I know that the flipping thing is a bit messy. ECAT 7 files do not nead to
be
flipped at spatial normalisation, so sptl_Ornt should be set to
[0 0 0 0 0 0 1 1 1 0 0 0] for these files to spatially normalise properly.
The orientation of ECAT 7 files is decided on line 69 of spm_vol_ecat7.m:
dircos = diag([-1 -1 -1]);
which results in all axes being flipped. Once a .mat file is written for
one of these images, then this flipping gets incorporated.
All the best,
-John
| I am confused as to how SPM99 treats the orientation of CTI7 files. Our
PET
| files are acquired in radiological orientation (L is R) and upside down
| (Superior is -Z) relative to SPM/Analyze conventions. I have the spatial
| normalization defaults for the affine transformation set to
|
| sptl_Ornt = [0 0 0 0 0 0 -1 1 1 0 0 0]; % Radiological Convention (L is
R)
|
| I've noted two odd things. First the spatially normalized files have the
| correct SPM superior-inferior orientation (Superior = -Z) even though the
affine
| transformation does not specify a superior/inferior flip.
|
| Second, even though the spatially normalized images have left side = - X,
the
| left/right orientation is incorrect, i.e., the right side is labeled as
left.
|
| Does SPM99 somehow read the orientation of CTI7 PET files from the image
header
| and then automatically perform the proper flips to bring the spatially
| normalized images into SPM/Analyze/Radiological orientation ? Is this
specified
| when the *.mat files are written as part of the realignment routine ? If
so,
| should the default for the affine transformation of CTI7 files be set to
no
| flips, i.e.. Neurological Convention (R is R)?
|
| sptl_Ornt = [0 0 0 0 0 0 1 1 1 0 0 0]; % Neurological Convention (R is
R)
|
| Oops. I mistyped the superior-inferior orientations of my CTI7 files. The
| correct orientation is:
|
| CTI7 Inferior = +Z Superior = -Z, i.e. Slice 0 = top of brain
(superior
| surface)
| SPM Inferior = -Z Superior = + Z , i.e., Slice 0 = bottom of brain
| (inferior surface)
|