Hi John,
> I've been reading through the DARTEL manual and the VBM manual and it
> states that it's good practice to realign images to MNI space before
> beginning the VBM processing.
My feeling is that, especially with new segment, this is a good idea as
sometimes the affine registration does not like heavily "disoriented"
brains (you know what I mean :)
> It states that one can achieve this by
> using the display feature in SPM and entering appropriate values. As
> I have quite a few brains to reorient, rather than apply this
> approximate correction to all subsequent images, or do them all
> one-by-one, is it possible to use the SPM coregitser 'estimate and
> reslice' option instead (target is the MNI T1152 template)?
Yes: we used to be doing this manually until somebody (probably John)
mentioned that one could simply do a coregistration of the anatomical
(source) to an appropriate template image (target). You could use a T1
for this - as the source image will not be enlarged but only juggled
around to better match the template's orientation, this is much easier
and in most cases quite appropriate.
> Are there
> any downsides to this ie would the subsequent VBM pipeline results be
> affected by the intepolation or reslicing of the data ? Or is such a
> step just slower and unnecessary?
You do not need to reslice, only estimate (this avoids your justified
worry about interpolation: as the updated orientation can be stored in
the header, there is no need to reslice). For larger populations, I
think this beats manual reorientation anytime :)
Best,
Marko
--
=====================================================================
Marko Wilke (Dr.med./M.D.)
[log in to unmask]
Universitäts-Kinderklinik University Children's Hospital
Abt. III (Neuropädiatrie) Dept. III (Pediatric neurology)
Hoppe-Seyler-Str. 1, D - 72076 Tübingen
Tel.: (+49) 07071 29-83416 Fax: (+49) 07071 29-5473
=====================================================================
|