PRINCIPAL COMPONENT ANALYSIS AND THE SCALED SUBPROFILE MODEL COMPAREDTO INTERSUBJECT AVERAGING AND STATISTICAL PARAMETRIC MAPPING .1. FUNCTIONAL CONNECTIVITY OF THE HUMAN MOTOR SYSTEM STUDIED WITH [O-15]WATERPET

Using [O-15]water PET and a previously well studied motor activation t ask, repetitive finger-to-thumb opposition, we compared the spatial ac tivation patterns produced by (1) global normalization and intersubjec t averaging of paired-image subtractions, (2) the mean differences of ANCOVA-adjusted voxels in Statistical Parametric Mapping, (3) ANCOVA-a djusted voxels followed by principal component analysis (PCA), (4) ANC OVA-adjustment of mean image volumes (mean over subjects at each time point) followed by F-masking and PCA, and (5) PCA with Scaled Subprofi le Model pre- and postprocessing. All data analysis techniques identif ied large positive focal activations in the contralateral sensorimotor cortex and ipsilateral cerebellar cortex, with varying levels of acti vation in other parts of the motor system, e.g., supplementary motor a rea, thalamus, putamen; techniques 1-4 also produced extensive negativ e areas. The activation signal of interest constitutes a very small fr action of the total nonrandom signal in the original dataset, and the exact choice of data preprocessing steps together with a particular an alysis procedure have a significant impact on the identification and r elative levels of activated regions. The challenge for the future is t o identify those preprocessing algorithms and data analysis models tha t reproducibly optimize the identification and quantification of highe r-order sensorimotor and cognitive responses.