I just helped debug a problem a colleague was having when trying to use
marsbar to extract betas from an ROI.
She was running into the problem where get_marsy exited with an error
indicating that it was unable to load image.
After stepping through the script, I determined that the image in
question was the first of 200 functional volumes.
get_marsy was using the information within SPM.mat to figure out what
functional image files needed to be loaded.
The problem was that, between the period that the SPM.mat file was
created and when marsbar was being used, these images were changed. In
our case, we had been using collapse_nii_scan to turn n 3D nifti files
into a single 4D nifti file, and then deleting the 3D files (which is
fine), but we did this AFTER the SPM.mat file was created (which is NOT
This problem could just as easily arise from moving or zipping up the
files, however. For example, if a colleague ran a first level analysis
in his/her home directory, and you copied the entire project directory
to your own home directory, your file locations will not match those
indicated in the SPM.mat files that your colleague generated.
Hope this helps avert some 1AM frustration.
Developmental Cognitive Neuroscience Laboratory
Northwestern University, Evanston, IL
DCNL Lab: http://tinyurl.com/9q35abd
Frontiers Profile: http://tinyurl.com/aut488k