Ones for voxels that contribute, zeros for those that don't. However, you may
have voxels that you want to contribute by some intermediate amount.
All the best,
-John
On Friday 18 April 2008 16:23, Neggers, S.F.W. wrote:
> Dear all,
>
> because of some regions in a special EPI scan that do most likely not
> move like a rigid body, I would like to in- or exclude some parts of the
> scanned functional volumes not to take part in the realigment
> cost-function, eg to exclude thos voxels.
>
> To do that, I am playing around with realigment in spm5, and was
> wondering what exactly the 'weighting' field in the estimation options
> for realigment is doing. I would guess it can simply modulate the
> contribution of a voxel for the difference cost function that is
> minimized in realigment, but I am not sure from the help text in the
> SPM5 interface nor the manual. It states:
>
> ----
> Weighting
>
> The option of providing a weighting image to weight each voxel of the
> reference image differently
>
> when estimating the realignment parameters. The weights are proportional
> to the inverses of the
>
> standard deviations. For example, when there is a lot of extra-brain
> motion - e.g., during speech,
>
> or when there are serious artifacts in a particular region of the
> images.
>
> -------
>
> Am I right that when I provide an image here with 1s for voxels that can
> contribute to the cost function, and 0s when not, that I can realize
> exclusion of voxels entering the cost function? The 'inverses of the
> standard deviations' in the above make me little uncertain whether this
> is the case.
>
> Otherwise, I can of course simply mask out parts of the functional
> images in a script before realigning, but the above seems more elegant
> and less disk-space consuming.
>
> Many thanks for any pointers, have a nice weekend,
>
> Bas
>
> --------------------------------------------------
> Dr. S.F.W. Neggers
> Division of Brain Research
> Rudolf Magnus Institute for Neuroscience
> Utrecht University Medical Center
>
> Visiting : Heidelberglaan 100, 3584 CX Utrecht
> Room B.01.1.03
> Mail : Huispost B.01.206, P.O. Box 85500
> 3508 GA Utrecht, the Netherlands
> Tel : +31 (0)88 7559609
> Fax : +31 (0)88 7555443
> E-mail : [log in to unmask]
> Web : http://www.fmri.nl/people/bas.html
> --------------------------------------------------
|