Date of Award
Doctor of Philosophy (PhD)
Although structural damage from stroke is focal, remote dysfunction can occur in regions of the brain distant from the area of damage. Lesions in both gray and white matter can disrupt the flow of information in areas connected to or by the area of infarct. This is because the brain is not an assortment of specialized parts but an assembly of distributed networks that interact to support cognitive function. Functional connectivity analyses using resting functional magnetic resonance imaging (fMRI) have shown us that the cortex is organized into distributed brain networks. The primary goal of this work is to characterize the effects of stroke on distributed brain systems and to use this information to better understand neural correlates of deficit and recovery following stroke. We measured resting functional connectivity, lesion topography, and behavior in multiple domains (attention, visual memory, verbal memory, language, motor, and visual) in a cohort of 132 stroke patients. Patients were followed longitudinally with full behavioral and imaging batteries acquired at 2 weeks, 3 months, and 1 year post-stroke. Thirty age- and demographic- matched controls were scanned twice at an interval of three months.
In chapter 1, we explore a central question motivating this work: how is behavior represented in the brain? We review progressing prospective – from basic functional localization to newer theories connecting inter-related brain networks to cognitive operations. In so doing, we attempt to build a foundation that motivates the hypotheses and experimental approaches explored in this work.
Chapters 2 and 3 serve primarily to validate approaches and considerations for using resting fMRI to measure functional connectivity in stroke patients. In chapter 2, we investigate hemodynamic lags after stroke. ‘Hemodynamic lag’ is a local delay in the blood oxygen level dependent (BOLD) response to neural activity, measured using cross-correlation of local fMRI signal with some reference brain signal. This work tests assumptions of the BOLD response to neural activity after stroke, but also provides novel and clinically relevant insight into perilesional disruption to hemodynamics. Significant lags are observed in 30% of stroke patients sub-acutely and 10% of patients at one-year. Hemodynamic lag corresponds to gross aberrancy in functional connectivity measures, performance deficits and local and global perfusion deficits. Yet, relationships between functional connectivity and behavior reviewed in chapter 1 persist after hemodynamic delays is corrected for. Chapter 3 provides a more extended discussion of approaches and considerations for using resting fMRI to measure functional connectivity in stroke patients. Like chapter 1, the goal is to motivate experimental approaches taken in later chapters. But here, more technical challenges relating to brain co-registration, neurovascular coupling, and clinical population selection are considered.
In chapter 4, we uncover the relationships between local damage, network wide functional disconnection, and neurological deficit. We find that visual memory and verbal memory are better predicted by connectivity, whereas visual and motor deficits are better predicted by lesion topography. Attention and language deficits are well predicted by both. We identify a general pattern of physiological network dysfunction consisting of decrease of inter-hemispheric integration and decrease in intra-hemispheric segregation, which strongly related to behavioral impairment in multiple domains.
In chapter 5, we explore a case study of abulia – severe apathy. This work ties together principles of local damage, network disruption, and network-related deficit and demonstrates how they can be useful in understanding and developing targeted treatments (such as transcranial magnetic stimulation) for individual stroke patients.
In chapter 6, we explore longitudinal changes in functional connectivity that parallel recovery. We find that the topology and boundaries of cortical regions remains unchanged across recovery, empirically validating our parcel-wise connectivity approach. In contrast, we find that the modularity of brain systems i.e. the degree of integration within and segregation between networks, is significantly reduced after a stroke, but partially recovered over time. Importantly, the return of modular network structure parallels recovery of language and attention, but not motor function. This work establishes the importance of normalization of large-scale modular brain systems in stroke recovery.
In chapter 7, we discuss some fundamental revisions of past lesion-deficit frameworks necessitated by recent findings. Firstly, anatomical priors of structural and functional connections are needed to explain why certain lesions across distant locations should share behavioral consequences. Secondly, functional priors of connectomics are needed to explain how local injury can produce widespread disruption to brain connectivity and behavior that have been observed.
Chair and Committee
David Van Essen
Maurizio Corbetta, Nico Dosenbach, Jin-moo Lee, Abraham Snyder,
Siegel, Joshua Sarfaty, "Understanding Stroke in the Connected Human Brain" (2018). Arts & Sciences Electronic Theses and Dissertations. 1580.