Questions for Z-map in functional connectivity

Submitted by carloscat on Sat, 05/24/2014 - 15:43

 Hi all:

I'm starting using DPARSF and REST in my studies for resting state functional connectivity.
I've already finished the voxel-wise functional connectivity calculation and got files including ROIxx_FCMap.txt, ROIxx_FCMap.mat, ROIxxFCMAP.nii, and zROIxx_FCmap.nii.
I knew that files of FCMap.txt or FCmap.mat stored the information of each correlation coefficiency.
But since I would like to correlate if there is any correlation between the psychological profiles and mean Z score of whole brain in each subjects, what should I need to do next for extracting the mean Z score of whole brain?
I really appreciate your assistance and feedback.

Best Regards.

Submitted by Wei Liao on Sat, 05/24/2014 - 23:09

Hi,
I guess that you should extract the mean Z value of whole brain by 'Extract ROI signals' in REST Utilities button. In this case, you should use a whole brain mask as 'ROI', and the zROIxx_FCmap.nii as input data set. Ofc, you can pool all subjects' zROIxx_FCmap.nii in one folder to extract each subject whole brain mean Z value simultaneously.
Hoping it is helpful.
 

Submitted by carloscat on Wed, 05/28/2014 - 16:11

Hello~
Thanks a lot for your advices.
I've tried this method and it seemed that the mean Z score of whole brain (using SPM's default brain mask) could be successfully extracted.
But I've check these mean Z-scores, the results were very small.
No matter which functional connectivity (such as DMN, SMN, SN...) showed that whole brain mean Z score in each subject was below 0.1, some even below 0.001.
I knew that the zROIxx_FCmap is the result after Fisher's Z transformation of the original correlation coefficiency of the specific voxel time course.
But is this normal for such a small whole brain mean Z score?
Thanks & Best Regards.