> I have noticed that the orbits and contents often show up on some of my
> segmentations, particularly the WM and GM segments. Does anyone:
>
> 1. have the same issue
This will depend on the contrast of your images. If the orbits have a similar
intensity to grey matter, then they may be classed as grey matter. The
segmentation in SPM2 is fairly simplistic, as it is based only on image
intensity and the approximate positions within the images.
> 2. think it’s an issue
For VBM, significant differences in this part of the image could be due to
differences in the orbits.
> 3. bother to edit images prior to segmentation, and if so how do you
> accomplish this?
It mat be easier to do this afterwards using software such as MRIcro or the
image display routine in FSL. Make sure that your headers are OK afterwards
though. You may also be able to do something with the various skull
stripping algorithms that are out there. If you have a mask, then just use
ImCalc to multiply the mask with the grey matter (again making sure that the
headers are OK in the resulting mask images - use Check Reg).
Alternatively, you could try it in SPM5b, in which case the data should be
compatible with FSL (should you still need to do any editing).
Best regards,
-John
|