[MINC-users] Very large dicom variables in the header crashing loni debabelizer.
Dylan David Wagner
minc-users@bic.mni.mcgill.ca
Wed Jul 20 20:24:05 2005
Hi,
I've been trying to use the loni debabelizer to convert some minc
files to nifti and/or analyze (couldn't find an in house converter for
nifti, do we have one?). I've been finding that certain files crash it,
other files aren't identified as MINC and a third group actually work!
The reason why some files crash the babelizer, or aren't
identified, seems to have to do with long headers caused by some of the
dicom variables in the minc header. At least that's the only difference
I can discern.
Older mincfiles scanned before the Siemens got an upgrade are
identified without fail. Aside from a few additional entries in the
study variable, the main differences is the dicom variables are huge and
the mincheader jumps from ~20kb to 700+ Kb.
Newer files scanned after the upgrade work if they've been
preprocessed with the fmri preprocessing tools (which seems to eliminate
the dicom entries from the header). Oddly enough if I use cygwin to do
the preprocessing with the same minctools, the dicom entries in the
header are preserved (and loni debabelizer crashes).
I'm guessing the fault lies with the Loni app, however does anyone
have any quick fixes to eliminate all the dicom info from the header?
I'd still like to keep the rest of the header intact.
Thanks,
DDW