The orientation of the raw images should be entered into SPM
either via the <Defaults> button (spatial normalisation,
Defaults for Parameter Estimation, Affine Starting Estimates?),
or by modifying the values for sptl_Ornt in spm_defaults.m.
Wheather or not the data is flipped at the spatial normalisation
stage depends on the values of these parameters.
[0 0 0 0 0 0 1 1 1] - no flipping.
[0 0 0 0 0 0 -1 1 1] - flip in x direction
[0 0 0 0 0 0 1 -1 1] - flip in y direction
[0 0 0 0 0 0 1 1 -1] - flip in z direction
If the orientation is enteded into spm_defaults.m, then it is a
permanent change. Otherwise, the orientation is only the default
for the remainder of the SPM session. The PostScript printout
produced by the spatial normalisation will tell you if the images
have been flipped or not.
Best regards,
-John
| I am having a problem in that I ran an analysis on a set of
| PET images.
| Data Flow as follows:
| 1) realignment of ECAT images (AIR) (radiologic)
| 2) conversion of ECAT to Analyze (still radiologic)
| 3) Normalize (neurologic)
| 4) smooth (neurologic)
| 5) stats....
|
| I got good results, and sent them to another group. They are
| trying to replicate this, but skipped step 2 opting to just
| have SPM read in the ECAT images. Their results are flipped
| laterally.
|
| my suspicion is that spm99 automatically flips ecat images
| when it reads them in, and thus a double flip has occurred.
|
| I have checked my images and I believe that they are in
| the correct orientation (one scan has bilateral thalamus
| flow allowing my to track orientation)
|
| So, does anyone know if spm99 automatically flips ecat7
| images when it reads them in? Any help will be accepted.
|