I would like to include rois I have made in the analysis.  I would like to remove from consideration the default ROIs.  WHen I remove BA.img form the roi folder, the conn program will not open.  When I remove it in the setup stage in ROI-data-set-up, I am given errors that my ROIs have not been defined.

(ERROR: Subject 2 ROI 2 file has not been defined )

when I go to select new ROIs in the selection window I am given the message "mismatched number of files"

Please advise about the proper way to consider only a select numebr of ROIs in this analyses.

Second, I have an unrelated question which is related to the fact that I do not have movement parameter files for all of my subjects.  I wonder if I should include them as covariates for the 50% or so of subjects I have these for, or not.

thank you,
Moriah

Hi Moriah,

You can remove some ROIs from consideration by going to the Setup->Rois tab, right-clicking on the desired ROI from the 'ROIs' list and selecting 'remove selected ROIs'. In order to add new ROIs, in the same Setup->Rois tab, select the first empty space in the 'ROIs' list, then enter a new name in the 'ROI name' field, and then (for subject-independent ROIs) select all of the subjects in the 'Subjects' list and then select an ROI file in the 'Select ROI definition files' section. For subject-specific ROIs you could: a) select sequentially one subject from the 'Subjects' list and one ROI file from the 'Select ROI definition files' section -and repeat these for all subjects-; or b) select all subjects from the 'Subjects' list and then select the same number of ROI files (one ROI file per subject) from the 'Select ROI definition files' section. The errors that you describe are possibly due to some subjects missing the associated ROI file (to double-check this, simply select each ROI in the 'ROIs' list, and then click sequentially on each subject in the 'Subjects' list to make sure all of them have an associated ROI file). Let me know if this helps. 

And regarding your second question, subject movement can have a dramatic confouding effect on connectivity measures (typically it introduces positive biases in the computation of connectivity values). I would recommend if possible to re-run the 'realignment' step on your functional volumes and enter the resulting realigned volumes as well as movement parameters into the analyses for all subjects (but I might be misinterpreting your question, I am not sure why you would have subject motion parameters for some subjects but not others; perhaps you used prospective motion correction in the scanner?)

Best regards
Alfonso

Originally posted by Moriah Thomason:

Dear Alfonso

I am glad to hear from you.  I was able to do these steps and they worked for me.  Trial and error can be a beautiful thing ;-)

In fact I also re-generated the movement (realignment parameter) files for all subjects and ran my analyses with those included (as you recommended) as well. To answer why some had them and others did not - the answer is that we are performing funcitonal connectivity in the fetal brain and fetuses do not take directions like "please stay still" to heart.  We need to extract the periods of (relative) stationarity from several sessions, and mask each length of time with a set-specific hand drawn mask, then manually transform into the right orientation.  So each segment (within a scan) gets preprocessed as its own session - thus we optimize for each length of time the fetus was still.  In this pilot analysis I selected to re-concatonate scans from within a session before submitting to your analysis.  Thus for some scans there were not complete realignment params. But, as I stated above, I went ahead and I ran realignment on all data in order to generate these files that would correspond (exactly) to the frames I submitted to your analysis.  So, all is well now & that explains why I had some work arounds.  Fetal imaging is a challenge & requires a bit of additional work and creativity.