Mt. Banich et al., fMRI studies of stroop tasks reveal unique roles of anterior and posteriorbrain systems in attentional selection, J COGN NEUR, 12(6), 2000, pp. 988-1000
The brain's attentional system identifies and selects information that is t
ask-relevant while ignoring information that is task-irrelevant. In two exp
eriments using functional magnetic resonance imaging, we examined the effec
ts of varying task-relevant information compared to task-irrelevant informa
tion. In the first: experiment, we compared patterns of activation as atten
tional demands were increased for two Stroop tasks that differed in the tas
k-relevant information, but not the task-irrelevant information: a color-wo
rd task and a spatial-word task. Distinct subdivisions of dorsolateral pref
rontal cortex and the precuneus became activated for each task, indicating
differential sensitivity of these regions to task-relevant, information (e.
g., spatial information vs, color). In the second experiment, we compared p
atterns of activation with increased attentional demands for two Stroop tas
ks that differed in task-irrelevant information, but not task-relevant info
rmation: a color-word task and color-object task. Little differentiation in
activation for dorsolateral prefrontal and precuneus regions was observed,
indicating a relative insensitivity of these regions to task-irrelevant in
formation However, Rie observed a differentiation in the pattern of activit
y for posterior regions. There were unique areas of activation in parietal
regions for the color-word task and in occipitotemporal regions for the col
or-object task. No increase in activation was observed in regions responsib
le for processing the perceptual attribute of color. The results of this se
cond experiment indicate that attentional selection in tasks such as the St
roop task, which contain multiple potential sources of relevant information
(e.g., the word vs, its ink color), acts more by modulating the processing
of task-irrelevant information than by modulating processing of task-relev
ant information.