Hi Eugene,

Thanks. Let me see if I've understood this properly.

T1.nii (1mm isotropic)
EPI_A.nii (3mm isotropic)
EPI_B.nii (3mm isotropic - coregistered with EPI_A, but 4D)

We want to coregister the EPI to the T1 (ignore the problems of doing
this for the moment). So we run flirt to get the transformation
matrix:

flirt -in EPI_A.nii -ref T1.nii -omat register.mat

We now want to transform EPI_B such that it is coregistered with the
T1, but the voxel sizes remain as 3mm isotropic. Do we do the
following?

applyxfm4D EPI_B.nii EPI_A.nii EPI_B_trans.nii register.mat -singlematrix

This should give me EPI_B coregistered with T1, but with the voxel
resolutions of EPI_A. Is that correct?

If you use EPI_A the voxel dimensions will be correct, but the sform/qform matrices may not match the t1, so that the image would not be in the same space as the t1 (e.g. in fslview, the positions in mm will be different). This could cause you problems. An easy way to create the appropriate 3mm reference image is to use the ApplyXFM GUI (thanks Steve!). Select Use identity transformation, use the t1 image is input and your EPI image as reference. ApplyXFM also allows you to explicitly specify the output image and voxel dims.

Best,

Eugene

Cheers,

Satra

On Fri, Mar 20, 2009 at 11:54 AM, Eugene Duff <[log in to unmask]> wrote:
> Hi Satra,
>
> The output voxel size is determined by the refvol you provide with the -ref
> option in flirt. This can be different from the target image used when the
> transformation was calculated. So you just need to use a refvol image with
> the image properties you desire.
>
> Hope that helps.
>
> Eugene
>
>
>
> 2009/3/20 Satrajit Ghosh <[log in to unmask]>
>>
>> I have a registration file that registers a functional image (3mm
>> isotropic) to a structural image (1mm isotropic). Is there a way to
>> use applyxfm4d or another tool such that when I transform the
>> functional images, I can specify the voxel resolution of the
>> transformed volume?
>>
>> Thanks,
>>
>> Satra
>>
>>
>> On Fri, Mar 20, 2009 at 11:21 AM, Mahinda Yogarajah
>> <[log in to unmask]> wrote:
>> > Dear FSL Users,
>> >
>> > Is there an option in FNIRT normalisation equivalent to the preserve
>> > concentrations option in SPM5 when one writes out normalised images ?
>> >
>> > Thanks.
>> >
>> > Mahinda
>> >
>
>
>
> --
>
> Eugene Duff
>
> FMRIB Centre,
> University of Oxford
> John Radcliffe Hospital, Headington OX3 9DU Oxford UK
>
> Ph: +44 (0) 1865 222 739 Fax: +44 (0) 1865 222 717
>
> --
>