The images probably have differing
distortions. If you have no other way of undistorting the images, you could use
the config file posted by Jesper for nonlinearly registering a FA to a T1 a
while back and use it on both images. You could then apply any warp field to
your original datasets and undistort them. You would need to take into account
any linear rotations that occurred as a part of the registration process in the
gradient files.
Peace,
Matt.
From: FSL - FMRIB's
Software Library [mailto:[log in to unmask]] On
Behalf Of Markus Gschwind
Sent: Wednesday, March 04, 2009
3:36 AM
To: [log in to unmask]
Subject: [FSL] Two DTI series from
1 subject form different scanners
Hello Experts!
I have 2 DTI series from the same subject but one from a Siemes 3T platform,
and one from a Philips 3T platform.
They used the identical table of gradients but the image dimensions are
unfortunately different.
I am not satisfied with the output of FLIRTing the one dti_FA to the other
(12deg) and neither with FLIRTing each of them to the T1 of the subject
(12deg).
Is there another way how I could compare those images directly?
Does anyone has got an experience whith that problem?
Thank you in advance!
Markus