Hello,
In your experience, is the "talairach_with_skull.lta" transform always reasonable, or does it sometimes fail (and if the latter, is the failure rate dependent on the FS version, as is the case for talairach.xfm ?)
I ask because the ICV ("eTIV") value is derived from the determinant of "talairach_with_skull.lta". So, if we want to use the eTIV values as a covariate, should we be verifying the accuracy of this transform? And if so, what would the syntax be to do this in tkregister2 ?
thanks, Mike H.
Those are good questions. Bruce, does this use the mni305 template or a different one?
On Tue, 19 Feb 2008, Michael Harms wrote:
Hello,
In your experience, is the "talairach_with_skull.lta" transform always reasonable, or does it sometimes fail (and if the latter, is the failure rate dependent on the FS version, as is the case for talairach.xfm ?)
I ask because the ICV ("eTIV") value is derived from the determinant of "talairach_with_skull.lta". So, if we want to use the eTIV values as a covariate, should we be verifying the accuracy of this transform? And if so, what would the syntax be to do this in tkregister2 ?
thanks, Mike H.
no, it's our own (.gca) template. It's the same type of linear registration that is used in the aseg, just this time to a template that has a skull in it On Tue, 19 Feb 2008, Doug Greve wrote:
Those are good questions. Bruce, does this use the mni305 template or a different one?
On Tue, 19 Feb 2008, Michael Harms wrote:
Hello,
In your experience, is the "talairach_with_skull.lta" transform always reasonable, or does it sometimes fail (and if the latter, is the failure rate dependent on the FS version, as is the case for talairach.xfm ?)
I ask because the ICV ("eTIV") value is derived from the determinant of "talairach_with_skull.lta". So, if we want to use the eTIV values as a covariate, should we be verifying the accuracy of this transform? And if so, what would the syntax be to do this in tkregister2 ?
thanks, Mike H.
You can try something like this:
mri_convert $FREESURFER_HOME/average/RB_all_withskull_2007-08-08.gca gca.mgh --frame 0
tkregister2 --targ gca.mgh --mov T1.mgz \ --lta transforms/talairach_with_skull.lta --reg tmp.reg
doug
On Tue, 19 Feb 2008, Bruce Fischl wrote:
no, it's our own (.gca) template. It's the same type of linear registration that is used in the aseg, just this time to a template that has a skull in it On Tue, 19 Feb 2008, Doug Greve wrote:
Those are good questions. Bruce, does this use the mni305 template or a different one?
On Tue, 19 Feb 2008, Michael Harms wrote:
Hello,
In your experience, is the "talairach_with_skull.lta" transform always reasonable, or does it sometimes fail (and if the latter, is the failure rate dependent on the FS version, as is the case for talairach.xfm ?)
I ask because the ICV ("eTIV") value is derived from the determinant of "talairach_with_skull.lta". So, if we want to use the eTIV values as a covariate, should we be verifying the accuracy of this transform? And if so, what would the syntax be to do this in tkregister2 ?
thanks, Mike H.
Hi Mike,
sorry, forgot to respond to this one. Not sure about tkregister2 - I'll leave that for Doug. I think you should be able to use mri_convert with the .gca file to extra the mean image frame (frame 0), then check the aligment to that.
cheers, Bruce
On Tue, 19 Feb 2008, Michael Harms wrote:
Hello,
In your experience, is the "talairach_with_skull.lta" transform always reasonable, or does it sometimes fail (and if the latter, is the failure rate dependent on the FS version, as is the case for talairach.xfm ?)
I ask because the ICV ("eTIV") value is derived from the determinant of "talairach_with_skull.lta". So, if we want to use the eTIV values as a covariate, should we be verifying the accuracy of this transform? And if so, what would the syntax be to do this in tkregister2 ?
thanks, Mike H.
p.s. remember if/when you check it that it is supposed to align skull, not necessarily brain. On Tue, 19 Feb 2008, Michael Harms wrote:
Hello,
In your experience, is the "talairach_with_skull.lta" transform always reasonable, or does it sometimes fail (and if the latter, is the failure rate dependent on the FS version, as is the case for talairach.xfm ?)
I ask because the ICV ("eTIV") value is derived from the determinant of "talairach_with_skull.lta". So, if we want to use the eTIV values as a covariate, should we be verifying the accuracy of this transform? And if so, what would the syntax be to do this in tkregister2 ?
thanks, Mike H.
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Doug Greve -
Michael Harms