Significant voxels were colored according to their level of significance and were overlaid on the averaged structural image. The averaged activation maps were intensity thresholded at P < 0.01, one-tailed, and each slice was subjected to a cluster analysis procedure to correct for multiple statistical comparisons by means of a spatial extent threshold that yielded a P < 0.01, one-tailed, significance level over the entire image. For the composite maps, the structural MRI scans were normalized into the same space to allow for the superimposition of statistical maps averaged across subjects onto an averaged structural image. Correlation coefficients were transformed into z scores. A Kendall's rank-order correlation was calculated for each voxel between event-related responses and subsequent memory classification of remembered, familiar, or forgotten. Subtracted values were integrated to measure each voxel's event-related response. The second statistical map, which revealed areas predicting subsequent memory, was created as follows: For each trial, the average value from baseline scans was subtracted from scans 10 through 15. The first statistical map, which revealed areas responding to picture presentation, compared signal from baseline scans to that of picture-classification scans for each voxel by means of a nonparametric Kolmogorov-Smirnov statistic. Scans 10 through 15 measured activation in response to picture classification.
Making memoires trial#
Scans 1 through 6 of each trial measured activation in response to baseline fixation.
0 kommentar(er)
