My solution was switching back to conn17f.

205 functional volumes for each of 58 subjects. Using solely "removal of initial scan" (5 scans) in conn18a without other preprocessing steps resulted in a surge from 205 to 41000 dimensions (it should be 205-5=200..) among the volumes and thus a mismatch to the first-level realignment covariate.

Hi there,

I experienced the same issue. I have 240 functional volumes, and removing the initial 4 scans resulted in 56640 dimensions (appears to equal 236 x 240) for functional image of each subject.
Does anyone have suggestions on how to fix this in Conn18a, other than switching back to previous version? Thanks in advance!

Best,
Qinggang

Hi

Sorry about that and thanks for reporting this issue. The problem seems to be in the GUI display, which is mis-reporting the dimensionality of the resulting functional series after removing a number of initial scans. I am attaching a patch to fix this issue (this patch is for conn18a, simply copy the attached file to the conn distribution folder overwriting the file with the same name there). After applying the patch please run the following commands with your conn project open, and then save your conn project to keep these changes (this step is just to avoid having to re-select your functional data and applying the removal of the first four scans again) :

   f = conn_module('get','functionals');
   conn_module('set','functionals',f);

Best
Alfonso
Originally posted by Qinggang Yu:

Hi Alfonso,

Thank you for your help! I think it works perfectly now!

Best,
Qinggang