Submitted by Rafael on Mon, 04/18/2011 - 17:02
Dear REST pro's
Tipically ALFF analysis is done on mALFF images, i.e. at each time point a voxel's ALFF is divided by the global ALFF (mean ALFF of all voxels). This normalization procedure might reduce intersunject variability in my opinion. Nevertheless it was recently used in a study comparing two groups (Zang et al. 2007 Brain & Development) in which interindividual variability is the effect of interest by definition, as I understand.
Why would it make sense to use mALFF images rather than ALFF images in such studies?
Thanks for any comments
Submitted by YAN Chao-Gan on Wed, 04/20/2011 - 01:17 Permalink
Re:
Hi!
The mALFF value in a given voxel reflects the degree of its raw ALFF value relative to the average ALFF value of the whole brain. If the global ALFF is a confounding effect across participants, then mALFF would be very important. Fox example, if most of the voxels showed reduced ALFF in patients, then mALFF could tell us where is still relative normal while where is relative abnormal.
May Professor Zang provide more comments.