I’m trying to get a 5TT output from freesurfer data as FSL would provide it using MRtrix wrapper 5ttgen fsl. mri_compute_volume_fractions looks like doing the job, but I can’t wrap my head around some tissue type that seemed to be missing.

For clarification, its infant data but it comes from a freesurfer based processing pipeline. If I’m using it like this:

I’ll get this error:

$Id: mri_compute_volume_fractions.c,v 1.22 2015/09/14 12:25:01 fischl Exp $

sysname  Darwin

hostname xxx.charite.de

machine  x86_64

user     martinbauer

setenv SUBJECTS_DIR /Users/martinbauer/Daten/MRI/FS_corrected

cd /Users/martinbauer

mri_compute_volume_fractions --o test --regheader 2035 /Users/martinbauer/Daten/MRI/FS_corrected/2035/mri/norm.mgz 

outstem test

regfile (null)

regtype 0

segfile aseg.mgz

wsurf white

psurf pial

TempVolFile (null)

USF 2

Reading in aseg and surfs from /Users/martinbauer/Daten/MRI/FS_corrected/2035

Loading /Users/martinbauer/Daten/MRI/FS_corrected/2035/mri/aseg.mgz

Changing type of seg from UCHAR to INT

Filling empty voxels with extracerebral CSF (if not there already), nDil=3

Computing registration from header

 1.00000   0.00000   0.00000   0.00000;

 0.00000   1.00000   0.00000   0.00000;

 0.00000   0.00000   1.00000   0.00000;

 0.00000   0.00000   0.00000   1.00000;

  t = 4.53563

Computing PVF (USF=2, resmm=0.500000)

  MRIhiresSeg(): filling unknowns with 257

ERROR: MRIseg2TTypeMap() tissue type for seg 172 Vermis not set

Fair enough, so I force it using finer parcellations doing:

mri_compute_volume_fractions --o test --regheader 2035 /Users/path/to/data/2035/mri/norm.mgz --seg aparc.a2009s+aseg.mgz

Vermis is now part of the game but:

$Id: mri_compute_volume_fractions.c,v 1.22 2015/09/14 12:25:01 fischl Exp $

sysname  Darwin

hostname xxx.charite.de

machine  x86_64

user     martinbauer

setenv SUBJECTS_DIR /Users/martinbauer/Daten/MRI/FS_corrected

cd /Users/martinbauer

mri_compute_volume_fractions --o test --regheader 2035 /Users/martinbauer/Daten/MRI/FS_corrected/2035/mri/norm.mgz --seg aparc.a2009s+aseg.mgz 

outstem test

regfile (null)

regtype 0

segfile aparc.a2009s+aseg.mgz

wsurf white

psurf pial

TempVolFile (null)

USF 2

Reading in aseg and surfs from /Users/martinbauer/Daten/MRI/FS_corrected/2035

Loading /Users/martinbauer/Daten/MRI/FS_corrected/2035/mri/aparc.a2009s+aseg.mgz

Filling empty voxels with extracerebral CSF (if not there already), nDil=3

Computing registration from header

 1.00000   0.00000   0.00000   0.00000;

 0.00000   1.00000   0.00000   0.00000;

 0.00000   0.00000   1.00000   0.00000;

 0.00000   0.00000   0.00000   1.00000;

  t = 4.40752

Computing PVF (USF=2, resmm=0.500000)

  MRIhiresSeg(): filling unknowns with 257

ERROR: MRIseg2TTypeMap() tissue type for seg 12104 ctx_rh_G_and_S_subcentral not set

What files are exactly read and can those files be adjusted? In FreeSurferColorLUT.txt Vermis and 12104 are included, I tried to add 12104 to the parcellation stats LUT but it didn’t solve the problem. I’m also open to go another way. 

Thank you so much in advance for pointing me towards the right direction!

All the best,

Martin