We found no significant activation in the hippocampus for maintaining constructed scenes in working memory above and beyond that for maintaining objects, and peak activity was significantly less for maintaining constructed scenes than for constructing scenes.
In anterior hippocampus, the response was significantly greater for constructing scenes than maintaining constructed scenes (t(24) = 5.05, P = 3.69e − 5).
There was no significant difference between maintaining perceived and maintaining constructed scenes in anterior (t(24) = 0.36, P = 0.72) or posterior (t(24) = 0.30, P = 0.77) hippocampus.
Reflecting the SPM result that scene construction was limited to anterior hippocampus, there was no significant difference between constructing and maintaining constructed scenes in posterior hippocampus (Fig. 4, bottom; t(24) = 1.00, P = 0.33).
Each model had PPI regressors for perceiving, constructing, maintaining perceived and maintaining constructed scenes, plus each of their object baselines (a PPI regressor being the time series of the ROI multiplied by the ON times of the conditions, which was then convolved with the canonical hemodynamic response function).
In order to maintain construct stability (see Brown 2006), we retained at least three items for each subscale.
This method of treatment was selected to evaluate long-term bioactivity of released IL-1ra since a direct co-culture of constructs and IL-1ra microspheres would require media changes every 3 days to maintain construct viability.
Further, it became obvious that for reasonable expression levels, freshly transformed cells, mid-log precultures, and high levels of ampicillin (200 μg/ml) were essential to maintain construct stability.
To identify voxels involved with scene maintenance, we performed the contrast of maintain constructed scenes – maintain constructed objects.
A contrast was calculated of the scene conditions relative to the object baselines ([perceive scenes + construct scenes + maintain perceived scenes + maintain constructed scenes] − [perceive objects + construct objects + maintain perceived objects + maintain constructed objects]), which revealed activation of a priori scene regions in addition to a large area of visual cortex (Fig. 2, Table 1).
