I think it could be tricky to achieve an accurate automatic registration
when the FOV is so small, and the dropout in the fMRI is so marked. One
thing you could try though, would be to skull-strip and bias correct the
anatomical scan. You may be able to do this via the Segment button (to
generate bias corrected, GM, WM and CSF images), followed by ImCalc to
mask out regions that are not GM, WM or CSF.
Best regards,
-John
On Fri, 2009-10-16 at 16:55 +0100, Cristián Modroño wrote:
> Dear all,
>
>
> I'm trying to coregister anatomical inplane slices
> with functional images (8 images per volume in both cases).
>
> In some of the series, as you can see in the attached .jpg, the result
> is not good
> (nevertheless with some subjects it works).
>
> I have modified the coregistration parameters unsuccessfully.
>
> Can anybody tell me whether this can be done and how? I have tried
> both SPM5 and SPM8.
>
> Thanks a lot,
>
>
> Cristián.
>
--
John Ashburner <[log in to unmask]>
|