Hi,
In general FNIRT warps contain both nonlinear and linear components, and subsequent resampling calls to applywarp usually only require the warp image and not extra affine matrices. (Note: one exception to this is when you are combining it with a different affine transformation to do something like go from space A to B to C, where B to C is the registration done with FNIRT and A to B is a separate registration, like is done in fMRI to go from example_func to highres to standard).
The dimension of the warp field in the "coef" convention is smaller because it is not an image but a collection of b-spline coefficients. If you want to see the warp as an image (we call this the "field" convention) then you need to either ask for this output directly, or create it from the coef file using fnirtfileutils.
All the best,
Mark
On 21 Mar 2013, at 13:36, xuemei <[log in to unmask]> wrote:
> Hi,
> After performing 'tbss_3_postreg', we can get a file named *_to_*_warp.nii.gz. In FSL website, I know that it is deformation field (aslo called a warpfiled) that produced using FNIRT. And it is a 4D image - three images combined together in the "fourth dimension". These correspond to x-displacements (volume 0), y-displacements (volume 1) and z-displacements (volume 2).
> 1. Does the file only contain the nonlinear part?
> 2. I find the Dimension of the file is smaller. Would you please tell me the reason?
> Thank for your help.
>
> Xuemei
|