Hello all, our research group is running into an odd error when we try to import HCP resting data into CONN.
We have tried this on various computers running Apple OS, Windows, and Linux and the error is always the same.
We have also tried importing multiple test files from the HCP database and the error is always the same.
Finally, we tried importing test files into a previous version of CONN (17). and the error is the same.
A screenshot of the generated error screen is attached.
After looking through the mailing list a bit it seemed that the file might be corrupted, but that is unlikely to be the case here as we downloaded multiple test files, on multiple machines. We verified in each case that the rest files open fine in mricron / mricrogl, so the files are apparently not corrupted.
Any insight would be appreciated!
Thanks
Eric Rawls
Postdoctoral Fellow
University of Minnesota Medical School
Psychiatry and Behavioral Sciences
rawls017@umn.edu

To add to the issue, task-fMRI from the HCP dataset loads with no errors (tested using the t-fMRI from the emotion task). The error seems to be restricted to the resting data.
E

Bump. Are we the only research group that has encountered this systematic issue reading Human Connectome Project data into CONN?
Cheers
E

 Hi Eric,

Sorry for the late reply. That is an odd error indeed (I have used HCP data wiithout a problem myself so I am pretty sure this should not a problem wth the data  ifself). The error message seems to indicate that the file ...MNINonLinear/Results/rfMRI_REST1_LR/rfMRI_REST1_LR.nii is not found or corrupted. Could you please double-check whether the file is in fact in that location and loadable from SPM? One possibility could be that the issue arises from not having write permissions into that folder. HCP data is originally in .nii.gz format and CONN/SPM will unzip that into uncompressed .nii format; this step may fail if the user does not have write permissions into that same folder. If that is the case here then you may have a rfMRI_REST1_LR.nii.gz file there but not a rfMRI_REST1_LR.nii file (and the solution would be simply to either gain write-permissions into those folders and then re-load the original .nii.gz files into CONN, or import the files into CONN using the "copy to BIDS folder and import" option so that CONN works with its own copy of the data instead of the one in the read-only folders).   

Hope this helps
Alfonso
 
Originally posted by Eric Rawls: