Dear Eric,
I would have assumed that some ICA components would correspond to
motion-related effects so that you don't have to discard volumes
beforehand. Removing volumes from a time series break serial correlation
assumptions.
Best regards,
Guillaume.
On 29/03/17 11:32, Eric HG wrote:
> Thanks to all for your reply!
>
> I was thinking to discard volumes with too much movement before
> performing ICA. Should I wait until the GLM and discard volumes then?
>
> Thanks a lot,
>
> Eric
>
> On Wed, Mar 29, 2017 at 12:13 PM, Guillaume Flandin <[log in to unmask]
> <mailto:[log in to unmask]>> wrote:
>
> Dear Eric,
>
> David's suggestion is the way to go if you were initially planning to
> discard volumes prior to a GLM.
>
> If you want to remove volumes from a NIfTI 4D file for other reasons (or
> just remove dummy scans), you can use the batch module:
> SPM > Util > 3D to 4D File Conversion
> where you select the volumes you wan to keep in the "3D Volumes" entry:
> 4dfile.nii,4
> 4dfile.nii,5
> 4dfile.nii,6
> ...
> See also:
> https://en.wikibooks.org/wiki/SPM/Working_with_4D_data
> <https://en.wikibooks.org/wiki/SPM/Working_with_4D_data>
>
> Best regards,
> Guillaume.
>
>
> On 29/03/17 09:46, Carmichael, David wrote:
> > Hi,
> >
> >
> >
> > Best way is in the GLM using regressor – one for each volume with a ‘1’
> > for the volume you want to excluded and ‘0’s otherwise – then you don’t
> > alter many other data features.
> >
> >
> >
> > David
> >
> >
> >
> >
> >
> >
> >
> > *From:*SPM (Statistical Parametric Mapping)
> [mailto:[log in to unmask] <mailto:[log in to unmask]>]
> > *On Behalf Of *Eric HG
> > *Sent:* 29 March 2017 00:11
> > *To:* [log in to unmask] <mailto:[log in to unmask]>
> > *Subject:* [SPM] Removing volumes from functional image
> >
> >
> >
> > Hi everyone,
> >
> >
> >
> > Does any of you have some code to remove specific volumes from
> > functional images? Like removing the following volumes: [1:3 43:58
> 121:142].
> >
> >
> >
> > Thanks a lot in advance,
> >
> >
> >
> > Eric
> >
>
> --
> Guillaume Flandin, PhD
> Wellcome Trust Centre for Neuroimaging
> University College London
> 12 Queen Square
> London WC1N 3BG
>
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|