you may relign the T1 to MNI template beforehand... I just read one
discussion online, hope it helps:
"I have seen results like this when the initial orientation of the
image is slightly oblique relative to the MNI alignment. One way to
fix the problem would be to adjust the AC-PC alignment of the image to
be closer to MNI orientation, manually by using the display-> Reorient
functionality. In searching for a more automatic way of avoiding this
problem in our legacy oblique scans, I have found that if I first use
the "Coregister: Estimate" function to register the subject image to
the (whole head) template, the resulting orientation is close enough
to the template (and priors) that the problem is almost always
eliminated. As John has pointed out on this list, the Unified
Segmentation routine in SPM5 is more sensitive to the initial
orientation than the old segmentation routine of SPM2 and earlier SPM
versions. My understanding is that in the older versions started off
by doing an affine registration of the image to the MNI template, but
in SPM5 the initial registration step is not used, and that is why the
initial orientation must be relatively close to the MNI alignment in
order to get good results. I hope this helps."
Good Luck!
Gao
On 19 May 2012 17:16, Ying Chu <[log in to unmask]> wrote:
> Note that these T1 images have not been slice timing, coregistered, and normalized (attached .doc).
> In Picture A, the cross-line is adjusted manually to the anterior commissure (AC). However, the Picture B is asymmetrical. I tried to adjust the pitch, roll, yaw on the panel and were not successful to realign the image. How can I adjust these images so that it will realign appropriately before slice timing, coregister, and normalize in SPM? Thank you.
|