|
Abstract:
|
One major goal of dynamic positron emission tomography (PET) imaging is the estimation of the spatial distribution of specific molecules throughout the brain. The generally accepted analytic strategy for dynamic PET data involves first fitting parametric models to each voxel/region time series, extracting a single summary measure from each model fit, and finally, conducting subsequent analysis on the summary measures to examine effects of factors such as age, sex, diagnostic group, etc. We will present an alternative flexible approach for the analysis of such data that is based on principles of functional data analysis and show the advantages of such an approach both with simulated data and real data analysis.
|
