help > RE: How CONN set up ROI as default?
Aug 14, 2019  04:08 AM | Pravesh Parekh - National Institute of Mental Health and Neurosciences
RE: How CONN set up ROI as default?
Hello,

You can visualize how the ROIs are defined by taking a look at the NIfTI files in the conn/rois folder (have a look at atlas.nii and networks.nii); the numbers in the file refer to the line number in the lookup table (for example, voxels with intensity 1 in atlas.nii correspond to frontal pole right (see atlas.txt); networks.nii is a 4D file where the volume number corresponds to the entries in networks.txt. You can also look at the description images of both of these to get an extent of how the regions are defined.

Conn uses the ROIs as is defined in the Setup. For example, if you have used "atlas.nii", then the entire region will be used for analysis. If you original ROI is a 5mm sphere centered somewhere, then Conn will use the ROI as such.

Hope this helps

Best
Pravesh


Originally posted by Mitch H:
Hi, 

I just came across this question too, and I think I have figured it out by looking in the REX gui. If you open up REX in the seed-voxel analysis results window, you can pull the correlation (or other measure) results for each subject and ROI by importing the Harvard-Oxford atlas in the "ROIs" selection.

So for example, I change the ROIs from the clusters found (from your first level analyses) to the Harvard-Oxford atlas (from the /conn/rois path). Then I change the measure I'm interested in to say, positive-voxels count. Then when I hit "Extract", I see the highlight ROI used in the bottom right corner of the REX display. As you can see from the attached REX display, the Frontal Pole R is not just a 5mm sphere but rather the whole atlas defined region. You can do the same for any of the correlation measurements too. Hope this helps. 

Best,
Mitch

Threaded View

TitleAuthorDate
Hyunwoo Kim Jun 27, 2018
Mitch H Aug 12, 2019
RE: How CONN set up ROI as default?
Pravesh Parekh Aug 14, 2019
max345 Apr 15, 2022
Jordon Tng May 28, 2019