Well, registering images that are of the same subject is much easier for a
registration algorithm than registering images of different subjects. Also,
the fMRIB template and the MNI template are not made from the same groups of
brains, so they do not look exactly the same. I would stick to the direct
approach if it gives you the best results (and in your next study remember
to acquire a field map to distortion correct your images so you won't have
to worry about this).
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Anastasia A Ford
Sent: Friday, January 09, 2009 11:19 AM
To: [log in to unmask]
Subject: Re: [FSL] T1 to b-zero registration for tractography
Matt,
The structures do align very well using the approach that Jesper has
suggested.
I also looked at the alignment using tbss scripts to align FA mpa to FMRIB
and using FNIRT to align my T1 to MNI 1mm. The resulting images o not have
their structures align, as the FA MNI image is much smaller than the T1. I
attached a screen shot of the two images.
Is there any reason why the direct registration using FNIRT works better
than using tbss approach?
thank you for your help.
Anastasia
|