Dear Gabor,
I am afraid fnirt hasn't really been designed for inter-modality-inter-
subject registration. Through its intensity modeling it offers some
robustness against "small" differences in e.g. T1 weighting and also
against differences in bias-fields. It cannot presently do e.g. T2->T1
though.
If you have both a T1 and a T2 weighted MR for a given subject you can
always do
T2->flirt->T1->fnirt->MNI_T1
I guess for the future we might consider supplying templates in
standard space also for e.g. T2 and PD weighted images. I can see how
that might potentially be useful. It is not something I can promise
will happen though.
Good Luck Jesper
On 8 Sep 2010, at 12:38, Gabor Perlaki wrote:
> Thanks a lot Jesper. And What about a T2 weighted MR? Can I use
> fnirt to register a T2 image to MNI152_T1? I haven't found any
> recommandation for register different modalities with fnirt.
>
> Thanks a lot, Gábor
>
|