Abstract
Visual working memory (VWM) is known to be a limited capacity system and recent neuroimaging studies have shown that the activations of the frontoparietal network which includes the superior frontal sulcus (SFS) and the intraparietal sulcus (IPS) are sensitive to VWM load (Linden, et al., 2003), although there are cases where subsets of this network are not sensitive under different conditions (Todd&Marois, 2004). In particular, it remains unclear in what conditions SFS is recruited into the network. Here, we focus on the effects of feature selection using multidimensional objects and attempt to dissociate these brain areas, with the use of functional magnetic resonance imaging (fMRI).
We conducted a single-probe test of change detection task where sample arrays were presented simultaneously at different locations. The items were defined by colored dynamic random dots, which possess three features; color, shape and motion direction. In feature-selection condition, the stimulus arrays always varied in each feature dimension and the subjects were required to memorize only one relevant feature, but ignore the others. On the other hand, in no-feature-selection condition, irrelevant feature dimensions were held constant; “white” for color, “square” for shape and “down” for direction.
As a result of the fMRI data analysis, the activation of IPS was sensitive to VWM load in both conditions. In contrast, SFS was load sensitive only in the feature-selection condition. These results suggest that the SFS plays an important role in feature selective maintenance, whereas the IPS may be associated to simple maintenance.
M.W. is supported by Grant-in-Aid for Scientific Research on Priority Areas -Higher-Order Brain Functions- from The Ministry of Education, Culture, Sports, Science and Technology (17022015)