Neuroscience 2000 Abstract
| Presentation Number: | 278.13 |
|---|---|
| Abstract Title: | Multivariate analysis of pharmacologically induced changes in MRI responses in alert rhesus monkeys. |
| Authors: |
Andersen, A. H.*1
; Rayens, W. S.1
; Gerhardt, G. A.1
; Schmitt, F. A.1
; Zhang, Z.1
; Gash, D. M.1
1Departments of Anatomy and Neurobiology, Statistics, Neurology, and the Magnetic Resonance Imaging and Spectroscopy Center, University of Kentucky College of Medicine, Lexington, KY |
| Primary Theme and Topics |
J. Disorders of the Nervous System and Aging - 131. Degenerative disease: Parkinson's |
| Session: |
278. Degenerative disease: Parkinson's--clinical I Poster |
| Presentation Time: | Monday, November 6, 2000 8:00 AM-9:00 AM |
| Location: | Hall G-J |
| Keywords: | Dopamine, Nigrostriatal System, Parkinson, BOLD |
Functional magnetic resonance imaging (fMRI) has been widely used for mapping blood oxygen level dependent (BOLD) signals as a marker of neuronal activity in the conscious human and nonhuman primate brain. Our group has reported mapping drug-induced changes in the basal ganglia by multiple gradient recalled echo fMRI in anesthetized as well as in alert rhesus monkeys. In addition to extensive demands placed on training and animal handling, new techniques of data analysis and statistical assessment are required to appropriately evaluate the pharmacological response as measured by MRI in alert animals. In this study we examined the use of robust multivariate statistical methods such as principal component analysis (PCA). These methods are data driven and make no a priori assumptions about timing or dynamic nature of the drug-induced temporal response. However, since they are based on variance summaries, the resulting principal components of the response are heavily influenced by the presence of aberrant or spurious effects in the data such as those caused by head movement artifacts. In fact, the desired response structure associated with pharmacological stimulation most often will not consolidate into a single component which summarizes the most variability, or it may be a nontrivial admixture of several principal components. To address these issues, the analysis must incorporate statistical influence measures for PCA to identify observations that disproportionately inflate the variance. These observations can then be downweighted or omitted so as to effectively focus the analysis, thus enabling us to detect small differences in the response associated with changes in brain function during normal aging or in Parkinson's disease.
Supported by USPHS grants NIA AG13494, NIMH MH012145 and NINDS NS39787.
Sample Citation:
[Authors]. [Abstract Title]. Program No. XXX.XX. 2000 Neuroscience Meeting Planner. New Orleans, LA: Society for Neuroscience, 2000. Online.
Copyright © 2000-2025 Society for Neuroscience; all rights reserved. Permission to republish any abstract or part of any abstract in any form must be obtained in writing by SfN office prior to publication.
