UCLA

Persistent tic disorder (PTD) is a neuropsychiatric disorder characterized by involuntary and stereotypical movements called tics. Despite a multitude of research, questions still remain regarding the full neurological bases of tic generative activity, as well as the degree to which certain cognitive functions are impaired in children with PTD. This dissertation aims to further elucidate the atypical neural dynamics associated with PTD through an EEG perspective. To facilitate the investigation of these neural dynamics, we first evaluate an adapted approach for performing group-level analyses on localized EEG measures at the cortical source level and compares performance to classical methods. We examined three approaches (a voxel approach, a region-of-interest (ROI) approach, and a k-means clustering approach) for detecting group differences in spectral activity in both a simulation analysis and visual attention task. Overall, we found that a voxel-approach produces reduced localization error, reduced spectral attenuation, and more accurate time-frequency detection of spectral effects, posing the method as an effective analysis approach. We then examined the degree to which behavioral performance and neural dynamics are atypical in children with PTD (compared to typically developing children) during an inhibitory control flanker paradigm, using measures of spectral power and effective connectivity. While task accuracy did not differ by diagnosis, children with PTD exhibited attenuated spectral activity in the anterior cingulate cortex (ACC), alongside greater information flow from the ACC to fronto-parietal network hubs relative to controls, while controls showed greater central and posterior connectivity. Correlations with clinical features (e.g., tic impairment) and task performance indicate these atypical activations may be neural adaptations from frequent engagement of inhibitory control pathways. Lastly, utilizing similar measures, we investigated the neural antecedents of tic expression, and the degree to which this activity is differentiable from normal resting state activity using a machine learning approach. Prior to tic occurrence, we observed increased spectral activity in the ACC, as well as changes in information flow in frontal, sensorimotor, and posterior regions, suggesting aberrant communication among multiple cognitive and sensory regions. These measurements were also found to be reliable discriminators between pre-tic activity and tic-free activity, using a Na�ve Bayes classification model. We then discuss future directions for each of the three preceding studies, as well as a potential path for developing more effective treatment protocols in PTD.