Brain fMRI Reactivity to Smoking-related Images Before and During Extended Smoking Abstinence (vol 17, pg 365, 2009)

The list of brain regions exhibiting more fMRI activity during extended smoking abstinence than during the pre-quit assessment should read: Greater fMRI activity was detected during extended smoking abstinence than during the pre-quit assessment subcortically in the caudate nucleus and cortically in prefrontal (BA 6, 8, 9, 10, 44, 46), primary somatosensory (BA 1,2,3), temporal (BA 22), parietal (BA 5, 7, 40), occipital (BA 17, 18), and posterior cingulate (BA 31) cortex. Methods: Monte Carlo analysis (page 367) now should state: For individual Scan 1 versus 2 analyses, 20 3x3x3mm resampled voxels (similar to 540mm(3)) are needed to correct an individual voxel type one error of p=.01 to a corrected threshold of p<.05. Due to a reduction in power when detecting an interaction effect, a slightly lower threshold was set when comparing Scan 1 versus 2 brain reactivity to smoking versus neutral images. Assuming an individual voxel Type I error of p=.05, a cluster size of 26 voxels (similar to 702mm(3)) corrects to a threshold of p<.05. Results: Imaging section (pages 367-368): for scans 1 and 2, all clusters similar to 540mm(3) are cluster corrected to p<.05. When comparing scans 1 and 2, all highlighted regions except the anterior cingulate (ACC) reached statistical significance at a p<.05 (corrected). The ACC activation did reach statistical significance at a lower statistical threshold (uncorrected p<.01). Re-thresholded data are presented in the revised Table 3 below.