> I have 3D anatomical scans for a VBM study obtained with a Siemens Sonata
> scanner. I have used MRIcro for conversion from Dicom to Analyze format.
> The display option in SPM2 shows image dimensions of 160*256*256 and a
> -1,2*1*1 voxel size. Images appear as they are supposed to.
>
> Using the Dicom toolbox in SPM2 I get image dimensions of 256*256*160 and a
> -1*1*1,2 voxel size! Images look good as well and are in the correct
> position.
>
> When I open the Dicom toolbox converted images in MRIcro, and click on the
> sagittal view button I get the coronar images (flipped by 90 degrees). With
> the transversal button I get sagittal images and so forth.
>
> Which conversion is right? What can I do?
They are both right - except that the MRIcro converted files are probably used
more easily in MRIcro. You may want to check the handedness of the images
though, and make sure they are as you intend them to be:
http://www.fil.ion.ucl.ac.uk/spm/software/spm2/#Compatability
When SPM converts DICOM images, it retains the same slices as in the DICOM
files. If you have saggital slices, then the .mat file produced by the SPM
conversion will encode this orientation. Because MRIcro doesn't make use of
the .mat files that the SPM conversion produces, it will view them in a
different orientation.
When MRIcro converts the data, it appears to change the ordering of the voxels
in the .img file so that the slices appear transverse.
Best regards,
-John
|