ROI-wise seed-based FC analysis in DPARSFA

Submitted by Mamtis13 on Fri, 11/01/2013 - 02:32

Dear forum members,

I set up an FC analysis with 2 ROIs to get zFC maps for correlations between two seed regions. In the FC folder (which gets generated in the Results folder) however I now have 72 files per subject, each apparently relating to one ROI. These files are labelled as follows: zROI1FCMAP_Subject1.nii, zROI1FCMAP_Subject2, zROI1FCMAP_Subject3, zROI1FCMAP_Subjectn, zROI2FCMAP_Subject1, zROI2FCMAP_Subject2, zROI2FCMAP_Subject3, zROI2FCMAP_Subjectn, ending with zROI72FCMAP_Subjectn ...

Would you know if something went wrong with my analysis or why I am getting so many ROI maps per subject? Also, which map do I now use for 2nd (group)  level analysis? Thank you very much in advance for your time, any advice or comments would be much aprpeciated!

Regards, Mamtis

Submitted by jigongjun on Fri, 11/01/2013 - 10:05

 1.  Would you know if something went wrong with my analysis or why I am getting so many ROI maps per subject?
What's the type of your ROI, spheres or predefined *.img? I suggest you to make sure that just two ROIs were added in DPARSFA.

2. Also, which map do I now use for 2nd (group)  level analysis?  
"zROI1FCMAP_Subject1.nii, zROI1FCMAP_Subject2, zROI1FCMAP_Subject3, ...... zROI1FCMAP_Subjectn" are FC maps of ROI1 for 2nd level analysis .

Submitted by Mamtis13 on Sun, 11/03/2013 - 20:13

Hello,

Thank you very much, one of the ROI masks was indeed not binary, thanks!

When I run the FC analysis with 2 seed regions (both are binary & anatomical, but one ROI is defined by an *.img file, the other one as a *.nii file), I am now getting zFCMaps for ROI1 and ROI2 (i.e. 2 per subject, e.g. zROI1FCMAP_Subject001, zROI2FCMAP_Subject1). My question: Do these results present the FC between region 1 nd 2, and again between region 2 and 1? I thought that the correlations would be the same between the two regions? Should it not be the same FC parameter for 1-2 as for 2-1? Why am I getting two zFC maps per subject?

Many thanks!
Mamtis

Submitted by ZHANG_RESTadmin on Sun, 11/03/2013 - 21:10

no, they are not FC between region 1 and 2, neither between 2 and 1.
They are correlation between region 1 and other voxels in the whole brain (zROI1FCMAP), and correlation between region 2 and other voxels in the whole brain (zROI2FCMAP).

To get the ROI-wise FC, you should use REST --> ROI wise FC, not voxel-wise FC.

Submitted by Mamtis13 on Sun, 11/03/2013 - 22:22

Hello,

I just watched Part1 of your multimedia course (Data_Process_of_Rest_fMRI_Part1_English.mp4) and found the answer to my question :-) Thanks again for your help as well as for providing the materials, everything has been really helpful!

Cheers,
Mamtis