Removal of confounding effects of global signal in functional MRI analyses

Local signals obtained from BOLD fMRI are generally confounded by global ef fects. In this paper, we make an essential distinction between global effec ts and the global signal. Global effects have a similar influence on local signals from a large proportion of cerebral voxels. They may reflect diffus e physiological processes or variations in scanner sensitivity and are diff icult to measure directly. Global effects are often estimated from the glob al signal, which is the spatial average of local signals from all cerebral voxels. If the global signal is strongly correlated with experimental manip ulations, meaningfully different results may be obtained whether or not glo bal effects are modeled (G. K. Aguirre ct al, 1998, NeuroImage, 8, 302-306) . In particular, if local BOLD signals make a significant contribution to t he global signal, analyses using ANCOVA or proportional scaling models may yield artifactual deactivations. In this paper, we present a modification t o the proportional sealing model that accounts for the contribution of loca l BOLD signals to the global signal. An event-related oddball stimulus para digm and a block design working memory task were used to illustrate the eff icacy of our model. (C) 2001 Academic Press.