Hi! I'm testing interoperability between AFNI and SPM5b via NIfTI
files, and
especially some new features I've added this week to improve
interoperability
between FSL and AFNI.
I've just been playing around with it until I come up with some formal
tests, but I've got one I can't explain right now: I have created from
a single
sample anatomical dataset 48 NIfTI datasets. These datasets contain exactly
the same data, except that it is stored in different orientations on
disk. The 48
permutations are every possible 90 degree square rotation for both
left-handed
and right-handed coordinate systems. AFNI definitely interprets all the
files as being
in the same spatial location and orientation; FSL also seems to (looking
at results from
the Brain Extraction Tool, which varies if FSL does not understand the
orientation
properly).
As an initial test, I loaded up two of these datasets into the Check
Registration
GUI for SPM. It will not display them in the same space.
I've started looking at the code to see if I can understand where AFNI
and SPM are
interpreting the location differently, but before I go to far, I thought
I'd ask the gurus
if you can suggest any reason why that is a Bad Test, or not a
meaningful test. Otherwise
I have to figure out how to reconcile these interpretations.
Also, I'm interested in hearing from any users who are using NIfTI for
software
interoperability--I'd like to hear what packages you're using, and what
success and
problems you've run into. Particularly if it involves AFNI.
I'll be glad to upload any or all of the test datasets I've generated if
somebody on the
SPM team needs more info.
thanks!
rich
--
\ Rich Hammett http://afni.nimh.nih.gov
/ [log in to unmask] sun dogs fire on the horizon
\ meteor rain stars across the night
/ this moment may be brief
\ but it can be so bright
|