[Mrtrix-discussion] T1/dwi mismatch

[Luis Concha]

Hello Hak.

You need to register your T1 and  your DWI-derived maps. If you did not use
some form of geometric distortion correction (revpe_distcorr
<https://github.com/jdtournier/mrtrix3/wiki/Tutorial-DWI-processing> or
topup <http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/TOPUP>, for example), then
getting a perfect registration between these two imaging modalities is
virtually impossible. However, you can get usable results using a
non-linear registration. You can spend months trying to improve your
registration, if you want to go down the rabbit hole...

The trick I most like for this purpose was described by Besson et al
<http://www.ncbi.nlm.nih.gov/pubmed/24814212> and it involves obtaining the
partial volume estimation of CSF from your T1 and to that you non-linearly
register your ADC map or b=0 image. You can also do this between the white
matter PVE estimation from T1 and your FA map.  My implementation below, i
hope it helps:


echo "  STEP:  Run FAST on the T1 volume"
fast -v -S 1 -n 3 -t 1 -I 1 -g -N \
  -o Fast_t1 \
  $t1

echo "  STEP:  Run FLIRT between the t1 CSF map and the ADC map"
flirt \
  -in Fast_t1_pve_0 \
  -ref $adc \
  -omat ${outbase}_lin.mat \
  -out lin_csf2adc_transformed

echo "  STEP:  Non-linear registration with FNIRT"
my_do_cmd fnirt -v \
  --in=Fast_t1_pve_0 \
  --ref=$adc \
  --fout=${outbase}_field \
  --aff=${outbase}_lin.mat

echo "  STEP: Apply the resulting warp field"
my_do_cmd applywarp -v \
  -i $t1 \
  -o ${outbase}_t1_to_dwi \
  -r $adc \
  -w ${outbase}_field

Dr. Luis Concha
Instituto de Neurobiología
Laboratorio C-13
UNAM, Campus Juriquilla
Boulervard Juriquilla 3001
Juriquilla, Querétaro.
C.P. 76230
México
Tel (442) 2 38 10 54
Fax (442) 2 38 10 46
http://personal.inb.unam.mx/lconcha/

On Thu, Apr 23, 2015 at 8:09 AM, Hak Fujiyama <
Hak.Fujiyama at faber.kuleuven.be> wrote:

>  Dear all,
>
>
>
> We got our DWI images through a Philips scanner. This spit out PAR/REC –
> files. Converting them to .nii could be done in FSL with the PARconv_v1.12
> command.
>
> However, since this is no DICOM format, we were unable to co-register our
> dwi images to our structural scan in MRTRIX.
>
>
>
> Is there a way to do this? It would be a great help!
>
>
> For now, we’re getting nice tracts but we can’t visualize them because of
> this dwi/T1 mismatch.
>
>
>
> Many thanks!
>
>
>
> H.
>
>
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20150423/09fb25bd/attachment.html>