Hi Aditya,
> 1. BET on 4D .nii diffusion data. My raw data is dicom
> format from 3.0 T GE Signa HDx, has 65 encoding directions,
> 23 slices, non-isotropic voxels. i converted the DICOM files
> to .nii.gz using MriCroN and got the bvals and bvecs files
> (see attached files- are these correct?), as suggested
> earlier.
> 2. prepare mask by fslroi (i will have to prepare two
> separate masks, isnt it? )
> 3. flirt
> 4. fnirt
> 5. run eddycorrect
> 6. run bedpostx
> 7. run probtrackx
> 8. run dtifit
> 9. view results in fslview
Not quite.
1.eddycorrect
2.fslroi to extract your b0 and then bet on it to get the mask of the brain
3.bedpostx (see manual)
4.dtifit to get the FA map
5.flirt FA to FMRIB58
6.if your subjects are kids:average the flirted FA to create your own template with the same number of autistic and healthy subjects
7.fnirt FA to either study-specific FA template (if kids) or to FRMIB58 if adults
8.probtrackx with warpfields generated from fnirt (see manual)
> Is this the right approach to do the dti analysis i
> refffered earlier for autism (n=6) vs control (n=7). Am i
> correct in assuming that output from one stage is input to
> the next stage.
> am i supposed to do coregistration of all datasets BEFORE i
> run eddycorrect?
No, see above.
> typically when and how should i prepare the mean images of
> the two groups for comparing FA values for autism>control
> and control>autism, and how do i do that from the gui? or
> can we do it using command line only?subsequently if i want
> to add more data sets to the study as more volunteers are
> recruited, do i need to run the entire analysis again?
>
>
> i also wanted to analyse the connectivity b/w broca and
> wernicke.
> please help and thanks a lot in advance.
You would need to know what you want to investigate. What are exactly your scientific/clinical questions regarding this analysis?
Please look at the archives and literature, it might help.
Cheers,
Gwenaelle
|