Exactly, just use a t1_brain (i.e. bet t1 first). 

Cheers

Stam

On 11 Jul 2014, at 11:29, Linden Parkes <[log in to unmask]> wrote:

Ah k

So if I run flirt with -in Diffusion.bedpostX/nodif_brain.nii.gz -ref /t1.nii.gz  -omat /nodif_brain.mat

the nodif_brain.mat is equivalent to the diff2str.mat from the tutorial? and then if the take the inverse of that using convert_xfm -inverse to get the str2diff.mat?

On 11 July 2014 19:16, Stamatios Sotiropoulos <[log in to unmask]> wrote:

Yes, these are the steps I described. Have a look at the commands I included.

Cheers

Stam

On 11 Jul 2014, at 10:09, Linden Parkes <[log in to unmask]> wrote:

Hi Stam,

Im basically trying to recreate this step (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FDT/UserGuide#Registration_within_FDT) using the command line

Cheers,

Linden

On 11 July 2014 19:06, Stamatios Sotiropoulos <[log in to unmask]> wrote:

Hi Linden

The bedpostx directory can be obtained using a bedpostx call. Just to clarify, is the diffusion-to-standard space registration commands that you are also after? You would use the T1 as an intermediate and then:

1. Register a b0 diffusion volume to the T1 using linear registration and flirt or using BBR and epi_reg.
2. Register the T1 to MNI space using non-linear registration and fnirt.
3. Concatenate the two transforms using convertwarp.

Cheers
Stam

On 11 Jul 2014, at 07:09, SUBSCRIBE FSL Linden M Parkes <[log in to unmask]> wrote:

> Hi FSLers,
>
> I'm using the registration within FDT tool from the GUI (version 5.0.6) and I'd like to script it but I can't find the appropriate command line call? Is it the same as the standard FSL registration? (fsl_reg)
> If so, can someone show me the command line call to replicate:
> BEDPOSTX directory: /001/Diffusion.bedpostX
> Main structural image: /001/t1_brain.nii.gz
> Non-betted structural (for nonlinear reg): 001/t1.nii.gz
> Standard space: /usr/local/fsl/data/standard/MNI152_T1_2mm_brain
>
> Cheers,
> Linden