MR SPECTROSCOPIC IMAGING IN ALZHEIMER'S DISEASE

Stanford University, Stanford CA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (Verbatim from the Applicant's Abstract): This application proposes to study brain structure and function assessed with proton MR spectroscopic imaging (MRSI) and cognitive neuropsychology I-Alzheimer's disease (AD). MRSI is a powerful approach for addressing important questions about the biology of the living human brain in health and disease. It is a safe, noninvasive method for longitudinal study, which is the essential design for characterizing disease progression and treatment monitoring. We have developed a method for the quantification of regional metabolite concentrations in gray matter and white matter separately. The aims combine simultaneous MRSI acquisition and image analysis development with clinical investigation for the following metabolites: N-acetyl compounds (predominantly N-acetyl-acetic acid, NAA), creatine (Cr), choline (Cho), and myo-inositol (rnI). This work includes the imaging of metabolites that have specificity for neurons (NAA) and for glia (mI). Upon successful completion of this work, the tools developed can be used to monitor disease progression in AD as well as treatment response. Robust regional quantification of these proton spectroscopic metabolites will permit rigorous investigation of the functional ramif aboutcations of brain biochemical status. Our Specific Aims are: 1) to conduct a longitudinal study of AD using a long-echo, 3D volumetric MRSI protocol with regional supratentorial quantification, and a short-echo, single slice MRSI protocol with sampling of the hippocampus; 2) to develop a method for short-echo, volumetric MRSI that will provide regional measures of mI as well as NAA, Cr, and Cho throughout the brain and to conduct a cross-sectional study of AD to test the new method. The cognitive correlates of the regional MRSI data will be examined by testing specific brain-behavior relationships through double dissociation models.