I notice that you are using MRIcro, rather than the NIfTI compatible
MRIcron. I would suggest trying Check Reg in SPM to see how well
aligned your images are to the tissue probability maps of SPM. If the
alignment is within about 3 cm and 15 degrees, then the segmentation is
much more likely to work.
Best regards,
-John
On Thu, 2009-11-12 at 18:44 -0500, fengshi wrote:
> Dear SPMers,
>
> I'm using the segmentation in SPM8 in a group of adult T1 images
> (shown in left panel of attached figure). Most of them have good
> segmentation results. But some of them always have twisted
> results even rerun them many times (shown in right panel of attached
> figure).
>
> I tried both the "segmentation" button and also the "New segment" in
> tools in SPM8, both of them do not work on these failed subjects.
>
> When using the SPM8, I just set the data as my data and use the
> default parameters.
>
> Please share some clues if you ever encountered the same problem.
>
> One of the adult T1 image is uploaded in blow link.
>
> http://www.med.unc.edu/uploads/withi.subjectt1t.gz
>
> Please note that this file will be purged in a week.
>
> Thanks!
>
> Best,
>
>
> Feng Shi
> ______________________________________________________________________
> 2009-11-11
>
--
John Ashburner <[log in to unmask]>
|