Meeting Banner
Abstract #2633

Hierarchical Linear Modeling of longitudinal Magnetic Resonance Spectroscopic Imaging

Yeh P, Meyererhoff D, Thayyullathil H, Kornak J
Northern California Institute for Research and Education

A comprehensive analysis of biomedical longitudinal data demands accommodation of between-subjects and within-subject variation, unbalanced designs and missing data. The longitudinal spectroscopic imaging data even complicate the problem by its large sources of variability. The goal of this study is to apply linear mixed models in the assessment of the change of 2 metabolites during short-term recovery from alcohol dependence and compare the results using different random effect model