On Wed, 3 Mar 2010, Perlaki Gabor wrote:
Hi all!
We overlayed the aseg.mgz on the norm.mgz and put some controll points where
the cerebellum segmentation was not accurate and we saved the controll points.
We reran recon-all like this: 'recon-all -autorecon2-cp -autorecon3 -subjid
our_subj'
After that we checked the volume of cerebellum, and it is the same as it was
before the whole procedure, and aseg.mgz is the same as well. How can we check
that how accurate the segmentation of cerebellum after the reran of recon-all,
and where are the new results?
Thank you very much,
Gabor Perlaki
On Fri, 26 Feb 2010 08:49:48 -0500 (EST), Bruce Fischl wrote
Hi Gergely,
overlay the aseg.mgz on the norm.mgz in tkmedit and you can visually
verify if it's accurate. If not, you can edit it with the tools in
tkmedit. We do have some internal tools for VBM-style analysis but
rarely use them as the results are so ambiguous. It should be easy
enough to stuff through FSL or SPM VMB though.
cheers,
Bruce
On Fri, 26 Feb 2010, Gergely Orsi wrote:
Hi all!
We have processed 100 subjects with autorecon.
I'd like to know how to check the cerebellum segmentation on the
subjects. It would be crucial, because the images are a bit dimmer at the
lower areas (it's far from the isocenter) and despite the intensity
normalisation it still has lower intensity compared to the dorsal part of
the cerebellum. After the skull-strip the cerebellum is still intact (the
whole cerebellum exists on the images), but I'd like to know wether
freesurfer recognised the lower intensity parts or not.
Unfortunatly BET in FSL simply cut off these areas, thus our VBM results
(showing significan changes at the ventral cerebellum) are compromised.
We have measured significant group differences in total cerebellar volume
using freesurfer, but we should check the segmentation first, and believe
the results later.
Could anyone help us in this?
And if the segmentation failed, is there any manual method to correct it (by
hand), because the whole cerebellum is on the images (despite of the
different intensity).
And one last question, is it possible to carry out VBM in freesurfer (after
checking/correcting the cerebellum if needed, although the intensity
difference will still be a big problem in VBM ).
Thank you very much.
Gergely Orsi
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
--
Pécsi Tudományegyetem Természettudományi Kar (http://www.ttk.pte.hu/)
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer