Print

Print


Hello Phil,

> First, thank you for your work in creating a toolbox that will allow
> creation of a custom TPM file. It is a very interesting toolbox, but
> I feel I may be missing something in the instructions.

First, thanks for the flowers :)

> To be specific, I am having issues in understanding exactly what
> images to use as the input when creating the TOM.mat file. I think I
> am suppose to use the normalized, unmodulated segmented files for
> each tissue type (c1* c2*, etc). Is this correct?

Well, yes, you would enter the wc*. I prefer the affine-only registered 
version, more below.

> If one uses the VBM8 toolbox, there is no option to write out all 6
> classes of images, only gm, wm and csf. How does one get the other
> tissue files?

I have actually never looked at that but it seems that one can trick the
toolbox into doing this when modifying lines 75-79 (in the current
version) of cg_vbm8_run:

% never write class 4-6
for i=4:6
     tissue(i).warped = [0 0 0];
     tissue(i).native = [0 0 0];
end

If you adapt those lines according to what you want, that should do it.

> If you use the results of new segment, as per the current VBM
> instructions in SPM8, one gets nonlinear transformed segmented images
> that may have some segmentation artifacts (at least that is the case
> for many of my elderly subjects).

I agree that especially for "unusual" subjects, VBM8 is likely to give
better results. One way around the non-linear normalization would be to
write out the tissue classes in native space and then do an affine-only
normalization onto the corresponding tissue class that comes with spm
(i.e., normalize c1 to grey.img and take along the other classes). In
the TOM paper, we actually normalized each class and averaged the
parameters, which seemed like a good idea at that time but I am not sure
this really makes that much of a difference.

> So if I want to create a 6 tissue TPM.nii file, what images should I
> use?

Assuming I was right, you have two choices now :)
Hope this helps,
Marko
-- 
____________________________________________________
PD Dr. med. Marko Wilke
  Facharzt für Kinder- und Jugendmedizin
  Leiter, Experimentelle Pädiatrische Neurobildgebung
  Universitäts-Kinderklinik
  Abt. III (Neuropädiatrie)


Marko Wilke, MD, PhD
  Pediatrician
  Head, Experimental Pediatric Neuroimaging
  University Children's Hospital
  Dept. III (Pediatric Neurology)


Hoppe-Seyler-Str. 1
  D - 72076 Tübingen, Germany
  Tel. +49 7071 29-83416
  Fax  +49 7071 29-5473
  [log in to unmask]

  http://www.medizin.uni-tuebingen.de/kinder/epn
____________________________________________________