On 11 Jun 2018, at 23:39, Camille Gasser <[log in to unmask]> wrote:

Hello,

I'm trying to run a very basic model to find regions active during learning trials compared to baseline. I tried setting the first-level model up by using a single regressor that contained the (jittered) onset of each trial, and includes motion-correction regressors as well. The single-run models ran with no problems, but when trying to run any kind of higher-level model (i.e., trying to get trial-related activity within-subject, across-runs), FSL won't allow me to load the lower-level feat directories into the GUI. I can't find any clear details in the resulting error message:

File "/usr/local/fsl/bin/imglob", line 78
   print "Usage: $0 [-extension/extensions] <list of names>"
                                                           ^
SyntaxError: Missing parentheses in call to 'print'. Did you mean print(int "Usage: $0 [-extension/extensions] <list of names>")?
 File "/usr/local/fsl/bin/imglob", line 78
   print "Usage: $0 [-extension/extensions] <list of names>"
                                                           ^
SyntaxError: Missing parentheses in call to 'print'. Did you mean print(int "Usage: $0 [-extension/extensions] <list of names>")?
   while executing
"exec sh -c "${FSLDIR}/bin/imglob $args" "
   (procedure "imglob" line 7)
   invoked from within
"imglob $feat_files(1)/stats/cope*.* "
   (procedure "feat5:updateselect" line 63)
   invoked from within
"feat5:updateselect $w"
   (procedure "feat5:multiple_check" line 74)
   invoked from within
"feat5:multiple_check .r 0 1 1 d"
   invoked from within
".dialog2.cancel invoke"
   ("uplevel" body line 1)
   invoked from within
"uplevel #0 [list $w invoke]"
   (procedure "tk::ButtonUp" line 22)
   invoked from within
"tk::ButtonUp .dialog2.cancel"
   (command bound to event)

--

I also tried running a model with the cope images from each single-run model as inputs instead of the lower level feat directories, and while this model at least started, it ultimately crashes. All I get is that same error message (above) in the FSL report log, and the text "Errors occured during the analysis" in big red font.

My best guess is that because there's only a single regressor with an onset every 8-11s, the model is somehow underspecified, since I'm pretty sure my approach would work if I had multiple trial conditions and could model those as separate regressors. My next thought is I'll need to model components of the trial separately, adding regressors for fixation screens, response screens, etc. Could anyone offer some more insight on this issue? Thanks!

########################################################################

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1