Segregation of cognitive and emotional function in the prefrontal cortex: a stereotactic meta-analysis.

Imaging studies of major depressive disorder and schizophrenia strongly implicate the prefrontal cortex. Interpretation of such studies is hindered by the limited knowledge of normal functional segregation. Different anatomical regions may be functionally specialised. Elucidating such specialisation may assist design and interpretation of patient studies. In this meta-analysis, 330 emotion induction and cognitive task studies of normal subjects published over the past decade reporting prefrontal activation have been examined. It was hypothesised that emotion induction would result in inferior medial activation and cognitive tasks dorsolateral activation. A significant difference in the pattern of reported activations was found in keeping with this hypothesis. Estimates of most likely reported activation loci for emotion induction and cognitive task studies have been made. In Montreal Neurological Institute (MNI) stereotactic space, these comprise of +/-5, 46, 18 and +/-5, 28, 31 for the medial prefrontal cortex, and +/-42, 28, -16 and +/-54, 28, 18 for the lateral prefrontal cortex, respectively. Additionally, estimates of the boundaries between emotional and cognitive-processing regions have been made. We restricted the effects of various potential sources of bias on the above estimates by attempting to include all relevant studies and independent selection by both authors of at most two activation loci from each study according to prespecified criteria. Such estimates of most likely reported activation loci may allow improved planning, analysis, and interpretation of imaging studies of psychiatric disorder and of normal function.