Yes, there is a solution -- I found the posting from Geoffrey Tan
below when I had this problem, and now use the following recipe:
Display anatomic scan, place the cursor at anterior commissure, and
enter the negative of the mm values, NOT the voxel values, in boxes
below for x y z.
then press "reorient" and choose all the images acquired for the
session (i.e. the anatomic image volume, any fieldmap data, and all
the functional runs). Then
proceed with pre-processing.
Hope this helps!
> Dear Francesco,
> With the GUI, click on display. In the graphics
> window click where you want the origin to be. The 3 values that pop up when
> you click on a coordinate are the x, y, z of the point relative to the
> current origin. Change the x, y, and z in the boxes below to the negative of
> the values and click on re-orient image. Choose the same image and click
> done. Origin should then be fine.
> Best Wishes,
> Geoffrey Tan
On 7/6/09, Huang, Wei (Psychiatry) <[log in to unmask]> wrote:
> Dear SPM experts,
> I got a wrong segmentation after coregistering my high resolution T1 3D
> anatomy to the mean EPI. The coregistration turned out to be fine when
> I checked registration. But the segmentation went completely wrong
> after the coregistration. When I checked the c1 and c2, they hardly
> have any shape of the brain. Has anyone had this problem before? Could
> it be the orientation of my anatomy? Is there a way to fix this?