Dear All
(dear John, I know, it is annoying. Feel free to delete, I am thinking
out loud). I have tried to look into this whole flipping issue on the
background that I looked into three programs for file conversion into
spm-usable files. My results were confusing, so I would like to share.
As a pretex, default.analyze.flip is set to 0 locally (since
I-don't-know-when).
The data is epi from a Siemens scanner (Sonata), and is routinely, from
the console, transferred to a DICOM server, where they are stored. The
3d is in register with the functional (if both are converted using the
dicom toolbox).
If I convert the files using MRIcro (conversion wizard OR convert
foreign to analyze, with NO flipping set in the options), the
functionals "look the other way" with regard to those converted using
the toolbox (sorry for attaching figures, I tried to make them small;
thank god for asymetrical brains and imperfect subject placing. Images
were displayed in MRIcro and saved as pictures). So at least there is no
correspondance here between the files converted using the DICOM-toolbox
and those converted using MRIcro (correspondance is also absent when
displaying them in spm2, presumably due to flip=0 ?).
Finally, files converted using MRIconvert that does *not* flip images in
lr-direction (according to the author) agree with the DICOM-results. So
it's 2:1 against MRIcro - now I'm confused. Interestingly enough, spm2
bails out and complains about the ones converted using mriconvert ("Not
enough volumes in ..."), while spm99, as MRIcro, reads them all right.
BTW, the 3D (if converted with MRIcro) is not in the correct orientation
(neither in SPM nor in MRIcro, where the coronal looks sagital and the
coronal axial) and thus not in register with either functional. So you
could say it is 3:1 and argues against using MRIcro (for this matter only!).
It is just that I am confused that two conversion routines outside of
spm2 do not agree, with both claiming not do to any flipping. The one
that does agree with the spm2-solution does not produce images that are
readable in the new version of spm. Bottomline, I will be happy to stick
to one conversion routine once I know what's going on...
> NEVER MESS WITH THE FLIP DEFAULT. IT SHOULD BE SET ONCE FOR YOUR
> LAB, AND KEPT AT THAT VALUE.
I agree. I did not and do not intend to change it. I will be a happy
camper just for being sure about the orientation in my small world here.
Best, and sorry if I make your day miserable,
Marko
--
=====================================================================
Marko Wilke (Dr.med./M.D.)
[log in to unmask]
Universitäts-Kinderklinik University Children's Hospital
Abt. III (Neuropädiatrie) Dept. III (Pediatric neurology)
Hoppe-Seyler-Str. 1, D - 72076 Tübingen
Tel.: (+49) 07071 29-83416 Fax: (+49) 07071 29-5473
=====================================================================
|