Brainiac . wrote:
> Hi All,
>
> I need to classify WM into different lobes (temporal, parietal, 
> frontal, occipital). I tried to do the lobe mapping using 
> parcellations from wmparc.mgz. 
> http://surfer.nmr.mgh.harvard.edu/fswiki/CorticalParcellation
>  
> My problem is the "unsegmented white matter" in wmparc.mgz. My 
> understanding is that these unsegmented areas are classified so 
> because they are beyond 5mm from any pre-classified cortex area.
>
> My questions:
> (1) So in order to classify "unsegmented WM" into lobes, do I need to 
> take each voxel in the unsegmented WM region and then assign it to the 
> closest cortex area to classify it into respective parcellations. Then 
> re-group into lobes. Is this correct?

>
> (2) Are there any tailor made function or any other easier way to do 
> this? Is there any easier way to segment WM into lobes directly that I 
> have missed?
>
> Thanks,
> Brain

You can do this in a couple of steps. First, you'll need to create a 
lobe annotation. You can do this in 3 ways:
   1. mri_annotation2label --lobesStrict 
  2. mri_annotation2label --lobes
  3. Custom: break up annotation and recombine:
      a. mri_annotation to create separate label files
      b. mri_mergelabels to create single label files for each lobe label
      c. mris_label2annot to create the lobe annotation

Next, create the WM segmentation with mri_aparc2aseg, specifying your 
new lobe annot and adding --labelwm. You can control the distance  with 
--wmparc-dmax

doug