Hi all,
I need some help for finding the right spatial correspondence between my functional and anatomical datasets. I inherited these data from a previous study and consequently I cannot modify the acquisition parameters (in particular the slices orientation). I know this is an old and already covered topic on the message board, but more information are needed.
The data were acquired on a GE Signa 1.5 T. I generated AFNI datasets for structural and functional images using to3d: it seems that the program automatically and correctly picks up info like TR, FOV, voxel size, xyz origin as they are also stored in the header of the image files.

My problem is the following: I open two AFNI controller windows, one with the anatomical dataset and the other with the functional dataset as underlays: I open an axial view of both and I lock the two windows. The slice shown by the ?anatomical window? is at a different level compared to the slice shown by the ?functional window?. I suppose this is a problem of z-origin setting. Functional and anatomical slices were acquired according to an oblique plane with different angles. Should I correct for this? If yes, what should I do? The following 4 lines contain some information coming from the header of my image files that can be useful:
ANATOMICAL: 0020,0037, Image Orientation (Patient)=0.999959\-0\0.00909443\0.00134279\0.989179\-0.146704
FUNCTIONAL: 0020,0037, Image Orientation (Patient)=0.999958\0.00027466\0.0289928\-0.000427253\0.999984\0.00555445

Functional images have zoff = -17.6415, while anatomical images have zoff= -24.8405. I applied
3drefit -dzorig -7.2 anat+orig
(-7.2=-24.8405+17.6415)
but things did not improve. By manual inspection, I guessed that a reasonable offset between the two types of images was about -27mm. Therefore, I applied 3drefit in the same way and I found better, even if not optimal, results (actually with ?better? I mean ?reasonable? according to the results I expect).
I know that this is a ? roughly empirical? approach and I am searching for a ?more scientific? one. Which kind of z-origin correction should I apply?

I read information from the header of my image files using the free software MRIcro, with the command ?Import --> open foreign?. In fact I tried to apply the programs 3dinfo, file_tool and ge_header, but several error messages appeared. I list such error messages below:

3dinfo -help
nVidia Corporation GeForce4 MX440SE (10de@0181):4:nvidia:16:::::

3dinfo volreg1+orig
nVidia Corporation GeForce4 MX440SE (10de@0181):4:nvidia:16:::::

file_tool -ge_all -ge_all -infiles i126639.MRDC.0674
i126639.MRDC.0674 : GE header failure : -2

ge_header i126639.MRDC.0674
-----------------------------------
File i126639.MRDC.0674 is not a GE image file [doesn't start with IMGF]

What do these error messages mean?


I think that each of the problems mentioned above is related to others. I wrote a long mail to better detail my difficulties. I hope I have not bored you.

Thanks for your patience while answering my questions.

Silvia