Dear all,
I have been trying to use the eTIV value FS outputs after the -autorecon1 and -autorecon2 stages, registering it in the aseg.stats file. The purpose would be to see how reliable that value is when compared to our own manual measurements. When reading the Buckner paper, I have noticed that a stringent requirement for optimal perfomance is the usage of an adequate template (i.e., age matched template), and I believe FS uses the MNI 305 template (at least in 2006 versions), which is not adequate for our population.
My questions are:
1) Is it possible to build our own template and then introducing it in the FS package? If so, where? (please notice that I am not interested in surface templates) 2) If 1) is possible, is the TIV estimation hindered by such a procedure? I know you have adjusted the calculation of the eTIV using bootstrap, which means that you don't use a pure version of the atlas scale factor (ASF), but you scale it by a fixed number you've obtained with this adjustment. My guess is that this estimate will no longer be valid. 3) I have adjusted the initial transform outputs by using mri_total, T1.mgz, brain.mgz and, in some cases manual adjustments. This affected talairach.xfm. Is this the adequate approach for TIV estimation purposes? I remember reading something about talairach_with_skull.lta...
Thank you very much for your patience and assistance!
Regards,
Joao Pereira
Hi Joao,
we actually don't use the MNI 305 for this purpose (or for much else, except reporting talairach coords). Are you finding the eTIVs don't match your manual measures? If you plot them, are they linearly related, or are the eTIVs just wrong? In general we don't need separate atlases for things like segmentation purposed because we keep much more information in the atlas than just the simple average that goes into the 305, so our hope was (and preliminary studies supported this) that we wouldn't need separate atlases for separate populations.
cheers, Bruce
On Tue, 24 Oct 2006, Joao Pereira wrote:
Dear all,
I have been trying to use the eTIV value FS outputs after the -autorecon1 and -autorecon2 stages, registering it in the aseg.stats file. The purpose would be to see how reliable that value is when compared to our own manual measurements. When reading the Buckner paper, I have noticed that a stringent requirement for optimal perfomance is the usage of an adequate template (i.e., age matched template), and I believe FS uses the MNI 305 template (at least in 2006 versions), which is not adequate for our population.
My questions are:
- Is it possible to build our own template and then introducing it in the
FS package? If so, where? (please notice that I am not interested in surface templates) 2) If 1) is possible, is the TIV estimation hindered by such a procedure? I know you have adjusted the calculation of the eTIV using bootstrap, which means that you don't use a pure version of the atlas scale factor (ASF), but you scale it by a fixed number you've obtained with this adjustment. My guess is that this estimate will no longer be valid. 3) I have adjusted the initial transform outputs by using mri_total, T1.mgz, brain.mgz and, in some cases manual adjustments. This affected talairach.xfm. Is this the adequate approach for TIV estimation purposes? I remember reading something about talairach_with_skull.lta...
Thank you very much for your patience and assistance!
Regards,
Joao Pereira
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Joao Pereira