Jagust Lab About Research Lab news Publications People Links        Member login

Imaging Interactions between Alzheimer's disease and cerebrovascular disease

SALSA

Computer Based Therapy

Imaging Brain Dopamine in Aging

Imaging Gene Therapy

Imaging Brain Amyloid in Vivo

Alzheimer’s Disease Neuroimaging Initiative
up to Research page 1 of 7 continue to SALSA
It is now well known that individuals with cerebrovascular disease (CVD) have a higher prevalence of AD and dementia, and that both vascular disease and AD pathology occur in the brain together in many older people. However, the mechanisms of this interaction are far less clear, especially in regard to small vessel vascular disease. This is the most common form of CVD, and does not result in large strokes, but causes small infarcts and changes in white matter (white matter hyperintensities, or WMH). A major hypothesis that guides this project is that underlying changes in white matter due to small vessel disease disconnect the frontal lobes from subcortical structures, producing cognitive decline that reflects frontal lobe damage. This frontal damage interacts with damage to the medial temporal-hippocampal memory system that is dysfunctional in AD, to produce severe cognitive deficits. The project is enrolling and longitudinally following a large cohort of older people with and without vascular disease, who undergo repeated cognitive testing, structural MRI scanning, and PET scanning with the glucose metabolic tracer [18F]fluorodeoxyglucose (FDG). Our goals are to define how subcortical CVD affects the FDG-PET scan, and how AD affects the FDG-PET scan, and define how the FDG-PET scan is related to underlying structural changes in WMH and hippocampal volume seen on the MRI. Evidence from this project so far confirms our hypotheses that subcortical vascular disease results in frontal lobe hypometabolism. We are in the process of analyzing longitudinal data to try to unravel how these metabolic deficits interact with structural and metabolic deficits produced by AD.

One of the techniques we have found useful for this project involves "virtual neurosurgery". Using this approach we can " carve" 3D brain volumes using intersecting planes. These brain volumes are obtained from MR datasets that have been previously coregistered to the PET datasets. This permits us to measure the structural volume (for tissue types, pathologies, etc.) and relate the brain structures to metabolic or chemical changes that have occurred in the PET scan. The images at right show a brain that has been carved into these separate volumes.