Hi
See http://www.fmrib.ox.ac.uk/fsl/fnirt/index.html (scroll to the very bottom of that page) for example usage, including the scenario you're asking about, i.e. using the estimated warps to register functional data to some template space.
hth
Christian
On 8 Feb 2010, at 21:30, Matthew Ward wrote:
> Hello everyone. I’ve recently successfully used FLIRT/FNIRT to register
> two different subjects’ anatomical volumes by first performing an affine
> transform and then utilizing FNIRT. While the FLIRT output includes the
> .mat file describing the affine transform, I am uncertain how/if it is
> possible to get FNIRT to generate a similar file so that I can apply it to a
> subject’s functional data. The ultimate goal is to generate an ‘average’
> brain for the subjects in the study in order to display group results (to my
> knowledge, there is no ‘template’ for the particular species I’m using).
> So, to summarize:
>
> 1. FSL registration step for each subject: functional-to-scout (3dof),
> scout-to-hi resolution
> volume (affine), functional-to-hi resolution (using affine .mat file).
> 2. Using FLIRT and FNIRT, register each subject to a primary subject’s
> anatomical (output
> to be used in creating an ‘average’ brain for subject pool).
> 3. Now, is it possible to generate/use the non-linear transforms from FNIRT
> (step 2) on
> the individual subjects’ functional data/statistical maps in order to
> get them all in the
> same non-linearly registered space?
>
> I apologize if my terminology is not standard, but I’ve been working on this
> process in isolation and may have made up my own terminology in the process.
> In any case, I hope the meaning comes through. Thanks for your consideration!
>
> Best,
>
> Matt
|