Hi,

I’m importing binarized ROI masks derived from the HCPex atlas (in MNI space) into CONN, focusing on visual and subcortical regions. I’ve read the documentation about ROI naming conventions and proper labeling.

To ensure correct alignment, I generated the masks directly from the HCPex.nii.gz file in MNI space and verified that they are properly coregistered with the functional and structural data in CONN (also in MNI space), using the “ROI tools > Check registration” feature.

However, when I process each network separately, CONN ends up grouping the ROIs into clusters like namemask_cluster00, namemask_cluster01, etc. I don’t understand on what basis this clustering is performed, and it interferes with my individual ROI-level analysis.

Is there a way to avoid this automatic grouping or to control it explicitly? I’ve also tried including a .txt file with the ROI names, but it didn’t solve the issue.

Thanks in advance for your help!

Best,

Guillem Moliner Michavila