ACTIVATION PATTERN REPRODUCIBILITY - MEASURING THE EFFECTS OF GROUP-SIZE AND DATA-ANALYSIS MODELS

The reproducibility of patterns from brain activation experiments has been examined only for suprathreshold spatially localized foci. Scatte r plots comparing signal levels across all pairs of Talairach voxels f or pairs of functional activation images provide an alternative approa ch for assessing reproducibility. Image-wide, signal-level reproducibi lity may be quantitatively summarized using pattern similarity measure s such as the Pearson product-moment correlation, rho. Empirical popul ation distributions of rho for many pair-wise image comparisons, gener ated using statistical resampling techniques, may be used to examine t he impact of a wide range of experimental variables. We demonstrate th e use of such empirical rho-histograms to measure changes in reproduci bility for [O-15]-water PET scans of a simple motor task as a function of group size and data analysis model. (C) 1997 Wiley-Liss, Inc.