help > RE: Meanuas file missing
Aug 14, 2019  09:08 AM | Alfonso Nieto-Castanon - Boston University
RE: Meanuas file missing
Dear Anna,

Yes, typically functional realignment will also align multiple sessions to the first sessions (so your funcitional data is all properly aligned across multiple scans as well as across multiple runs/sessions), and it will also create a representative single "meanfunctional" file aggregating across all time-points and sessions (and named after the first functional in your first run/session). In your case, if you have multiple session-specific structural images and would like to do a direct coregistration between each functional run/session and each corresponding session-specific structural image (e.g. if you would like to use session-specific Freesurfer cortical tessellations for surface-based analyses), you could do that in either of the following two ways:

1) running realignment separately for each session. To do this, when you run the functional realignment step in CONN, simply uncheck the 'All Sessions' checkbox and select 'Sesssion1' and then click 'Start' to realign separately the first-session functional data (and then repeat selecting now 'Session2' only). Having run realignment separately for each session will produce a different representative "meanfunctional" file for each session, which CONN will then use for reference in the corresponding 'functional coregistration' step (note: there is no need to run also 'functional coregistration' separately for each session, CONN will handle those separately in any way given that you have entered a different structural image per session)

or 2) selecting "first functional volume as reference" during coregistration. If, instead of attempting to use the "meanfunctional" image as reference , you check the 'select first functional volume as reference' checkbox there, it will instead use the first functional scan/timepoint (separately for each session) as reference image for function-to-structural coregistration, so that should also work without a problem in the case of having session-specific structural volumes per subject

Hope this helps

Originally posted by Anna Vasilevskaya:
Dear experts, 

I am new to Conn processing, and ran into an error while preprocessing longitudinal fMRI data. I could not find the answer anywhere on the forum. I have 2 fMRI sessions and 2 anatomical scans for each participant. 

After running the "functional realignment & unwarp (subject motion estimation and correction)" preprocessing step, Conn created the meanuas file just for the session 1 scans, however not for the session 2 scans. Therefore, when I am trying to run the "functional coregistration to structural" step, which is next in my preprocessing pipeline, I get the following error message (copied below). It is looking for the meanuas file, however it wasn't created during the previous step. I also kept "first functional volume as reference' as unchecked. Would you happen to know what I should do? 

Thank you


Error using conn_setup_preproc (line 1618)
Error preparing files for coregistration. Mean functional file not found (err/Volumes/LaCie/Anna/CTE_longit_rsfMRI2/Functional/1034_2/meanuas110088B_04OCT2018-0005-00001-000121-01.img)
Error in conn (line 894)
Error in conn_menumanager (line 120)
CONN v.17.f
SPM12 + DEM FieldMap MEEGtools
Matlab v.2018a
storage: 969.6Gb available
spm @ /Applications/spm12
conn @ /Applications/conn"

Threaded View

Anna Vasilevskaya Aug 14, 2019
RE: Meanuas file missing
Alfonso Nieto-Castanon Aug 14, 2019