Abnormal large-scale neural dynamics and the role of secondary thalamic dysfunction in stroke

Abstract

Stroke is a leading cause of chronic disability globally, affecting virtually all domains of human brain function. Although tissue death due to stroke is typically localized, neural activity can be disrupted widely throughout surviving brain tissue, as evidenced by the broad shift towards slower dynamics consistently observed with electroencephalography (EEG) and magnetoencephalography (MEG) following stroke. Associations between abnormal slow activity and poorer outcomes suggest it could serve as a therapeutically-useful marker of underlying brain dysfunction, but this has yet to inform applications that meaningfully improve patient recovery. The primary goals of this thesis were therefore to better characterize these abnormal neural dynamics and identify their neurophysiological causes, thereby promoting biomarker and treatment development. Using MEG recordings from chronic stroke patients, Study 1 first characterized post-stroke slow activity in both the frequency and time domains, finding that elevated slow activity was better explained by a change in aperiodic neural dynamics --- not elevated delta oscillations, as is commonly assumed. However, oscillatory abnormalities were indeed found at higher frequencies, namely alpha oscillation slowing and reduced beta oscillation amplitude. Study 2 directly investigated the causes of this abnormal activity using a computational corticothalamic circuit model capable of emulating both aperiodic and periodic neural dynamics. Fitting these models to MEG data identified thalamic dysfunction as the cause of the abnormal dynamics, despite the absence of direct thalamic injury in these patients. Specifically, the model identified abnormally reduced inhibition in the thalamus as the cause of slowing, and this was reflected in magnetic resonance imaging (MRI) measures indicating secondary thalamic injury. Finally, Study 3 replicated the major findings of both prior studies using EEG recordings from a group of subacute and chronic stroke patients. By linking electrophysiological slowing to secondary thalamus dysfunction, this work explains a longstanding finding in stroke research, establishes electrophysiological slowing as a potentially useful biomarker for stroke research and rehabilitation, and highlights the thalamus as an important target for future development of therapeutic interventions.