[RESOLVED] ROI list for Resting-state global brain Connectivity (GBC)

Use Questions
1

To run Resting-state global brain Connectivity (GBC), one of the parameters used is:
–target=“<which_roi_to_use>” … Array of ROI codes that define target ROI
[default: FreeSurfer cortex codes]

What exactly does “Array of ROI codes that define target ROI” mean? That is, what is the specific format expected for the value of this parameter?

For example, if I am to use parcellated connectivity data that use the Glasser parcellation, what would be format for this parameter’s value that include the following parcels: R_V1_ROI
R_MST_ROI
R_V6_ROI
R_V2_ROI
R_V3_ROI
R_V4_ROI
R_V8_ROI

As a second example, if I wanted to include all parcels for the Glasser parcellation, how could that be done?

Thank you.

Estephan

2

Estephan, hi!

The target parameter is used to limit the computation of GBC in the case of dense input data, to a specific ROI or set of ROIs. To use the parameter, it has to be combined with an ROI names file (see Region of interest specification files) and the target parameter needs to list the integer codes for the ROIs that should be included in the GBC computation. If limiting dense volume or surface-based GBC to a subset of ROI is your interest, then I would advise you to use the fc_compute_gbc3 command directly. You can find examples of this use of “restricted GBC” in these papers:

Do note that the additional parameters rsmooth and rdilate (used for smoothing and dilating the resulting mask) are only implemented for volume based data.

If you are not interested in restricted GBC, then you can leave the parameter empty. If you are interested in parcellated restricted GBC, then you would need to prepare an ROI file for the parcellated data (in essence a plabel file) and a corresponding ROI specification file.

Let me know if you are moving in the direction of restricted GBC and would need further information.

All the best,

Grega

3

Hi Grega, I’m back into trying to run a GBC analysis using a CIFTI parcellated label file as the ROI file. Looking at the description for the ROI names file https://qunex.readthedocs.io/en/latest/wiki/Overview/file_names.html, it only mentions a volume mask file:

The first line in the names file provides a path to the group level mask file to be used for ROI definition. A mask file is a single volume image file in which each voxel is coded either with 0 or an integer code denoting a specific ROI.

So my questions are:

Thank you.

Estephan