I am seeking advice regarding the well-known flipping problem associated with
running results in SPM.
I currently have two datasets, one dataset aquired in neurologic convention and the
other acquired in radiologic convention. Unique sets of spm_defaults were properly
set-up prior to reorientation for each group. Upon running stats by script (same script
for both, separate defaults), no problem in flipping was detected and running t-tests
produced no errors for within or between group comparisons. All con files upon
display revealed the dimensions set in the script: 3 x 3 x 3. However, the creation of
additional contrasts beyond those produced by script yielded a flip in x to –x for
some of the subjects, resulting in the dimensions: -3 x 3 x 3. Changing the
spm_defaults and re-running the results does not alleviate this problem. I would
greatly appreciate advice regarding this perplexing issue. The error that results
during the mapping of files is:
Images do not all have same orientation and voxel size
??? Error while evaluating uicontrol Callback.
Kristina DePeau
Sperling laboratory
Boston, MA
|