamma activity in the visual cortex. Here we found in the SAM analysis) for the oblique condition, a positive correlation of the oblique main effect in a cortical location in the medial visual cortex (at a frequency range of 30--‐70 Hz) to behavioural thresholds. However here we did not detected an oblique effect when we compared oblique to cardinal condition. In the final experimental chapter we look into the relation between GABA and training effects using however two GABA scanning protocols (with and without macromolecule suppression). Here we find that training effects depend on GABA concentration (as found in unpublished findings by Edden et al.). Additionally here we find negative correlations with behavioural thresholds and GABA however these are strongest for the untrained sessions.

Current practice for analysing functional neuroimaging data is to average the brain signals recorded at multiple sensors or channels on the scalp over time across hundreds of trials or replicates to eliminate noise and enhance the underlying signal of interest. These studies recording brain signals non-invasively using functional neuroimaging techniques such as electroencephalography (EEG) and magnetoencephalography (MEG) generate complex, high dimensional and noisy data for many subjects at a number of replicates. Single replicate (or single trial) analysis of neuroimaging data have gained focus as they are advantageous to study the features of the signals at each replicate without averaging out important features in the data that the current methods employ. The research here is conducted to systematically develop flexible regression mixed models for single trial analysis of specific brain activities using examples from EEG and MEG to illustrate the models. This thesis follows three specific themes: i) artefact correction to estimate the `brain' signal which is of interest, ii) characterisation of the signals to reduce their dimensions, and iii) model fitting for single trials after accounting for variations between subjects and within subjects (between replicates). The models are developed to establish evidence of two specific neurological phenomena - entrainment of brain signals to an α band of frequencies (8-12Hz) and dipolar brain activation in the same α frequency band in an EEG experiment and a MEG study, respectively.

In daily life, we are immersed in a continuous flow of stimuli targeting each of our different senses. Far from being independently processed, accumulating evidence has been widely documented by studies showing that stimuli from different modalities largely interact. However, despite the increasing interest, the interpretations of the results of experiments studying multisensory interaction are still controversial and the underlying mechanisms remain broadly unknown. The aim of this thesis is to investigate the interactions that occur between the senses of audition and touch. Audiotactile interactions have been far less studied than the ones existing between other modality pairings. Maybe because they go often unnoticed though being well present in many everyday life situations. This thesis focuses mainly on two aspects that concern interactions: understanding the impact of the relative saliency between the stimuli and investigating the mechanism behind perceptual integration. These questions are addressed respectively in two studies conducted by means of magnetoencephalography. The thesis is structured as following: in chapter 1, I provide the theoretical background to my scientific questions. A brief synthesis of the two main studies is presented in chapter 2. The two studies are entirely reported under the form of manuscripts in chapter 4. Finally, in appendix a behavioral study that investigates spatial aspects of AT interactions is reported. Although the results of this study are of pertinence of the project, given the preparatory character and the preliminary state of the study we decided to show them in the appendix rather than include them in the main body of the thesis.

In daily life, we are immersed in a continuous flow of stimuli targeting each of our different senses. Far from being independently processed, accumulating evidence has been widely documented by studies showing that stimuli from different modalities largely interact. However, despite the increasing interest, the interpretations of the results of experiments studying multisensory interaction are still controversial and the underlying mechanisms remain broadly unknown. The aim of this thesis is to investigate the interactions that occur between the senses of audition and touch. Audiotactile interactions have been far less studied than the ones existing between other modality pairings. Maybe because they go often unnoticed though being well present in many everyday life situations. This thesis focuses mainly on two aspects that concern interactions: understanding the impact of the relative saliency between the stimuli and investigating the mechanism behind perceptual integration. These questions are addressed respectively in two studies conducted by means of magnetoencephalography. The thesis is structured as following: in chapter 1, I provide the theoretical background to my scientific questions. A brief synthesis of the two main studies is presented in chapter 2. The two studies are entirely reported under the form of manuscripts in chapter 4. Finally, in appendix a behavioral study that investigates spatial aspects of AT interactions is reported. Although the results of this study are of pertinence of the project, given the preparatory character and the preliminary state of the study we decided to show them in the appendix rather than include them in the main body of the thesis.

Cette thèse d'informatique et de mathématiques s'applique au domaine des neurosciences, et plus particulièrement aux recherches sur la modélisation de l'activité cérébrale humaine par électrophysiologie et imagerie. Dans ce champ, la tendance est actuellement d’expérimenter avec des stimuli naturels, comme le visionnage d’un film ou l’écoute d’une piste audio, et non plus avec des stimuli étroitement contrôlés mais outrageusement simples. L’analyse de ces stimuli « naturels » et de leurs effets demande toutefois de disposer d’une immense quantité d’images, par ailleurs très coûteuses. Sans outils mathématique, identifier l'activité neuronale à partir des données est quasi impossible. Toutefois, ces stimuli sont compliqués à modéliser et à analyser, car l'utilisation de méthodes fondées sur des régressions est limitée par la difficulté de modéliser les stimuli. C'est ce qui motive l'utilisation de méthodes non-supervisées qui ne font pas d'hypothèses sur ce qui déclenche les activations neuronales. Dans cette thèse, nous considérons d'abord le cas du modèle de réponse partagée (MRP), dans lequel les sujets sont supposés partager une réponse commune. Ce modèle est utile pour réduire la dimension des données, mais son entraînement est coûteux pour les données d'imagerie fonctionnelle (IRMf) dont la dimension peut être immense. Nous présentons une version bien plus rapide et beaucoup plus économe en mémoire. Mais le MRP fait des hypothèses irréalistes sur les données d'imagerie. Des hypothèses plus réalistes sont utilisées dans l'analyse en composantes indépendantes (ACI) mais cette méthode est difficile à généraliser aux jeux de données qui contiennent plusieurs sujets. Nous proposons alors une extension de l'ACI appelée ACI multi-vue, fondée sur le principe de maximum de vraisemblance et qui convient à des jeux de données multi-sujets. L’ACI multi-vue a une vraisemblance en forme fermée qui peut être maximisée efficacement. Toutefois, cette méthode suppose la même quantité de bruit pour tous les sujets. Nous présentons donc l’ACI partagée, une généralisation de l’ACI multi-vue qui s'accompagne d'un modèle de bruit plus général. Contrairement à presque tous les modèles fondés sur l'ACI, l’ACI partagée peut séparer des sources gaussiennes et non gaussiennes et propose une estimation optimale des sources communes, qui pondère chaque sujet en fonction de son niveau de bruit estimé. En pratique, l’ACI partagée et l’ACI multi-vue permettent d'obtenir, en magnéto-encéphalographie et en IRMf, une estimation plus fiable de la réponse commune que leurs concurrents. Enfin, nous utilisons l'ACI comme base pour faire de l'augmentation de données. Plus précisément, nous présentons l’ACI conditionnelle, une méthode d'augmentation de données qui exploite la grande quantité de données d'IRMf non étiquetées pour construire un modèle génératif en utilisant seulement un petit nombre de données étiquetées. L’ACI conditionnelle permet d'augmenter de façon appréciable la précision du décodage sur huit grands jeux de données d'IRMf. Nos principaux apports nous semblent consister dans l’accélération de l’entraînement du MRP ainsi que dans l’introduction de deux modèles plus réalistes pour l’analyse de l’activité cérébrale de sujets exposés à des stimuli naturels : l’ACI multi-vue et l’ACI partagée. Enfin, nos résultats sont prometteurs concernant l’utilisation de l’ACI pour faire de l’augmentation de données. Nous présentons pour finir quelques pistes qui pourraient guider des travaux ultérieurs. D’un point de vue pratique, des modifications mineures de nos méthodes pourraient permettre l’analyse des données d’imagerie obtenues sur des sujets au repos en faisant l’hypothèse d’une organisation spatiale partagée. D’un point de vue théorique, les travaux futurs pourraient se concentrer sur la compréhension de la façon dont la réduction de dimensions et l'identification de la réponse partagée peuvent être réalisées conjointement
This thesis in computer science and mathematics is applied to the field ofneuroscience, and more particularly to the mapping of brain activity based on imaging electrophysiology. In this field, a rising trend is to experiment with naturalistic stimuli such as movie watching or audio track listening,rather than tightly controlled but outrageously simple stimuli. However, the analysis of these "naturalistic" stimuli and their effects requires a huge amount of images that remain hard and costly to acquire. Without mathematical modeling, theidentification of neural signal from the measurements is very hard if not impossible. However, the stimulations that elicit neural activity are challenging to model in this context, and therefore, the statistical analysis of the data using regression-based approaches is difficult. This has motivated the use of unsupervised learning methods that do not make assumptions about what triggers brain activations in the presented stimuli. In this thesis, we first consider the case of the shared response model (SRM), wheresubjects are assumed to share a common response. While this algorithm is usefulto perform dimension reduction, it is particularly costly on functional magneticresonance imaging (fMRI) data where thedimension can be very large. We considerably speed up thealgorithm and reduce its memory usage. However, SRM relies on assumptions thatare not biologically plausible. In contrast, independent component analysis (ICA) is more realistic but not suited to multi-subject datasets. In this thesis, we present a well-principled method called MultiViewICA that extends ICA to datasets containing multiple subjects. MultiViewICA is a maximum likelihood estimator. It comes with a closed-formlikelihood that can be efficiently optimized. However, it assumes the same amount of noise for all subjects. We therefore introduce ShICA, a generalization of MultiViewICA that comes with a more general noise model. In contrast to almost all ICA-based models, ShICA can separate Gaussian and non-Gaussian sources and comes with a minimum mean square error estimate of the common sources that weights each subject according to its estimated noise level. In practice, MultiViewICA and ShICA yield on magnetoencephalography and functional magnetic resonance imaging a more reliable estimateof the shared response than competitors. Lastly, we use independent component analysis as a basis to perform data augmentation. More precisely, we introduce CondICA, a data augmentation method that leverages a large amount of unlabeled fMRI data to build a generative model for labeled data using only a few labeled samples. CondICA yields an increase in decoding accuracy on eight large fMRI datasets. Our main contributions consist in the reduction of SRM's training time as well as in the introduction of two more realistic models for the analysis of brain activity of subjects exposed to naturalistic stimuli: MultiViewICA and ShICA. Lastly, our results showing that ICA can be used for data augmentation are promising. In conclusion, we present some directions that could guide future work. From apractical point of view, minor modifications of our methods could allow theanalysis of resting state data assuming a shared spatial organization instead of a shared response. From a theoretical perspective, future work could focus on understanding how dimension reduction and shared response identification can be achieved jointly

Absence seizures in Childhood Absence Epilepsy (CAE) are 5 10 second episodes of impaired consciousness that are characterized on electroencephalography (EEG) by frontally-predominant, 3 4 Hz spike and wave discharges (SWD). The aims of this study were to use simultaneous EEG, functional magnetic resonance imaging (fMRI), and behavioral testing to identify the neural networks involved in absence seizures as well as to examine the timecourse of those ictal fMRI changes. It was hypothesized that absence seizures involve wide-reaching neural networks including the areas traditionally associated with normal attention processing and that absence seizures produce fMRI signal changes not only during the seizure, but before and after it as well. In this study, we recorded 88 absence seizures from a cohort of 42 children with pure CAE. These seizures were recorded as subjects participated in simultaneous EEG-fMRI scanning while engaged in a continuous performance task (CPT) of attentional vigilance or a repetitive tapping task (RTT) requiring repetitive motor activity. Using a novel, voxel-based percent fMRI change analysis combined with a volume of interest analysis, the second-by-second fMRI signal timecourse of the absence seizures were examined across numerous brain regions of interest, from 20 seconds before seizure onset through 40 seconds after seizure onset. EEG frequency analysis revealed seizures with a mean duration of 6.6 seconds and an abrupt onset and ending that were comprised of frontally-predominant, 3 4 Hz SWD. Ictal behavioral testing demonstrated abrupt onset of impairments during periods of SWD. These behavioral impairments were typical of CAE absence seizures in that impairments were greater in the CPT of attentional vigilance (omission error rate, OER = 81%) than in RTT testing (OER = 39 %) (p < 0.003). The ictal fMRI changes we observed varied depending upon the method of fMRI signal analysis used. Using the traditional general liner model, and assuming the standard hemodynamic response (HRF) function, this study replicated results consistent with previous ictal absence fMRI studies showing ictal activations primarily in the thalamus and ictal deactivations in traditional default mode areas. Using a more data-driven, novel voxel-based fMRI percentage change analysis to examine the ictal fMRI timecourse on a second-by-second basis, both ictally as well as pre- and post- ictally, this study, however, demonstrated ictal involvement of diverse brain regions before, during, and after the seizure. Activation was demonstrated up to 16 seconds before seizure onset, starting first in the parietal and orbital-medial frontal cortices and progressing to lateral frontal and lateral temporal cortices followed by the occipital and Rolandic cortices and finally the thalamus. Deactivation followed a similar anatomic progression and lasted up to 17 seconds after the end of SWD. These findings reveal a complex and long-lasting sequence of fMRI changes in CAE absence seizures that are not detectable by conventional HRF modeling and are important in the understanding and eventual treatment of absence seizures associated with CAE.

Introduction: The world’s population is aging at an unprecedented rate. By 2050, the number of adults older than 60 years will double from 10% to 20%. This trend has immense implications, due to the prevalence of impaired physical and cognitive functions among older adults. Therefore, it is important to understand the underlying mechanisms for these impairments and identify effective prevention strategies. White matter hyperintensities (WMHs) are common findings on MRI scans of older adults, and are associated with both physical and cognitive decline. Key risk factors for WMHs are related to metabolic and cardiovascular health. Thus, due to the established and significant benefit of targeted exercise training on metabolic and cardiovascular health in older adults, we hypothesized that one mechanism by which exercise, and specifically resistance training (RT), promotes physical and cognitive functions is by reducing WMH progression among older adults. Methods: We explored the associations between WMHs and physical and cognitive functions in Chapters 2 to 4. In Chapter 5, we presented a randomized controlled trial of 52-week RT. Participants were randomized to either once-weekly RT, twice-weekly RT, or twice-weekly balance and tone. We investigated the effect of RT on WMH progression. Results: Results from Chapters 2 and 3 suggest that reduced WMH progression may translate to maintained, or improved, physical and cognitive functions. Chapter 4 demonstrated that physical function is important for cognitive health. Chapter 5 provided proof-of-concept evidence that RT has beneficial effects on WMH progression, which may translate to improved physical and cognitive function. Specifically, we found that reduced WMH progression was significantly iii associated with improved gait speed. Moreover, our results suggest this effect may be dose-dependent, as the significant reduction in WMH progression was only observed among those in the twice-weekly RT group, and not in the once-weekly RT group. Conclusion: We provided converging evidence from four separate studies leading to the conclusion that RT has beneficial effects on WMH progression. Since WMHs are demonstrated to have significant associations with physical and cognitive dysfunctions, we believe that exercise-induced reductions of WMHs progression might translate to improvements in physical and cognitive functions in older adults.
Medicine, Faculty of
Medicine, Department of
Experimental Medicine, Division of
Graduate

The work undertaken involves the use of functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) as separate but complementary non-invasive functional brain imaging modalities. The aim in combining fMRI and MEG is centred around exploitation of the high temporal resolution available in MEG, and the high spatial resolution available in fMRI. However, whilst MEG represents a direct measure of neuronal activity, BOLD fMRI is an indirect measure and this makes the two modalities truly complementary. In both cases, the imaging signals measured are relatively poorly understood and so the fundamental question asked here is: How are the neuromagnetic effects detectable using MEG related to the metabolic effects reflected in the fMRI BOLD response? Initially, a novel technique is introduced for the detection and spatial localisation of neuromagnetic effects in MEG. This technique, based on a beamforming approach to the MEG inverse problem, is shown to yield accurate results both in simulation and using experimental data. The technique introduced is applied to MEG data from a simple experiment involving stimulation of the visual cortex. A number of heterogeneous neuromagnetic effects are shown to be detectable, and furthermore, these effects are shown to be spatially and temporally correlated with the fMRI BOLD response. The limitations to comparing only two measures of brain activity are discussed, and the use of arterial spin labelling (ASL) to make quantitative measurements of physiological parameters supplementing these two initial metrics is introduced. Finally, a novel technique for accurate quantification of arterial cerebral blood volume using ASL is described and shown to produce accurate results. A concluding chapter then speculates on how these aCBV measurements might be combined with those from MEG in order to better understand the fMRI BOLD response.

Thesis (Ph. D. in Medical Engineering and Medical Physics)--Harvard-MIT Program in Health Sciences and Technology, 2013.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 163-182).
Two different hemodynamic-based neuroimaging techniques were studied in this work. Near-Infrared Spectroscopy (NIRS) is a promising technique to measure cerebral hemodynamics in a clinical setting due to its potential for continuous monitoring. However, the presence of strong systemic interference in the signal significantly limits our ability to recover the hemodynamic response without averaging tens of trials. Developing a new methodology to clean the NIRS signal from systemic interference and isolate the cortical signal would therefore significantly increase our ability to recover the hemodynamic response opening the door for clinical NIRS studies such as epilepsy. Toward this goal, a new method based on multi-distance measurements and state-space modeling was developed and further optimized to remove systemic physiological oscillations contaminating the NIRS signal. Furthermore, the cortical and pial contributions to the NIRS signal were quantified using a new multimodal regression analysis. Functional Magnetic Resonance Imaging (fMRI) based on the Blood Oxygenation Level Dependent (BOLD) response has become the method of choice for exploring brain function, and yet the physiological basis of this technique is still poorly understood. Despite the effort, a detailed and validated model relating the signal measured to the physiological changes occurring in the cortical tissue is still lacking. Modeling the BOLD signal is challenging because of the difficulty to take into account the complex morphology of the cortical microvasculature, the distribution of oxygen in those microvessels and its dynamics during neuronal activation. Here, we overcome this difficulty by performing Monte Carlo simulations over real microvascular networks and oxygen distributions measured in vivo on rodents, at rest and during forepaw stimulation, using two-photon microscopy. Our model reveals for the first time the specific contribution of individual vascular compartment to the BOLD signal, for different field strengths and different cortical orientations. Our model makes a new prediction: the amplitude of the BOLD signal produced by a given physiological change during neuronal activation depends on the spatial orientation of the cortical region in the MRI scanner. This occurs because veins are preferentially oriented either perpendicular or parallel to the cortical surface in the gray matter.
by Louis Gagnon.
Ph.D.in Medical Engineering and Medical Physics

The ageing brain undergoes many structural changes. High resolution imaging with MRI has allowed visualisation and quantification of many aspects of this neurological degeneration. One particular feature of the ageing brain is the increased presence of hyperintensities. These can be clearly visualised with T2-weighted imaging. Manual scoring of these lesions is time consuming and prone to inter- and intra-observers variability. We developed automatic methods for quantification and classification of the hyperintensity volumes using T1-weighted and fluid attenuation inversion recovery images of Aberdeen Birth Cohort of 1936. The hyperintensities were classified into different brain regions given by the local Scheltens’ scale, i.e., grey matter, infratentorial, deep white matter and periventricular white matter. The automatically generated hyperintensity volumes were compared with the local scores and investigated in relation to respiratory function and smoking history. Ageing is also associated with cognitive decline. We investigated the role of the structural complexity of ageing brains in the life course changes of cognitive ability. We hypothesised that fractal descriptors of white matter would be associated with childhood IQ, suggesting early cognitive maturation, better fluid cognitive performance and less decline in late life. This was done using fractal measures of white matter structure. Our results show that the automatic quantification and classification methods provide a promising alternative to manual grading with strong correlations (p<.05). The automatic hyperintensity volumes broadly show the same pattern association as shown by the local scores when investigating with the respiratory function, and stronger correlations with daily cigarette consumption and smoking burden than did the local scores. The results of fractal measures of brain complexity demonstrate the potential of fractal measures as an estimate of structural maturation.

Sickle Cell Disease (SCD) is a collection of genetic haemoglobinopathies, the most common and severe being homozygous sickle cell anaemia. In the UK, it has been estimated that 1 in 2000 children are born with SCD. The disease is characterised by chronic anaemia, recurrent pain crises and vascular occlusion. Neurologically, there is a high incidence of stroke in childhood, as well as cognitive dysfunction. Newborn screening programmes and preventative treatments have allowed a much longer lifespan; however recently, neurological research has shifted to characterising subtler aspects of brain development and functioning that may be critically important to the individual’s quality of life. This thesis overviews the neurological and neurocognitive complications of SCD, and how magnetic resonance imaging (MRI) can provide biomarkers for severity of disease. During the PhD, retrospective and prospective cognitive and MRI data were collected and analysed. Diagnostic clinical MRI sequences and advanced MRI sequences were applied, as well as a neuropsychological test battery aimed at intelligence and executive function. First, this thesis reviews the intelligence literature in SCD and includes previously unreported data, finding patients, regardless of abnormality seen on conventional MRI, have lowered full-scale intelligence quotient than controls. Then, to determine imaging biomarkers, volumetric differences and diffusion characteristics were identified. Patients were found to have decreased volumes of subcortical structures compared to controls, in groups corresponding to disease severity. Results from a three-year longitudinal clinical trial suggest evidence of atrophy in paediatric patients, with no apparent protective effect of treatment. Diffusion tensor imaging revealed reduced white matter integrity across the brain, correlating with recognised markers of disease severity (i.e. oxygen saturation and haemoglobin from a full blood count). Overall, the four experiments bridge a gap in the cognitive and neuroimaging literature of the extent of neurological injury in children with SCD.

A statistical meta-analysis combines the results of several studies that address a set of related research hypotheses, thus increasing the power and reliability of the inference. Meta-analytic methods are over 50 years old and play an important role in science; pooling evidence from many trials to provide answers that any one trial would have insufficient samples to address. On the other hand, the number of neuroimaging studies is growing dramatically, with many of these publications containing conflicting results, or being based on only a small number of subjects. Hence there has been increasing interest in using meta-analysis methods to find consistent results for a specific functional task, or for predicting the results of a study that has not been performed directly. Current state of neuroimaging meta-analysis is limited to coordinate-based meta-analysis (CBMA), i.e., using only the coordinates of activation peaks that are reported by a group of studies, in order to "localize" the brain regions that respond to a certain type of stimulus. This class of meta-analysis suffers from a series of problems and hence cannot result in as accurate results as desired. In this research, we describe the problems that existing CBMA methods are suffering from and introduce a hierarchical mixed-effects image-based metaanalysis (IBMA) solution that incorporates the sufficient statistics (i.e., voxel-wise effect size and its associated uncertainty) from each study. In order to improve the statistical-inference stage of our proposed IBMA method, we introduce a nonparametric technique that is capable of adjusting such an inference for spatial nonstationarity. Given that in common practice, neuroimaging studies rarely provide the full image data, in an attempt to improve the existing CBMA techniques we introduce a fully automatic model-based approach that employs Gaussian-process regression (GPR) for estimating the meta-analytic statistic image from its corresponding sparse and noisy observations (i.e., the collected foci). To conclude, we introduce a new way to approach neuroimaging meta-analysis that enables the analysis to result in information such as “functional connectivity” and networks of the brain regions’ interactions, rather than just localizing the functions.

Longitudinal designs are widely used in medical studies as a means of observing within-subject changes over time in groups of subjects, thereby aiming to improve sensitivity for detecting disease effects. Paralleling an increased use of such studies in neuroimaging has been the adoption of pattern recognition algorithms for making individualized predictions of disease. However, at present few pattern recognition methods exist to make full use of neuroimaging data that have been collected longitudinally, with most methods relying instead on cross-sectional style analysis. In this thesis we develop a feature construction method that uses longitudinal high dimensional data to improve the predictive performance of pattern recognition algorithms when classifying early neurodegeneration. Our method can be applied to data from a wide range of longitudinal study designs and permits an arbitrary number of time-points per subject. We apply the method to two problems: discriminating subjects with mild cognitive impairment (MCI) from healthy controls and discriminating subjects at risk for Parkinson’s disease from healthy controls. We show substantial improvements in predictive accuracy relative to cross-sectional classifiers for discriminating disease subjects from healthy controls on the basis of structural magnetic resonance (MR) images. In addition, our method allows for the transfer of longitudinal information from one set of subjects to make disease predictions in another set of subjects. The proposed methodology is simple and, as a feature construction technique, flexible with respect to the choice of classifier, imaging modality and image registration algorithm.

Small vessel disease (SVD) accounts for approximately 25% of all strokes and 45% of all dementias. Although the small vessels cannot be visualised with conventional neuroimaging, the pathological changes in the cerebral white and deep grey matter secondary to SVD has been adopted as markers of SVD. These are best appreciated with magnetic resonance imaging (MRI) and includes recent small subcortical infarcts, white matter hyperintensity (WMH), lacunes, cerebral microbleeds and enlarged perivascular spaces (PVSs). There are however a number of outstanding questions regarding these surrogate neuroimaging markers of SVD and how these markers may influence clinical management. First, although a high burden of microbleeds have been associated with an increased risk of intracerebral haemorrhage (ICH) and possibly recurrent ischaemic stroke in patients with TIA or ischaemic stroke, how microbleeds should influence antithrombotic treatment use after TIA or ischaemic stroke remains uncertain. Second, the long-term prognostic implications of enlarged PVSs in patients with TIA or ischaemic stroke have not been studied. Third, although previous studies have shown possible ethnic differences in prevalence of microbleeds, whether there are any ethnic differences in prevalence of other neuroimaging markers of SVD remains unclear. Fourth, although a Total SVD Score was recently proposed to measure the global SVD burden, the prognostic value of this score in patients with TIA or ischaemic stroke has yet to be studied. Fifth, the relationships of long-term premorbid blood pressure with global SVD burden is unknown. Finally, the age and sex specific associations between renal impairment, carotid and cerebral pulsatility with burden of SVD has yet to be studied. The aim of my thesis was therefore to determine the clinical correlates, ethnic differences and long-term prognostic implications of a range of neuroimaging markers and global burden of SVD. I also aimed to determine the relationships of global SVD burden with long-term mean premorbid blood pressure, renal impairment and carotid pulsatiltiy. I have collected, collated and analysed clinical and neuroimaging data from two independent cohorts - the Oxford Vascular Study (OXVASC) and The University of Hong Kong (HKU). In particular I worked as one of the Clinical Research Fellows at OXVASC and was involved in regular recruitment, assessment and follow up of study patients. In OXVASC, 1080 predominantly Caucasians with TIA or ischaemic stroke who had a cerebral MRI performed at baseline was recruited during 2004 to 2014. I interpreted all these MRIs, specifically coding the burden of microbleeds, enlarged perivascular spaces and lacunes. I was involved in obtaining funding and developing the HKU cohort, which includes 1003 predominantly Chinese with ischaemic stroke recruited during 2008-2014 who had a cerebral MRI performed at baseline. I saw about 25% of the patients in the cohort and was involved in interpreting all of the MRIs of the cohort. All patients from both cohorts were followed-up regularly and adverse events including recurrent ischaemic stroke and ICH was determined. Presence and burden of periventricular and subcortical WMH, lacunes, microbleeds, basal ganglia and centrum semiovale PVSs was determined for all patients and the global burden of SVD estimated according to the Total SVD Score. There are several clinically relevant findings in this thesis. First, I have shown that in Caucasians and Chinese with ≥5 microbleeds, withholding antiplatelet drugs during the first year after TIA or ischaemic stroke may be inappropriate, especially early after TIA. However, the risk of ICH is likely to outweigh any benefit thereafter. Second, I have shown that TIA or ischaemic stroke patients with microbleeds on warfarin had an increased risk of subsequent ICH. However, this risk was not different from that of antiplatelet users with microbleeds. Third, I have shown that a high burden of MRI-visible basal ganglia PVSs is independently associated with an increased risk of recurrent ischaemic stroke, but not ICH. However, the prognostic value of MRI-visible centrum semiovale PVSs in the TIA or ischaemic stroke population is limited. Fourth, I demonstrated significant ethnic differences in underlying prevalence and burden of neuroimaging markers of SVD - Chinese had a greater prevalence of microbleeds, lacunes and subcortical WMH, whilst Caucasians had a greater prevalence of periventricular WMH and PVSs. Fifth, I validated the Total SVD Score and showed that the SVD Score is able to predict risk of recurrent ischaemic stroke and ICH in Caucasians and Chinese, but is unable to identify patients at high risk of ICH from those at high risk of recurrent ischaemic stroke. Sixth, I showed that mean premorbid blood pressure, especially diastolic blood pressure measurements taken 10-20 years prior to TIA or ischaemic stroke was most strongly associated with global SVD burden suggesting a latency effect of hypertension on the pathogenesis of SVD. Finally, I demonstrated age-specific associations between renal impairment, internal carotid artery pulsatility index and SVD burden.

Cerebrovascular diseases (CD) are the third most common cause of death and the leading cause of disability in adults in developed countries (Carmichael, 2012; World Health Organization, 2004). Specifically, ischemic stroke and white matter lesions (WML) often result in multiple neurological, cognitive impairment and behavioral and emotional disorders (Gorelick et al., 2011; Troncoso et al., 2008). Strokes are responsible for damage in the core of the ischemic lesion but may also cause alterations in remote areas from the primary ischemic lesion. The thalamus is a key structure in the cortico-subcortical circuits (Alexander et al., 1986; Byne et al., 2009) and is involved in multiple cognitive functions (Herrero et al., 2002; Sherman, 2005) especially in functions executive, one of the most affected cognitive domains after suffering a stroke. Although it is known that the cortico- subcortical circuits are involved in cognitive functions, to date their neuroimage correlates are unknown. The overall objective of this thesis was to study the effects of a disruption in the cortico-subcortical circuits, due to a direct or remote damage, in executive functions. For the study of remote thalamic abnormalities we use the technique of diffusion tensor image (DTI) for both ischemic stroke and WML. Moreover, due to attention and cognitive inhibition are one of the most important functions of executive domain, we studied the relationship between a specific white matter (WM) tract -called Front aslant Tract (FAT)- and these functions. The study results showed that remote thalamic microstructural abnormalities secondary to a cerebrovascular lesion can occur in both ipsilateral and contralateral thalamus, in healthy subjects with WML and in patients with cerebral ischemic stroke. These thalamic abnormalities may be related to a disruption in the cortico-subcortical circuits associated with executive dysfunction. In addition, the right FAT is involved in attention and response inhibition functions in community-dwelling subjects and participants with ischemic stroke. In conclusion, the results obtained in this thesis suggest that stroke can affect the cortico-subcortical circuits through thalamic microstructural abnormalities and these could be related to cognitive dysfunction. Finally, the novel technique of DTI can play an important role in understanding the cognitive functioning in both ischemic stroke and WML.
Los accidentes cerebrovasculares (ACV) son la tercera causa más común de muerte y la causa principal de discapacidad en adultos en los países desarrollados (Carmichael, 2012; Organización Mundial de la Salud, 2004). Concretamente, el ictus isquémico y las lesiones de sustancia blanca (LSB) frecuentemente dan lugar a múltiples secuelas neurológicas, deterioro cognitivo y alteraciones conductuales y emocionales (Gorelick et al., 2011; Troncoso et al., 2008). Los ACV son responsables de daño en la zona primaria de la lesión isquémica pero también pueden producir alteraciones en áreas remotas a ésta. El tálamo es una estructura clave en los circuitos cortico-subcorticales (Alexander et al., 1986; Byne et al., 2009) y está involucrado en múltiples funciones cognitivas (Herrero et al., 2002; Sherman, 2005) especialmente en las funciones ejecutivas, uno de los dominios cognitivos más afectados después de sufrir un ACV. Aunque se sabe que los circuitos cortico-subcorticales están implicados en las funciones cognitivas, hasta la fecha sus correlatos de neuroimagen se desconocen. El objetivo general de esta tesis ha sido estudiar los efectos de una interrupción en los circuitos cortico-subcorticales debido a una lesión directa o remota en las funciones ejecutivas. Para el estudio de las anomalías talámicas remotas usamos la técnica de la Imagen por Tensor de Difusión (ITD), tanto para el ictus isquémico como para las LSB. Además, dado que la atención y la inhibición cognitiva son una de las funciones más importantes de las funciones ejecutivas, estudiamos la relación entre un tracto de sustancia blanca (SB) -llamado Frontal Aslant Tract (FAT)- y estas funciones. Los resultados de los estudios mostraron que anomalías secundarias microestructurales talámicas remotas a la lesión cerebrovascular pueden ocurrir tanto en el tálamo ipsilateral como en el tálamo contralateral, en sujetos sanos con LSB y en pacientes con un ictus cerebral isquémico. Estas anomalías talámicas pueden estar relacionadas con una disrupción en los circuitos cortico-subcorticales asociado con disfunción ejecutiva. Además, en sujetos de la comunidad y con un ictus isquémico, el FAT derecho está implicado en atención e inhibición de respuesta. En conclusión, los resultados obtenidos en la presente tesis doctoral sugieren que los ACV puede afectar los circuitos cortico-subcortical a través de anomalías microstructurales talámicas y éstas podrían estar relacionadas con la disfunción cognitiva. Finalmente, la novedosa técnica de la ITD puede tener un papel relevante en el conocimiento del funcionamiento cognitivo tanto en el ictus isquémico como en las LSB.

INTRODUCCIÓN: La parálisis cerebral es una de las causas más frecuentes de discapacidad física en niños y supone una condición que persiste a lo largo de la vida (Krageloh-Mann & Cans, 2009). Concretamente, la prevalencia media mundial es de aproximadamente de 2 por cada 1000 nacidos vivos (Stavsky et al., 2017). El término parálisis cerebral incluye un grupo de trastornos permanentes del movimiento y / o postura y de la función motora debido a una afectación en el cerebro en desarrollado o inmaduro que pueden ir acompañados por diferentes comorbilidades (Rosenbaum et al., 2007; Surveillance of Cerebral Palsy in Europe, 2000). La parálisis cerebral es una condición heterogénea tanto en cuanto a etiología como en el tipo y gravedad de la afectación motriz. En este sentido, se considera útil categorizar a los individuos con parálisis cerebral en grupos para proporcionar un mayor nivel de detalle sobre las características y necesidades específicas de las diferentes personas (Surveillance of Cerebral Palsy in Europe, 2007). Los esquemas de clasificación tradicionales se han centrado principalmente en agrupar los casos según la distribución de las extremidades afectadas y el tipo predominante de tono o anormalidad del movimiento, lo que da lugar a tres grandes grupos de parálisis cerebral: el espástico, el discinetico y el atáxico (Platt, Krageloh‐Mann, & Cans, 2009). La presente tesis doctoral se centra en el estudio del segundo tipo más frecuente, la parálisis cerebral discinètica, que se caracteriza por movimientos involuntarios, descontrolados, recurrentes y, ocasionalmente, estereotipados. La resonancia magnética ha aumentado nuestra comprensión de la parálisis cerebral (Vandborg et al., 2015), pero mientras que la mayoría de los estudios se centran en la parálisis cerebral espástica, los estudios de neuroimagen centrados en la parálisis cerebral discinética son aun escasos. Tanto los procesos cognitivos globales como específicos pueden verse afectados en la parálisis cerebral y casi el 50% de la población de parálisis cerebral presenta un cociente intelectual por debajo de 70 (Novak, Hines, Goldsmith, & Barclay, 2012). La consideración de las funciones cognitivas es uno de los aspectos centrales en el estudio de la parálisis cerebral (Schiariti, Selb, Cieza, & O’Donnell, 2015) siendo la función ejecutiva de particular interés debido a su papel clave en la adquisición efectiva de nuevas habilidades, conocimiento y la aplicación de estas conocimiento en el día a día (Anderson & Ylvisaker, 2009). La asociacione entre la calidad de vida y la función ejecutiva se ha descrito en la población general y en otras condiciones neurológicas y psiquiátrica distintas a la parálisis cerebral (Barf, Post, Verhoef, Gooskens, & Prevo, 2010; Davis, 2010; Pattanayak, Sagar, & Mehta, 2012; Ritsner, 2007; Sherman, Slick, & Eyrl, 2006). Existe acuerdo en que la parálisis cerebral discinética presenta mayor gravedad motriz que otros tipos parálisis cerebral (Himmelmann et al., 2009). Sin embargo, existen pocos estudios que comparen las capacidades cognitivas entre este tipo de parálisis cerebral y otros subtipos y, de hecho, la mayoría de ellos no controla los resultados por el nivel de gravedad motriz. Ningún estudio hasta la fecha se ha enfocado específicamente en estudiar la función ejecutiva ni su asociación con las características de la estructura cerebral en este subtipo de parálisis cerebral. Las lesiones de los ganglios basales y el tálamo son especialmente frecuentes en personas con parálisis cerebral discinética (Bax, Tydeman, & Flodmark, 2006; Himmelmann, Ahlin, Jacobsson, Cans, & Thorsen, 2011) y dado que los circuitos fronto-estriatales juegan un papel crucial en el funcionamiento ejecutivo para las personas con un desarrollo normal (Bottcher, 2010; Krause et al., 2012), seria esperable encontrar este correlato anátomo-funcional en la parálisis cerebral discinética. Sin embargo, muy pocos trabajos han estudiado la relación entre la estructura cerebral y el funcionamiento ejecutivo en esta población. De hecho, no existe ningún estudio que analice esta relación en las personas con parálisis cerebral discinética (Weierink, Vermeulen, & Boyd, 2013). OBJETIVOS: El objetivo general de la tesis fue caracterizar el funcionamiento ejecutivo y el funcionamiento intelectual general y sus bases biológicas en imagen de resonancia magnética estructural convencional y de difusion en la parálisis cerebral discinética. Específicamente, en la presente tesis se formularon cuatro objetivos que han sido abordados mediante la realización de cuatro estudios. Primero, analizar el impacto de la función ejecutiva en la calidad de vida en personas con parálisis cerebral discinética (Estudio 1). El segundo objetivo fue identificar el perfil de funcionamiento ejecutivo e intelectual en las personas con parálisis cerebral discinética (Estudio 2). Tercero, identificar las alteraciones de la estructura cerebral en participantes con parálisis cerebral discinética (Estudio 3 y Estudio 4). Cuarto, investigar el correlato neural del funcionamiento ejecutivo e intelectual general en participantes con parálisis cerebral discinética (Estudio 3 y Estudio 4). RESULTADOS: En consecuencia, los principales hallazgos de los estudios son los siguientes. 1) Uno de los componentes de las funciones ejecutivas, la flexibilidad cognitiva, es un predictor importante de la calidad de vida en personas con parálisis cerebral discinética. 2) Las personas con parálisis cerebral discinética presentan dificultades tanto en el rendimiento intelectual general como en el funcionamiento ejecutivo. Sin embargo, las habilidades de planificación parecen ser similares a las de los controles con desarrollo normativo. Además, las personas con parálisis cerebral discinética muestran un mejor funcionamiento intelectual y ejecutivo que las personas con parálisis cerebral espástica, lo que indica una tendencia general hacia un mejor rendimiento cognitivo en lugar de un déficit disejecutivo específico. 3) Existen lesiones observables en la substancia blanca y gris, así como una reducción de la integridad de la substancia blanca en la parálisis cerebral discinética. Concretamente, las lesiones en el tálamo lateral posterior y en el lóbulo frontal son los más comunes en nuestra muestra de personas con parálisis cerebral discinética. Además, la pérdida en la integridad de la sustancia blanca en la parálisis cerebral discinética predomina en regiones posteriores subyacentes, principalmente en la corteza parietal. 4) El funcionamiento intelectual general está relacionado con la integridad de la substancia blanca en varias regiones cortico-corticales y cortico-subcorticales y con lesiones cerebrales observables principalmente del tálamo posterior. Las funciones ejecutivas se relacionaron con la microestructura de la sustancia blanca en regiones que contienen las vías fronto-corticales y cortico-subcorticales posteriores, así como con las lesiones cerebrales observables especialmente en el lóbulo parietal y de cuerpo calloso medio y posterior. Contrariamente a nuestra hipótesis, no se identificaron relaciones significativas entre la función ejecutiva y las vías fronto-estriatales. REFRENCIAS: Anderson, V., & Ylvisaker, M. (2009). Executive function and the frontal lobes: Themes for child development, brain insult and rehabilitation. Developmental Neurorehabilitation, 12(August), 253–254. https://doi.org/10.3109/17518420903086899 Barf, H. A., Post, M. W. M., Verhoef, M., Gooskens, R. H. J. M., & Prevo, A. J. H. (2010). Is cognitive functioning associated with subjective quality of life in young adults with spina bifida and hydrocephalus? Journal of Rehabilitation Medicine, 42(1), 56–9. https://doi.org/10.2340/16501977-0481 Bax, M., Tydeman, C., & Flodmark, O. (2006). Clinical and MRI correlates of cerebral palsy: the European Cerebral Palsy Study. Jama, 296(13), 1602–8. https://doi.org/10.1001/jama.296.13.1602 Bottcher, L. (2010). Children with spastic cerebral palsy, their cognitive functioning, and social participation: a review. Child Neuropsychology : A Journal on Normal and Abnormal Development in Childhood and Adolescence, 16(3), 209–228. https://doi.org/10.1080/09297040903559630 Davis, J. (2010). The independent contribution of executive functions to health related quality of life in older women. BMC Geriatrics, 10(1), 16. Retrieved from http://www.biomedcentral.com/1471-2318/10/16 Himmelmann, K., Ahlin, K., Jacobsson, B., Cans, C., & Thorsen, P. (2011). Risk factors for cerebral palsy in children born at term. Acta Obstetricia et Gynecologica Scandinavica, 90(10), 1070–1081. https://doi.org/10.1111/j.1600-0412.2011.01217.x Himmelmann, K., McManus, V., Hagberg, G., Uvebrant, P., Krageloh-Mann, I., & Cans, C. (2009). Dyskinetic cerebral palsy in Europe: trends in prevalence and severity. Archives of Disease in Childhood, 94(12), 921–926. https://doi.org/10.1136/adc.2008.144014 Krageloh-Mann, I., & Cans, C. (2009). Cerebral palsy update. Brain & Development, 31(7), 537–544. https://doi.org/10.1016/j.braindev.2009.03.009 Krause, M., Mahant, N., Kotschet, K., Fung, V. S., Vagg, D., Wong, C. H., & Morris, J. G. L. (2012). Dysexecutive behaviour following deep brain lesions - Adifferent type of disconnection syndrome? Cortex, 48(1), 97–119. https://doi.org/10.1016/j.cortex.2011.03.014 Novak, I., Hines, M., Goldsmith, S., & Barclay, R. (2012). Clinical prognostic messages from a systematic review on cerebral palsy. Pediatrics, 130(5), e1285-312. https://doi.org/10.1542/peds.2012-0924 Pattanayak, R. D., Sagar, R., & Mehta, M. (2012). Neuropsychological performance in euthymic Indian patients with bipolar disorder type I: correlation between quality of life and global functioning. Psychiatry and Clinical Neurosciences, 66(7), 553–563. https://doi.org/10.1111/j.1440-1819.2012.02400.x Platt, M. J., Krageloh‐Mann, I., & Cans, C. (2009). Surveillance of cerebral palsy in Europe: reference and training manual. Medical Education, 43, 495–496. Ritsner, M. S. (2007). Predicting quality of life impairment in chronic schizophrenia from cognitive variables. Quality of Life Research, 16(6), 929–937. https://doi.org/10.1007/s11136-007-9195-3 Rosenbaum, P., Paneth, N., Leviton, A., Goldstein, M., Bax, M., Damiano, D., … Jacobsson, B. (2007). A report: the definition and classification of cerebral palsy April 2006. Developmental Medicine and Child Neurology. Supplement, 109, 8–14. Schiariti, V., Selb, M., Cieza, A., & O’Donnell, M. (2015). International Classification of Functioning, Disability and Health Core Sets for children and youth with cerebral palsy: A consensus meeting. Developmental Medicine and Child Neurology, 57(2), 149–158. https://doi.org/10.1111/dmcn.12551 Sherman, E., Slick, D., & Eyrl, K. (2006). Executive dysfunction is a significant predictor of poor quality of life in children with epilepsy. Epilepsia, 47(11), 1936–1942. https://doi.org/10.1111/j.1528-1167.2006.00816.x Stavsky, M., Mor, O., Mastrolia, S. A., Greenbaum, S., Than, N. G., & Erez, O. (2017). Cerebral Palsy—Trends in Epidemiology and Recent Development in Prenatal Mechanisms of Disease, Treatment, and Prevention. Frontiers in Pediatrics, 5(February), 1–10. https://doi.org/10.3389/fped.2017.00021 Surveillance of Cerebral Palsy in Europe. (2000). Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Developmental Medicine & Child Neurology, 42(12), 816–824. https://doi.org/10.1111/j.1469-8749.2000.tb00695.x Surveillance of Cerebral Palsy in Europe. (2007). The definition and classification of cerebral palsy. Developmental Medicine & Child Neurology, 49(109), 1–44. https://doi.org/http://dx.doi.org/10.1111/j.1469-8749.2007.00001.x Vandborg, P. K., Hansen, B. M., Greisen, G., Mathiasen, R., Kasper, F., & Ebbesen, F. (2015). Follow-up of extreme neonatal hyperbilirubinaemia in 5- to 10-year-old children: A Danish population-based study. Developmental Medicine and Child Neurology, 57(4), 378–384. https://doi.org/10.1111/dmcn.12603 Weierink, L., Vermeulen, R. J., & Boyd, R. N. (2013). Brain structure and executive functions in children with cerebral palsy: A systematic review. Research in Developmental Disabilities, 34(5), 1678–1688. https://doi.org/10.1016/j.ridd.2013.01.035
Magnetic resonance imaging has increased our understanding of cerebral palsy [1] but most studies have focused on spastic cerebral palsy, whilst neuroimaging studies of dyskinetic cerebral palsy remain scarce.Global and specific cognitive processes may be affected in cerebral palsy, with almost 50% of the cerebral palsy population having an intellectual quotient below 70 [2]. Cognitive functions are considered one of the essential “Core Sets” in cerebral palsy [3], executive function being of particular interest because of its key role in the effective acquisition of new skills, knowledge, and the application of this knowledge in life [4]. Associations between quality of life and executive function have been described in the general population, in psychiatric conditions, and in neurological conditions other than cerebral palsy [5–9].It is agreed that dyskinetic cerebral palsy is associated with poorer motor outcomes than other cerebral palsy types [10] but very few studies compare cognitive abilities in dyskinetic cerebral palsy with those of other cerebral palsy subtypes, particularly in groups with similar levels of motor ability. No study to date has specifically focused on executive function nor its association with brain magnetic resonance imaging characteristics in this cerebral palsy subtype. Basal ganglia and thalamus lesions are frequently described in people with dyskinetic cerebral palsy [11, 12] and fronto-striatal circuitry play a crucial role on executive functioning for typically developing people [13, 14]. However, there is a paucity of brain imaging studies focussing on executive functioning, with no studies including participants with dyskinetic cerebral palsy [15].The overall aim of the thesis was to characterize executive functioning and general intellectual functioning and its biological bases in dyskinetic cerebral palsy, as measured by diffusion and structural magnetic resonance imaging. Specifically, the current thesis formulated the following four aims, which were accomplished through four studies. First, to analyse the impact of executive function on quality of life in people with dyskinetic cerebral palsy (Study 1). Second, to map executive and intellectual functioning in people with dyskinetic cerebral palsy (Study 2). Third, to identify brain structure alterations in participants with dyskinetic cerebral palsy (Study 3 and Study 4). Fourth, to investigate the neural correlate of executive and general intellectual functioning in participants with dyskinetic cerebral palsy (Study 3 and Study 4).The main findings of the studies are the following. (1) An executive function domain, cognitive flexibility, is an important driver of quality of life in people with dyskinetic cerebral palsy. (2) People with dyskinetic cerebral palsy present difficulties in both general intellectual and executive functioning but goal setting abilities are close to those in typically developing controls. Additionally, people with dyskinetic cerebral palsy display better intellectual and executive functioning than people with spastic cerebral palsy, indicating a general tendency towards a better cognitive level rather than a specific dysexecutive deficit.(3) Observable white and grey matter lesions as well as white matter integrity are involved in dyskinetic cerebral palsy. Specifically, posterior lateral thalamus and the frontal lobe lesions are the most common in our sample of people with dyskinetic cerebral palsy. In this sample, the loss in the integrity of the white matter predominantly appears outside of the frontal lobe, mainly in the parietal cortex. (4) General intellectual functioning is related to white matter integrity in several cortico-cortical and cortico-subcortical regions and with observable brain lesions particularly in the posterior thalamus. Executive functions were related with white matter microstructure in regions containing fronto-cortical and posterior cortico-subcortical pathways and with observable brain lesions particularly in the parietal lobe and the middle and posterior corpus callosum. Unexpectedly, neither in terms of white matter microstructure nor in terms of observable lesions, was there a significant relationship between executive function and the fronto-striatal pathways.References1. Vandborg, P. K., Hansen, B. M., Greisen, G., Mathiasen, R., Kasper, F., & Ebbesen,F. (2015). Follow-up of extreme neonatal hyperbilirubinaemia in 5- to 10-year-old children: A Danish population-based study. Developmental Medicine and Child Neurology, 57(4), 378–384.2. Novak, I., Hines, M., Goldsmith, S., & Barclay, R. (2012). Clinical prognostic messages from a systematic review on cerebral palsy. Pediatrics. DOI:10.1542/peds.2012-09243. Schiariti, V., Selb, M., Cieza, A., & O’Donnell, M. (2015). International Classification of Functioning, Disability and Health Core Sets for children and youth with cerebral palsy: A consensus meeting. Developmental Medicine and Child Neurology. DOI:10.1111/dmcn.125514. Anderson, V., & Ylvisaker, M. (2009). Executive function and the frontal lobes: Themes for child development, brain insult and rehabilitation. Developmental neurorehabilitation. DOI:10.3109/175184209030868995. Barf, H. A., Post, M. W. M., Verhoef, M., Gooskens, R. H. J. M., & Prevo, A. J. H. (2010). Is cognitive functioning associated with subjective quality of life in young adults with spina bifida and hydrocephalus? Journal of rehabilitation medicine. DOI:10.2340/16501977-04816. Davis, J. (2010). The independent contribution of executive functions to health related quality of life in older women. BMC Geriatrics. DOI: 10.1186/1471-2318-10- 16.7. Pattanayak, R. D., Sagar, R., & Mehta, M. (2012). Neuropsychological performance in euthymic Indian patients with bipolar disorder type I: Correlation between quality of life and global functioning. Psychiatry and clinical neurosciences, 66(7), 553–563. 8. Ritsner, M. S. (2007). Predicting quality of life impairment in chronic schizophrenia from cognitive variables. Quality of Life Research, 16(6), 929–937.9. Sherman, E., Slick, D., & Eyrl, K. (2006). Executive dysfunction is a significant predictor of poor quality of life in children with epilepsy. Epilepsia. DOI:10.1111/j.1528-1167.2006.00816.x10. Himmelmann, K., McManus, V., Hagberg, G., Uvebrant, P., Krageloh-Mann, I., & Cans, C. (2009). Dyskinetic cerebral palsy in Europe: trends in prevalence and severity. Archives of Disease in Childhood, 94(12), 921–926.11. Bax, M., Tydeman, C., & Flodmark, O. (2006). Clinical and MRI correlates of cerebralpalsy: the European Cerebral Palsy Study. JAMA, 296(13), 1602–1608.12. Himmelmann, K., Ahlin, K., Jacobsson, B., Cans, C., & Thorsen, P. (2011). Risk factors for cerebral palsy in children born at term. Acta Obstetricia et Gynecologica Scandinavica, 90(10), 1070–1081.13. Krause, M., Mahant, N., Kotschet, K., Fung, V. S., Vagg, D., Wong, C. H., & Morris,J. G. L. (2012). Dysexecutive behaviour following deep brain lesions – A different type of disconnection syndrome? Cortex, 48(1), 97–119.14. Bottcher, L. (2010). Children with spastic cerebral palsy, their cognitive functioning, and social participation: a review. Child neuropsychology: A journal on normal and abnormal development in childhood and adolescence, 16(3), 209–228.15. Weierink, L., Vermeulen, R. J., & Boyd, R. N. (2013). Brain structure and executive functions in children with cerebral palsy: A systematic review. Research in Developmental Disabilities, 34(5), 1678–1688.

The importance of investigating Emotion Regulation (ER) may be self-evident, given that emotions have a substantial impact on our daily lives. ER encompasses set of processes that people go through in order to cultivate their feelings that arise at the moment and produce some response. Brain-imaging studies of ER have broadly focused on examining cognitive strategies, such as reappraisal, in order to understand underlying variables that contribute to the development of this particular process of emotions. The main focus in this paper was to summarize some of the observation done by functional Magnetic Resonance Imaging (fMRI) on neural processes underlying cognitive reappraisal. Furthermore, the paper will discuss some of these experiments that have been made through the last 15 years in the field where indications have been somewhat confusing when it comes to certain aspects of presented data, especially in comparison with other studies. Finally, a brief overview and some of the significant contributions, such as a process model of ER, to the field of ER have been presented and discussed. Cognitive reappraisal has been shown to effectively down-regulate subjective emotional experience. Even though many studies have been performed in measuring brain-activity when engaging in cognitive reappraisal, a unified and accepted agreement has yet not been found. In broader terms, brain-responses when engaging in cognitive reappraisal seem to operate in a particular manner where different parts of prefrontal and parietal cortex execute control over subcortical regions, such as amygdala.

Animal studies have shown that the striatal cholinergic system plays a role in reversal learning, but there has been no attempt to study this in humans due to a lack of appropriate non-invasive techniques. This body of work aimed to address the gap in the literature concerning the role of the human striatal cholinergic system and its thalamic inputs in cognitive flexibility in vivo. To do this, we used a combination of proton magnetic resonance spectroscopy ('H-MRS) and high-resolution functional magnetic resonance imaging (fMRI), together with a reversal learning task and computational modelling to investigate the relationship between cholinergic function and cognitive flexibility. First, we measured individual levels of choline (CHO) at rest in both the dorsal and ventral striatum using 'H-MRS, and examined their relationship with performance on a reversal learning task. We found that average levels of CHO in the human dorsal striatum (DS) are associated with performance during reversal, but not during initial learning. Specifically, lower levels of CHO in the dorsal striatum were associated with faster perseveration. Moreover, choline levels explained variance in the number of perseverative trials over and above that explained by learning rates from negative prediction errors. Second, we used functional-'H-MRS (fMRS) to measure changes in CHO levels in the DS during performance on the reversal learning task. We found a task-dependent decrease in CHO levels in the human DS specifically during reversal learning. We interpret this to reflect an increase in ACh levels, which is in line with findings from the animal literature. Lastly, we used fMRI along with a combination of parametric modulation and psychophysiological interaction (PPI) analysis to investigate the role of the DS and thalamostriatal interactions during human reversal learning. Parametric modulation revealed a performance-dependent dynamic increase in activation in the dorsal striatum. Additionally, PPI analysis revealed task-dependent changes in connectivity between the DS and the centromedial parafascicular complex, which is the main input to the cholinergic system . Moreover, the strength of connectivity correlated with the ability to flexibly alter behaviour. Taken together, these three experiments provide multimodal in vivo evidence to demonstrate a role for the human cholinergic system in the DS in behavioural flexibility as measured using a reversal learning paradigm. Additionally, this body of work demonstrates that it is possible to use in-vivo 'H-MRS, both at rest and as a functional measure to investigate the human striatal cholinergic system. This method not only helps to bridge the gap between animal and human studies, but importantly may provide a novel method of studying disorders that comprise cholinergic dysfunction in humans in vivo (e.g. Parkinson's Disease, Alzheimer's Disease), extending our understanding of the neural mechanisms underlying these disorders.

The aim of this doctoral dissertation was to examine effects of olfactory imagery on other sensory and perceptual processes, and to explore brain areas involved in generation of olfactory mental images. Four studies, three behavioral and one functional neuroimaging (Positron Emission Tomography, or PET), were conducted, and healthy volunteers participated in all four studies. In Study 1, participants were better at detecting weak odors when they simultaneously imagined the same compared with a different odor as the one being detected. This effect of olfactory imagery was specific, as the request to imagine objects visually did not have any effect on detection of weak odors. In Studies 2 and 3, effects of presented and imagined odors on taste perception were compared. Effects of imagined odors were equivalent to the effects of presented odors when an objective measure of taste perception (detection of a weak tastant, Study 3) was used, and comparable but more limited when a subjective measure of taste perception (intensity ratings, Study 2) was used. In Study 4, PET technology was used to investigate changes in cerebral blood flow (CBF) associated with odor imagery. Participants were screened and selected for their odor imagery ability, using the behavioral paradigm developed in Study 1. Increased CBF associated with odor imagery was revealed in several areas relevant for olfaction: the left primary olfactory cortical region including piriform cortex, the left secondary olfactory cortical region (posterior orbitofrontal cortex), and the rostral insula bilaterally. Interestingly, increased activity in the primary olfactory cortex and the rostral insula was observed both in the odor imagery and the odor perception subtraction. Based on the obtained findings, I concluded that the effects of imagined odors on sensory processes are specific when compared with visual imagery, and similar to the effects of presented odors. Furthermore, the neural

The primary objective of this thesis was to investigate body image at both the behavioural and neural levels. We describe three studies aimed at: (1) developing a novel digital methodology with which to assess perceptual aspects of body image during adolescence; (2) investigating perceptual accuracy and sensitivity to changes in the size/shape of body images among healthy adolescents; and (3) identifying the neural mechanisms of body perception using functional magnetic resonance imaging (fMRI). A novel library of digital images of adolescent bodies was created and used to characterize natural covariations in body size and shape using principal components analysis. Identified principal components were used to morph body images in a realistic manner to generate larger or smaller bodies. These morphed body-image stimuli were then used in a behavioural investigation of self body-image perception among adolescents. Male and female adolescents overestimated the size of their bodies. When compared with males, females overestimated their body size to a greater extent and showed greater sensitivity in detecting changes in body size. Overestimation of body size and detection sensitivity increased with subject age. Detection sensitivity decreased as a function of subjects' body mass index (BMI). In order to identify the underlying neural mechanisms of these effects, functional block-design and fMR-adaptation experiments were completed in healthy young adults. During both experiments, females, and not males, showed greater fMR signal in the right versus left hemisphere in the extrastriate body area (EBA) and fusiform body area (FBA). During the block-design experiment, females also demonstrated greater right EBA response compared with males. Observer BMI modulated the EBA hemispheric effect in both experiments. A significant recovery from adaptation was found in EBA and FBA with body-image morphing, indicating that both regions were sensitive to body-size changes. Ultimately, we demonstrated the successful use of a novel body-morphing method for the assessment of body image, established that sex, age, and BMI modulate accuracy of self body-size estimation and detection of changes in body size, and described evidence of EBA and FBA as the likely neural substrates of these behavioural effects.

This thesis investigates the neuroimaging and neuropsychological characteristics of Posterior Cortical Atrophy (PCA) which is most often caused by Alzheimer’s disease (AD) pathology. Posterior cortical atrophy describes a predominantly posterior pattern of atrophy and cognitive deficits. PCA has been poorly characterized and is likely to have been under-recognized. Using magnetic resonance imaging (MRI), patterns of cortical thickness were assessed in patients with pathologically-confirmed AD and different clinical presentations during life (including amnestic, visual and behavioural phenotypes). In addition to atrophy in the medial temporal lobe, tissue loss in posterior regions is indicative of AD pathology. Since medial temporal lobe atrophy is not specific to AD, posterior atrophy may aid distinction between AD and other dementias. Using easily-applied visual rating scales for medial temporal and posterior atrophy in patients with pathologically-confirmed AD and frontotemporal lobar degeneration (FTLD), it was shown that posterior atrophy ratings improve classification accuracy of AD from FTLD and controls. Cross-sectional and longitudinal image analysis techniques were used to characterize atrophy patterns in PCA compared with controls and typical amnestic AD. Whilst the cross-sectional analysis revealed differential patterns of tissue loss in these two groups, with PCA showing greatest atrophy in posterior parietal regions, and typical AD predominantly in medial temporal lobe regions, longitudinal results showed that at five years disease duration, both PCA and typical AD had global grey matter loss and cortical thinning compared with controls. The nature of visual deficits in PCA was assessed by administering detailed neuropsychological tests. The behavioural data showed that visual deficits were not uniformly affected in PCA, with considerable heterogeneity of visual impairments shown. Cortical thickness measures were used to assess atrophy patterns in PCA patients with predominant space versus object perception impairments, revealing overlap in cortical thinning patterns between these two PCA subgroups. In summary this thesis investigates the common and differential atrophy patterns of atypical AD presentations as well as the degree of heterogeneity of deficits which exist within the PCA presentation.

Neuroimaging studies have shown that natural language processes engage left hemisphere perisylvian brain regions. Artificial Grammars (AG), which are designed to emulate aspects of language syntactic structure, recruit comparable brain areas. Nonhuman animals have been shown to learn a range of different AGs. However, no data is currently available regarding the brain areas that support these processes. In this thesis, I combined behavioural artificial grammar learning (AGL) and fMRI experiments to generate insights regarding language evolution, and as a first step to developing animal model systems for aspects of language processing. These experiments provide novel evidence that nonhuman primates are able to learn a non-deterministic AG, designed to emulate some of the variability of the structure of sentences in natural language, and demonstrated notable correspondences between the brain regions involved in macaque and human AGL. I developed a quantitative method to compare AGL abilities across species and studies, and a novel eye-tracking technique with which to collect objective behavioural data. Using this technique, and a refined version of a traditional video-coding paradigm, I demonstrated that Rhesus macaques notice violations of the AG structure and that these results could not be explained by reliance on simple cues. Common marmosets also showed evidence of AGL however, these results may have been driven by simple learning strategies. Comparative fMRI experiments showed that, in humans, violations of the AG activated a number of perisylvian brain regions associated with language processing, including the ventral frontal cortex (vFC), temporal and temporo-parietal regions, although not Broca’s area (BA44/45). In Rhesus macaques, comparable patterns of activation were seen in the ventral frontal cortex and temporo-parietal regions. Additional activation in BA44/45 in macaques provides interesting insights into the evolution of this region. These experiments provide novel evidence regarding the AGL capabilities of nonhuman primates, and the brain areas that support them, suggesting that some language related functions may represent generic, rather than language specific processes. Therefore, some of the brain regions involved in AGL in both species might share a common evolutionary heritage, and therefore Rhesus macaques might represent a valuable animal model system for aspects of language processing.

PET is a non-invasive molecular imaging technique that is increasingly being used for medical imaging and drug development. Carbon-11 (11C; half-life 20.4 min) is one of the most commonly used radionuclides for PET molecular imaging. 11C is usually produced in the form of [11C]CO2 and converted into more reactive secondary precursors such as [11C]methyl iodide and [11C]carbon monoxide for radiolabelling. Although such secondary precursors are undoubtedly useful, given the short half-life of 11C, it would be advantageous to use [11C]CO2 directly from the cyclotron without additional time-consuming processing. Therefore, the development of radiochemical methods to efficiently radiolabel compounds directly with [11C]CO2 for applications in PET neuroimaging is an important goal and is the focus of this thesis. This work includes the development of novel radiolabelling methodology utilising [11C]CO2 for the radiolabelling of molecules based on urea and carbamate scaffolds. These functional groups are found in a plethora of biologically active molecules and pharmaceuticals. As proof of concept, the utility of the developed radiochemistry methods were applied to the synthesis of novel GABA and glutamate radiotracers. GABA and glutamate are major excitatory and inhibitory neurotransmitters in the brain. Although implicated in many diseases, the in vivo function of these neurotransmitter system is poorly understood. Their dysfunction are implicated in pathologies such as addiction, Alzheimer’s disease, Parkinson’s disease and autism. Monitoring the expression of the receptors in vivo and in vitro would enable better understanding of these diseases, their progression and treatment. The research described in this thesis unveils new methods to radiolabel novel molecules for these targets with 11C thereby enabling more opportunities to study them in vitro using autoradiography and in vivo using PET molecular imaging.

Parkinson's disease (PD) is a common, disabling, neurodegenerative disease characterised by three core motor symptoms: tremor, rigidity and bradykinesia. These symptoms arise from degeneration of dopaminergic (DA) cells in the substantia nigra (SN) and the subsequent loss of dopaminergic terminals within the striatum, and circuits to cortical areas, critical in the control of movement. Other, non-DA systems are now known to be involved in the pathogenesis of PD, defective cognitive functions and side effects of DA medication treatments. Thus, the use of non-invasive in vivo techniques such as magnetic resonance imaging (MRI) has allowed a reliable, albeit in-direct method of assessing alterations in the PD brain. It is now widely considered that motor control is dependent upon the integrated operation of large-scale distributed brain networks. Recent methodological advances in MRI techniques allow both structural and functional connectivity between critical regions of motor control to be investigated and increase our understanding of the impact of PD pathology on motor networks and its subsequent effect on symptomatology. In this thesis, I present three studies that combine both structural and functional MRI techniques to assess the neural PD motor network and to test the general hypothesis that loss of effective motor control in PD arises from disrupted connectivity. I demonstrate in a sizable cross-sectional study that as disease burden increases, effective functioning in key motor areas and functional connectivity between regions in both the active and resting state is initially compromised but does show evidence of compensatory mechanisms. In addition, I show that compensatory mechanisms are likely to possess a neural reserve property rather than permit a period of normal functioning. Next, I present a follow-up study that assessed the active and resting neural motor network longitudinally. This study clearly shows that functional connectivity of the active and resting neural motor network is compromised as the disease progresses with evidence suggesting the initiation of compensatory mechanisms. Finally, structural properties of key regions related to PD pathology (substantia nigra and striatum) have been assessed to elucidate the effect of PD progression on diffusion indices and clinical symptoms. This work identifies the importance of multi-modal assessment of neural networks in PD to evaluate the effect of disease on neural motor control.