Recent research has suggested disrupted interactions between brain regions may contribute to some of the symptoms of motor disorders such as Parkinson’s Disease (PD). It is therefore important to develop models for inferring brain functional connectivity from data obtained through non-invasive imaging technologies, such as functional magnetic resonance imaging (fMRI). The complexity of brain activities as well as the dynamic nature of motor disorders require such models to be able to perform complex, large-scale, and dynamic system computation. Traditional models proposed in the literature such as structural equation modeling (SEM), multivariate autoregressive models (MAR), dynamic causal modeling (DCM), and dynamic Bayesian networks (DBNs) have all been suggested as suitable for fMRI data analysis. However, they suffer from their own disadvantages such as high computational cost (e.g. DBNs), inability to deal with non-linear case (e.g. MAR), large sample size requirement (e.g. SEM), et., al. In this research, we propose applying Probabilistic Boolean Network (PBN) for modeling brain connectivity due to its solid stochastic properties, computational simplicity, robustness to uncertainty, and capability to deal with small-size data, typical for fIVIRI data sets. Applying the proposed PBN framework to real fMRI data recorded from PD subjects enables us to identify statistically significant abnormality in PD connectivity by comparing it with normal subjects. The PBN results also suggest a mechanism of evaluating the effectiveness of L-dopa, the principal treatment for PD. In addition to PBNs’ promising application in inferring brain connectivity, PBN modeling for brain ROTs also enables researchers to study dynamic activities of the system under stochastic conditions, gaining essential information regarding asymptotic behaviors of ROTs for potential therapeutic intervention in PD. The results indicate significant difference in feature states between PD patients and normal subjects. Hypothesizing the observed feature states for normal subject as the desired functional states, we further explore possible methods to manipulate the dynamic network behavior of PD patients in the favor of the desired states from the view of random perturbation as well as intervention. Results identified a target ROT with the best intervention performance, and that ROl is a potential candidate for therapeutic exercise.

Recent research has suggested disrupted interactions between brain regions may contribute to some of the symptoms of motor disorders such as Parkinson’s Disease (PD). It is therefore important to develop models for inferring brain functional connectivity from data obtained through non-invasive imaging technologies, such as functional magnetic resonance imaging (fMRI). The complexity of brain activities as well as the dynamic nature of motor disorders require such models to be able to perform complex, large-scale, and dynamic system computation. Traditional models proposed in the literature such as structural equation modeling (SEM), multivariate autoregressive models (MAR), dynamic causal modeling (DCM), and dynamic Bayesian networks (DBNs) have all been suggested as suitable for fMRI data analysis. However, they suffer from their own disadvantages such as high computational cost (e.g. DBNs), inability to deal with non-linear case (e.g. MAR), large sample size requirement (e.g. SEM), et., al. In this research, we propose applying Probabilistic Boolean Network (PBN) for modeling brain connectivity due to its solid stochastic properties, computational simplicity, robustness to uncertainty, and capability to deal with small-size data, typical for fIVIRI data sets. Applying the proposed PBN framework to real fMRI data recorded from PD subjects enables us to identify statistically significant abnormality in PD connectivity by comparing it with normal subjects. The PBN results also suggest a mechanism of evaluating the effectiveness of L-dopa, the principal treatment for PD. In addition to PBNs’ promising application in inferring brain connectivity, PBN modeling for brain ROTs also enables researchers to study dynamic activities of the system under stochastic conditions, gaining essential information regarding asymptotic behaviors of ROTs for potential therapeutic intervention in PD. The results indicate significant difference in feature states between PD patients and normal subjects. Hypothesizing the observed feature states for normal subject as the desired functional states, we further explore possible methods to manipulate the dynamic network behavior of PD patients in the favor of the desired states from the view of random perturbation as well as intervention. Results identified a target ROT with the best intervention performance, and that ROl is a potential candidate for therapeutic exercise.

Recent research has suggested disrupted interactions between brain regions may contribute to some of the symptoms of motor disorders such as Parkinson’s Disease (PD). It is therefore important to develop models for inferring brain functional connectivity from data obtained through non-invasive imaging technologies, such as functional magnetic resonance imaging (fMRI). The complexity of brain activities as well as the dynamic nature of motor disorders require such models to be able to perform complex, large-scale, and dynamic system computation. Traditional models proposed in the literature such as structural equation modeling (SEM), multivariate autoregressive models (MAR), dynamic causal modeling (DCM), and dynamic Bayesian networks (DBNs) have all been suggested as suitable for fMRI data analysis. However, they suffer from their own disadvantages such as high computational cost (e.g. DBNs), inability to deal with non-linear case (e.g. MAR), large sample size requirement (e.g. SEM), et., al. In this research, we propose applying Probabilistic Boolean Network (PBN) for modeling brain connectivity due to its solid stochastic properties, computational simplicity, robustness to uncertainty, and capability to deal with small-size data, typical for fIVIRI data sets. Applying the proposed PBN framework to real fMRI data recorded from PD subjects enables us to identify statistically significant abnormality in PD connectivity by comparing it with normal subjects. The PBN results also suggest a mechanism of evaluating the effectiveness of L-dopa, the principal treatment for PD. In addition to PBNs’ promising application in inferring brain connectivity, PBN modeling for brain ROTs also enables researchers to study dynamic activities of the system under stochastic conditions, gaining essential information regarding asymptotic behaviors of ROTs for potential therapeutic intervention in PD. The results indicate significant difference in feature states between PD patients and normal subjects. Hypothesizing the observed feature states for normal subject as the desired functional states, we further explore possible methods to manipulate the dynamic network behavior of PD patients in the favor of the desired states from the view of random perturbation as well as intervention. Results identified a target ROT with the best intervention performance, and that ROl is a potential candidate for therapeutic exercise. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

No description available.

To understand how the brain functions, we must investigate the transient interactions that underpin communication between cortical regions. EEG possesses the optimal temporal resolution to capture functional connectivity, but it lacks the spatial resolution to identify the cortical locations responsible. To circumvent this problem electrophysiological connectivity should be investigated at the source level. There are many quantifiers of connectivity applied to EEG data, but some are not sensitive to the direct, or indirect, influence of one region over another, and others require the specification of a priori models so are unsuitable for exploratory analyses. Transfer Entropy (TE) can be used to infer the direction of linear and non-linear information exchange between signals over a range of time-delays within EEG data. This thesis explores the creation of a new method of mapping dynamic brain connectivity using a trial-based TE analysis of EEG source data, and the application of this technique to the investigation of semantic and number processing within the brain. The first paper (Chapter 2) documents the analyses of a semantic category and number magnitude judgement task using traditional ERP techniques. As predicted, the well-known semantic N400 component was found, and localised to left ATL and inferior frontal cortex. An N365 component related to number magnitude judgement was localised to right superior parietal regions including the IPS. These results offer support for the hub-and-spoke model of semantics, and the triple parietal model of number processing. The second paper (Chapter 3) documents an analysis of the same data with the new trial-based TE analysis. Word and number data were analysed at 0-200ms, 200-400ms, and 400-600ms following stimulus presentation. In the earliest window, information exchange was occurring predominately between occipital sources, but by the latest window it had become spread out across the brain. Task-dependent differences of regional information exchange revealed that temporal sources were sending more information to occipital sources following words at 0-200ms. Furthermore, the direction and timing of information movement within a front-temporal-parietal network was identified during 0-400ms of the number magnitude judgment. The final paper (Chapter 4), documents an attempt to track the influence of TMS through the brain using the TE analysis. TMS was applied to bilateral ATL and IPS because they are both important hubs in the brain networks that support semantic and number processing respectively. Left ATL TMS influenced sources located primarily in wide-spread left temporal lobe, and inferior frontal and inferior occipital cortices. The anatomical connectivity profile of the temporal lobe suggests that these are all plausible locations, and they exhibited excellent spatial similarities to the results of neuroimaging experiments that probed semantic knowledge. The analysis of right ATL TMS obtained a mirror image of the left. Left parietal stimulation resulted in a bilateral parietal, superior occipital, and superior prefrontal influence, which extended slightly further in the ipsilateral hemisphere to stimulation site. A result made possible by the short association and callosal fibres that connect these areas. Again, the results at the contralateral site were a virtual mirror image. The thesis concludes with a review of the experimental findings, and a discussion of methodological issues still to be resolved, ideas for extensions to the method, and the broader implications of the method on connectivity research.

It is well established that childhood adversity is associated with long lasting effects on development including both negative physical and mental health outcomes. Research demonstrates that adverse childhood experiences influence neurodevelopment and propose that this may be a mechanism linking adversity and psychopathology. However, little is known how the timing and type of maltreatment experiences may differentially impact longitudinal changes in neural processes of risk-related decision making. Using conditional growth curve modeling, we examined how abuse and neglect across three developmental periods (early childhood, school age, and adolescence) are associated with longitudinal changes in task-based functional connectivity during risk-processing and cognitive control. The current sample included 167 adolescents (13-14 years old at Time 1; 53% male), assessed annually for six years. At each of the six time points, adolescents completed a lottery choice task and a cognitive control task while blood-oxygen-level-dependent (BOLD) responses were monitored with functional magnetic resonance imaging (fMRI). Adolescents reported on maltreatment experiences occurring during ages 1 to 18. Generalized psychophysiological interactions (gPPI) was used to examine task- based functional connectivity in the insula and dACC (dorsal anterior cingulate cortex) during both risk processing and cognitive control, respectively. Although no sensitive periods emerged for the effects of abuse or neglect on functional connectivity during risk processing, chronic abuse (abuse occurring in more than one developmental period) significantly predicted weaker insula-dACC connectivity in late adolescence. For functional connectivity during cognitive control, adolescence emerged as a potential sensitive period for neglect, such that those with neglect experiences occurring during ages 13 to 18 showed slower improvements in dACC- insula connectivity across adolescence. Chronic neglect was also associated with slower improvements in dACC-insula connectivity. Additionally, chronic abuse was significantly associated with stronger improvements in dACC-insula connectivity across adolescence. Collectively, these results suggest that abuse may be linked to a delayed maturation in neural connectivity associated with valuation, but an accelerated maturation in neural connectivity associated with cognitive control. Furthermore, neglect may be linked to a delayed maturation in neural connectivity associated with cognitive control. Both sets of findings involved functional connectivity in both the dACC and insula, important regions involved in salience processing. These findings elucidate the distinct effects of abuse and neglect on connectivity in regions involved in risk-related decision making, including valuation and cognitive control. Future work will benefit from examining how these different pathways may lead to outcomes such as health risk behaviors and psychopathology. / M.S. / Childhood adversity is associated with long lasting effects on development including both negative physical and mental health outcomes. Research shows that adverse childhood experiences influence brain development. However, little is known how the timing and type of maltreatment experiences may differentially impact changes in brain processes of risky decision making across adolescence. We examined how abuse and neglect across three developmental periods (early childhood, school age, and adolescence) are associated with changes in functional connectivity during risk-processing and cognitive control. The current sample included 167 adolescents (13-14 years old at Time 1; 53% male), assessed annually for six years. At each of the six time points, adolescents completed a lottery choice task and a cognitive control task while blood-oxygen-level-dependent (BOLD) responses were monitored with functional magnetic resonance imaging (fMRI). Adolescents reported on maltreatment experiences occurring during ages 1 to 18. Generalized psychophysiological interactions (gPPI) was used to examine task- based functional connectivity in the insula and dACC (dorsal anterior cingulate cortex) during both risk processing and cognitive control, respectively. Results showed that chronic abuse (abuse occurring in more than one developmental period) significantly predicted weaker insula- dACC connectivity in late adolescence. For functional connectivity during cognitive control, those with neglect experiences occurring during ages 13 to 18 showed slower improvements in dACC-insula connectivity across adolescence. Chronic neglect was also associated with slower improvements in dACC-insula connectivity. Additionally, chronic abuse was significantly associated with stronger improvements in dACC-insula connectivity across adolescence. Both sets of findings involved functional connectivity in both the dACC and insula, important regions involved in salience processing. These findings elucidate the distinct effects of abuse and neglect on connectivity in regions involved in risk-related decision making, including valuation and cognitive control. Future work will benefit from examining how these different pathways may lead to outcomes such as health risk behaviors and psychopathology.

Understanding pathways of neurological disorders requires extensive research on both functional and structural characteristics of the brain. This dissertation introduced two interrelated research endeavors, describing (1) a novel integrated approach for constructing functional connectivity networks (FCNs) of brain using non-invasive scalp EEG recordings; and (2) a decision aid for estimating intracranial volume (ICV). The approach in (1) was developed to study the alterations of networks in patients with pediatric epilepsy. Results demonstrated the existence of statistically significant (p

Preterm birth (defined as birth before 37 weeks) is a leading cause of neurocognitive impairment in childhood, including difficulties in social cognition and executive function. Microstructural divergence from typical brain development in the preterm brain can be quantified using diffusion magnetic resonance imaging (dMRI) tractography during the neonatal period. The relationship between dMRI tractography metrics and later cognitive difficulties remains inconclusive. A general measure of white matter microstructure (gWM) offers a neural basis for cognitive processes in adults, however it remains unclear when gWM is first detectable in the developmental trajectory. Eye-tracking is a technique which assesses eye-gaze behaviour in response to visual stimuli, which permits inference about underlying cognitive processes, such as social cognition and executive function in infancy. The primary aims of this thesis were to test the hypotheses: dMRI tractography reveals significant differences in tract-average fractional anisotropy (FA) and mean diffusivity (MD) between preterm and term infants, and variance in tract-average FA and MD is shared across major tracts. Secondly, infants born preterm have altered social cognition and executive function compared to term born peers, assessed by eye-tracking and finally, neonatal MRI gWM is associated with cognitive function in infancy. Preterm (birth weight ≤ 1500g) and term infants (born ≥ 37 weeks’ post-menstrual age [PMA]) were recruited and underwent a MRI scan at term equivalent age (between 38 - 42 weeks’ PMA) and an eye-tracking assessment six to nine months later. Preterm infants were assessed at two years using the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III). dMRI tractography metrics were generated using probabilistic neighbourhood tractography (PNT) in eight pre-defined tracts-of-interest. Principal component analyses (PCA) were used to determine the correlations between the eight tracts-of-interest for four tract-averaged water diffusion parameters. dMRI metrics were compared to the eye-tracking performance and two year outcome data. Quantitative microstructural changes were identifiable within the preterm brain when compared to infants born at term. PCA revealed a single variable that accounts for nearly 50% of shared variance between tracts-of-interest, and all tracts showed positive loadings. Eye-tracking revealed group-wise differences in infant social cognition, attributable to preterm birth, but executive functions inferred from eye-tracking did not differ between groups. dMRI tractography metrics within the neonatal period did not relate to later outcome measures. This thesis shows that variance in dMRI parameters is substantially shared across white matter tracts of the developing brain and suggests that anatomical foundations of later intelligence are present by term equivalent age. Social cognition is altered by preterm birth, however social cognitive ability in infancy is independent of gWM.

A better, more optimized processing pipeline for functional connectivity (FC) data will likely accelerate practical advances within the field of neuroimaging. When using correlation-based measures of FC, researchers have recently employed a few data-driven methods to maximize its predictive power. In this study, we apply a few of these post-processing methods in both task, twin, and subject identification problems. First, we employ PCA reconstruction of the original dataset, which has been successfully used to maximize subject-level identifiability. We show there is dataset-dependent optimal PCA reconstruction for task and twin identification. Next, we analyze FCs in their native geometry using tangent space projection with various mean covariance reference matrices. We demonstrate that the tangent projection of the original FCs can drastically increase subject and twin identification rates. For example, the identification rate of 106 MZ twin pairs increased from 0.487 of the original FCs to 0.943 after tangent projection with the logarithmic Euclidean reference matrix. We also use Schaefer’s variable parcellation sizes to show that increasing parcellation granularity in general increases twin and subject identification rates. Finally, we show that our custom convolutional neural network classifier achieves an average task identification rate of 0.986, surpassing state-of-the-art results. These post-processing methods are promising for future research in functional connectome predictive modeling and, if optimized further, can likely be extended into clinical applications.

This dissertation introduces a novel approach at gauging patterns of informa- tion flow using brain connectivity analysis and partial directed coherence (PDC) in epilepsy. The main objective of this dissertation is to assess the key characteristics that delineate neural activities obtained from patients with epilepsy, considering both focal and generalized seizures. The use of PDC analysis is noteworthy as it es- timates the intensity and direction of propagation from neural activities generated in the cerebral cortex, and it ascertains the coefficients as weighted measures in formulating the multivariate autoregressive model (MVAR). The PDC is used here as a feature extraction method for recorded scalp electroencephalograms (EEG) as means to examine the interictal epileptiform discharges (IEDs) and reflect the phys- iological changes of brain activity during interictal periods. Two experiments were set up to investigate the epileptic data by using the PDC concept. For the investigation of IEDs data (interictal spike (IS), spike and slow wave com- plex (SSC), and repetitive spikes and slow wave complex (RSS)), the PDC analysis estimates the intensity and direction of propagation from neural activities gener- ated in the cerebral cortex, and analyzes the coefficients obtained from employing MVAR. Features extracted by using PDC were transformed into adjacency matrices using surrogate data analysis and were classified by using the multilayer Perceptron (MLP) neural network. The classification results yielded a high accuracy and pre- cision number. The second experiment introduces the investigation of intensity (or strength) of information flow. The inflow activity deemed significant and flowing from other regions into a specific region together with the outflow activity emanating from one region and spreading into other regions were calculated based on the PDC results and were quantified by the defined regions of interest. Three groups were considered for this study, the control population, patients with focal epilepsy, and patients with generalized epilepsy. A significant difference in inflow and outflow validated by the nonparametric Kruskal-Wallis test was observed for these groups. By taking advantage of directionality of brain connectivity and by extracting the intensity of information flow, specific patterns in different brain regions of interest between each data group can be revealed. This is rather important as researchers could then associate such patterns in context to the 3D source localization where seizures are thought to emanate in focal epilepsy. This research endeavor, given its generalized construct, can extend for the study of other neurological and neurode- generative disorders such as Parkinson, depression, Alzheimers disease, and mental illness.