Dear listmembers,
Following Manuel Guirao's question (May 8th), we were wondering if someone
has a routine to convert images in Analyze-format back to ECAT-7. We have
been using SPM's normalization function on our ECAT-7 PET files,
which have been converted to Analyze format, so SPM can handle them.
We use one sum file from which we determine normalization parameters,
and apply these parameters to write normalized an entire set of for instance
39 files
(one dynamic FDG-PET file, converted to Analyze).
After this we convert the Analyze files back to ECAT-7 format, to be read in
the CAPP software.
Our question is: How do we produce normalized nfilename.img files that are
exactly equivalent in size to the original filename.img files (2 MB)?
Up until now, the normalized nfilename.img files did not approach the size
of the original files,
being either 16 or 8 times to small, or just 1/4 to large. We've tried using
the voxel size and bounding box dimensions
of the original, and this seems to work better (1/4 to large), but the files
are still not readable in ECAT-7 format.
We therefore assume that they have to be exactly equal in size, because
un-normalized files can be reconverted without difficulty.
Does someone have a solution to this problem?
Tomas Knapen
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|