HEMISPHERIC ASYMMETRY ANALYSES THROUGH COMPUTATIONAL NEUROSCIENCE MODELS WITH EMPHASIS ON EEG MICROSTATES : EEG-FMRI DATA INTEGRATION APPROACH

Abstract

It is often assumed that there is a direct correlation between the knowledge an individual possesses and that individual's actions. However, many hidden processes influence decisionmaking processes. Asymmetric processing of affective, cognitive, and sensory information has long been one of the fascinating properties of human brain function. Thus, understanding hemispheric asymmetry as one of those hidden processes can bridge the gap between what a person knows and what one decides to do. One widely used technique for analysis of brain asymmetry is Electroencephalography (EEG), whose simplicity, portability, and high temporal resolution enable its usage in a relatively wide-range of real-world environments. Howbeit, it also poses a drawback of less spatial resolution as the localization of an active site is limited to several centimeters. The hemispheric difference between EEG alpha activity over the frontal regions has been termed as frontal EEG asymmetry. This phenomenon was first linked to patterns of emotion processing decades ago. functional Magnetic Resonance Imaging (fMRI) is another technique that provides a unique view of the human brain by detecting changes in blood oxygenation. It poses an advantage of high spatial resolution; however, it possesses a low temporal resolution. The hemispheric dominance in fMRI has been called the laterality index. The laterality index enables one value per subject per contrast as a descriptor for activation pattern based hemispheric dominance. Also, vii simultaneous recordings and analysis of EEG-fMRI techniques, which can counteract the limitations posed by EEG-FMRI, have recently gained attention and can be gauged for effectiveness in the hemispheric asymmetry research. The current thesis aims to corroborate the hemispheric asymmetry research by exploring the resting-state EEG/fMRI hemispheric asymmetry models after simultaneous EEG-fMRI data acquisition. These resting-state models of hemispheric asymmetry in the brain may serve as potential parameters for comprehending the human actions when engaged in any exogenously directed task. Thus, the standard resting frontal alpha EEG hemispheric asymmetry model was first examined before engagement in a Situational awareness (SA) task to vindicate the relationship between pre-task resting asymmetry and SA-task performance. SA is the knowledge of the environment, and maintenance of SA is crucial for optimal performance in the aviation and military domain. Thus, understanding the linkage of the neural mechanisms underlying the pre-task resting frontal alpha asymmetry model with subsequently performed SA-task can improve SA. For this purpose, initially, an SA-task with influence from the Stroop effect was designed and developed, and pre-task resting EEG absolute alpha power and its frontal alpha hemispheric asymmetry were assessed. This study revealed a strong association of SA-task performance measures with resting frontal alpha hemispheric asymmetry. Further, the neural mechanisms underlying pre-SA task resting absolute alpha power and its frontal asymmetry, as assessed through the EEG-informed fMRI approach, significantly influenced the SA-task's neural mechanisms. After examining the relationship between the pre-task resting alpha EEG asymmetry model with subsequently performed SA-task, the association of this standard asymmetry model with affect and approach/withdrawal measures of an individual was gauged. The purpose of this viii study was to understand the significance of real-time standalone recordings of pre-task resting alpha EEG asymmetry in terms of its connectedness with measures of positive/negative affect and approach/withdrawal behavior. Further, to strengthen the findings, the mapping between pretask resting alpha EEG asymmetry model and fMRI through EEG-informed fMRI analysis was explored. For this purpose, initially, the robust correlation of standard resting frontal alpha asymmetry with affect and approach/withdrawal measures was carried out. Next, the neural underpinnings and Hemodynamic Lateralization Index, HLI (based on these neural underpinnings) for standard resting frontal alpha asymmetry were assessed. The results yielded no significant relationship between the standard resting frontal asymmetry and its HLI with any psychological measures. This ambivalence on the validity of standard resting frontal alpha asymmetry in terms of its association with affect and approach/withdrawal psychological measures motivated us towards the estimation of a novel microstate-based frontal alpha asymmetry model and assessment of this model’s linkage with positive/negative affect and approach/withdrawal measures. The microstates represent global electrical brain activity on the scalp that remains semi-stable for brief transient periods. The utilization of microstates was based on the evidence that supported the importance of stable EEG patterns in bringing forth the interrelation of affect and approach/withdrawal measures with resting frontal alpha asymmetry. The results revealed that the microstate-based resting frontal alpha asymmetry model correlated significantly with negative affect, and its neural underpinning’s HLI significantly correlated with positive/Negative affect and approach/withdrawal measures. Thus, the novel microstate-based microstate-based resting frontal alpha asymmetry model proved efficacious in bringing forth the association with affect and approach/withdrawal measures. ix In addition, to understand the role of subcortical regions, and their interaction with cortical regions in bringing forth the hemispheric asymmetries of affect and approach/withdrawal behavior, a study based on the hemispheric asymmetry model of resting fMRI graph theory functional connectivity metrics was carried out, as the viability to detect subcortical signals through EEG is still debatable. In this analysis, we report the neuroimaging finding based on Region of Interest (ROI) based analysis and graph-theory measures for global networks and subnetworks. The study revealed the involvement of emotion and memory-related subcorticalcortical interactions in positive and negative affect and basal ganglia structures in approachwithdrawal dichotomy. Further, lateralization of the strength of degree-measures of the corticalregions vital for subcortical-cortical interaction revealed higher connectivity within the lefthemisphere for affective measures. Thus, the current thesis demonstrates the benefit of assessing the standard resting hemispheric asymmetry model before a complex cognitive task such as SA, which holds paramount importance for the ergonomics community and for military/aviation domains. Further, the outcomes also offer an unprecedented attempt towards the development of a novel microstates-based resting hemispheric asymmetry model for bringing forth the relationship of resting EEG based asymmetry with psychological measures of affect and approach/withdrawal behavior Also, the key findings of subcortical regions and their interaction with cortical regions dominating the affect and approach/withdrawal measure can be further explored in clinical as well as task-based studies.