You are not adding any new information to your data by upsampling. (Well, one could argue that you are adding interpolation artifact, but that's another topic.) If whatever you are using for subcortical segmentation is advertised to work at your native resolution, my personal opinion is that you're better off not upsampling.
On Wednesday, September 12, 2012 11:16:02 you wrote:
> Thanks. A quick follow up though. If I do not unsample the data and leave
> it as is, couldn't this have an impact on volumetric and vertex analyses?
>
> Victor
>
> On Wed, Sep 12, 2012 at 11:10 AM, Benjamin Kay <[log in to unmask]> wrote:
>
> > Victor, upsampling your data will probably not help you do your
> > subcortical segmentation and will only make things take longer/require more
> > memory. If you really want to do it, check out the documentation for flirt.
> > You can use flirt with -applyisoxfm. or use -applyxfm with a -ref volume in
> > your desired resolution and coordinate space.
> >
> > On Wednesday, September 12, 2012 09:38:14 you wrote:
> > > Dear FSL experts,
> > >
> > > I have a data set that I would like to do subcortical segmentation on.
> > Before I get underway with the automatic segmentation, I was wondering if
> > it is possible to convert the image resolution in FSL from 2x2x2 mm to
> > 1x1x1 mm.
> > >
> > > I couldn't find much in the documentation regarding this. Any guidance
> > would be greatly appreciated!
> > >
> > > Thank you,
> > >
> > > Victor
> > >
> >
>
|