SUCCESS!!
> Hi - these have different Z voxel dimensions - that probably explains
> the
> difference in FLIRT output as they are really quite different.
this is the difference!! wow. i simply do not understand why the
difference in the z voxel dimension would make such a difference such
that FLIRT would rotate the to-be-registered volume so severly. but
thank you, thank you, thank you!
the difference between the z voxel dimensions is that VolumeA's zvox,
4mm, is slice thickness + slice gap. the slice thickness is 3mm and
the slice gap is 1mm. VolumeB's zvox is 3mm -- slice thickness only;
slice gap is not added.
when i change VolumeB's zvox to 4mm (the same value as VolumeA), the
registration to the standard volume is successful, ie, the FEAT
Registration report shows VolumeB in the same orientation as the
standard!
steve, i must admit i was utterly skeptical of the idea that the zvoxel
size would make a difference. if you (or anyone) would not mind
further elucidation, i would be most grateful for any insight as to why
did the zvoxel dimension made a difference -- THE difference? could
the difference be because VolumeB seemed like only a "partial" brain
(ie, missing approximately 37mm [the volume is 37 slices]) and FLIRT
simply did the best it could?
thanks to all for your input and help.
- bettyann
|