[VoxBo] Problem with setorigin -m

[Daniel Y Kimberg]

Junghoon Kim wrote:
[problem with origin mapping using setorigin]

setorigin -m uses the "AbsoluteCornerPosition" field to figure out how
two volumes are offset.  This field comes from your raw DICOM data,
and should have been set by the scanner to reflect absolute position
in the bore.  As long as you have these coordinates, setorigin -m
should work fine.

It looks like for your functionals, one of the values is "nan," which
means there's no sensible way for setorigin -m to work, at least in
that dimension.  I'm not sure if the problem was with your data files
or with how VoxBo reads them.  If you can point me to the raw data, I
should be able to figure it out.

Setting the origin by eye is probably good enough for many purposes,
but if you need really accurate coregistration, you may want to
coregister explicitly.  That might not be a bad idea in general, just
because subjects can move between runs, making the scanner coordinates
less useful.

dan