Date of Award

Spring 5-15-2017

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



Alzheimer’s disease (AD) affected approximately 48 million people worldwide in 2015. Its devastating consequences have stimulated an intense search for AD prevention and treatment. Clinically, AD is characterized by memory deficits and progressive cognitive impairment, leading to dementia. Over the past two to three decades, researchers have found that amyloidbeta (Aβ) plaques and neurofibrillary tau tangles occur during a long pre-symptomatic period (preclinical stage) before the onset of clinical symptoms. As a result, identification of the preclinical stage is essential for the initiation of prevention trials in asymptomatic individuals. Currently, Positron Emission Tomography (PET) imaging with injected 11C or 18F containing radiotracers (e.g., Pittsburgh compound B, PiB or florbetapir-fluorine-18, 18F-AV-45) is widely used to detect amyloid deposition in vivo and to identify this preclinical stage. However, PET scans are time consuming (about 1 hour), require injection of a radiotracer, thus, exposing the patient to ionizing radiation. After the preclinical stage, AD patients begin to show clinical symptoms, referred as a very mild or mild AD group. Post-mortem studies show that neuronal damage is the most proximate pathological substrate of cognitive impairment in AD compared with amyloid and tau deposition. Thus, a diagnostic tool is needed for detection of neuronal loss in vivo. As a faster, non-invasive, and radiation free imaging technique, Magnetic Resonance Imaging (MRI) plays an important role in the diagnosis of cognitive diseases. Conventional MRI yields superb definition of brain anatomy and structure and provide important volumetric information (e.g., brain atrophy). However, conventional MRI cannot provide microstructural and functional insight into the pathology of AD. The approach developed in Yablonskiy’s lab is based on the Gradient Echo Plural Contrast Imaging (GEPCI) protocol, which provides quantitative in vivo measurements of transverse relaxation properties of the tissue water 1H spins as determined from the gradient echo MRI signal. The measurements are corrected for macroscopic magnetic field inhomogeneity effects and physiologic-motion-driven fluctuations in magnetic field as these are the major artifacts present with the gradient echo technique. The principal relaxation property used in this dissertation is the tissue-specific transverse relaxation rate constant, R2*. The R2* value reflects the microscopic and mesoscopic magnetic field inhomogeneities rising from the complex tissuewater-environment within the human brain. In turn, changes in R2* reflect changes in the tissue’s microscopic and mesoscopic tissue structure. However, because of the presence of the cerebral blood vessel network, the magneticsusceptibility-driven blood-oxygen-level dependent (BOLD) effect also makes a significant contribution to R2*. A previously developed approach, quantitative BOLD (qBOLD), allows the separation of R2* into a tissue specific R2t* without blood vessel effects and the BOLD component. Quantifying the BOLD component allows the calculation of cerebral hemodynamics parameters, such as oxygen extraction fraction (OEF) and deoxygenated cerebral blood volume (dCBV). These parameters (R2*, R2t*, OEF, dCBV) describe structural and functional properties of tissue at the microstructural level in the human brain. In the study of normal aging, quantitative GEPCI measurements showed that R2t* increases with age while hemodynamic parameters, i.e., relative OEF and dCBV remain constant in most cerebral cortical regions. The comparison between quantitative GEPCI measurements and literature information suggest that (a) age-related increases in the cortical R2t* mostly reflect the age-related increases in the cellular packing density (or neuronal density); (b) regions in a brain characterized by higher R2t* contain a higher concentration of neurons with less developed cellular processes (dendrites, spines, etc.); and (c) brain regions characterized by lower R2t* represent regions with lower concentration of neurons but more developed cellular processes. In the Alzheimer study, R2* and R2t* together demonstrated significant differences among the normal, preclinical and mild AD groups. First, the results uncovered strong correlations between R2* and Aβ deposition measured by the PiB PET-tracer in several cortical regions (e.g., medial temporal lobe and precuneus). This finding indicates that R2* may be a potential surrogate marker for Aβ deposition. The strongest correlation was found in the medial temporal lobe (MTL), particularly in the parahippocampal cortex, which can be used to distinguish the normal and preclinical groups. Second, R2t* in the hippocampus, which characterized the hippocampal cellular integrity demonstrated much stronger correlations with psychometric tests than volume quantification of hippocampal atrophy. Importantly, decreased R2t* characterizing cellular damage was detected even in the hippocampal areas not affected by atrophy. In addition, R2t* significantly decreased in the mild AD group but was preserved in the preclinical group compared with the normal group. These results indicate a significant cellular density decrease in the mild group but not in the preclinical group, which is consistent with previous histological studies. In summary, GEPCI provides a new approach for evaluation of AD-related tissue pathology in vivo in the preclinical and early symptomatic stages of AD. Since MRI is widely available worldwide and does not require radiation exposure, it provides the opportunity to obtain new information on the pathogenesis of AD and for pre-screening cohorts (stratification) for clinical drug trials.


English (en)

Chair and Committee

Joseph Ackerman

Committee Members

Alexander Barnes, Anne Cross, Dewey Holten, Amitriy Yablonskiy


Permanent URL: