Hi Michael
we found that the conversion to unsigned character sometimes compressed
the brain into only a couple of bits of the dynamic range. We know use
the talairach xform to locate the center of the brain and make sure that
a ball centered at the tal origin occupies a lot of the histogram. This
helped in a number of cases, but maybe the makes the tal xform more
sensitive to bias fields and strange intensity scalings.
Bruce
On Wed, 2 Nov 2011,
Michael Waskom wrote:
Hi,
What was the reason for this switch? Intuitively it seems that a
bias-corrected image is likely to get a better registration (since I
assume the target has essentially uniform intensities).
Michael
On Wed, Nov 2, 2011 at 1:15 PM, Nick Schmansky <ni...@nmr.mgh.harvard.edu>
wrote:
no, v5.0 uses nu.mgz as input to the talairach_avi routine, whereas v5.1
uses orig.mgz.
its hard to say why it would fail for some subjects but not others,
other than head positioning and to some degree ventricle size makes a
difference.
n.
On Wed, 2011-11-02 at 16:10 -0400, Joshua Lee wrote:
> Thanks Nick,
>
> Is use-mritotal equivalent to what version 5.0 does by default?
>
> Also, any suggestions on why this talairach error would occur only on
> a portion of my data, despite all being collected with the same
> imaging protocol, same conversions from dicom, etc?
> -
>
> Joshua Lee
>
> Graduate Student
> Center for Mind and Brain &
>
> Department of Psychology
>
> University of California, Davis
> 530.747.3805
>
>
> On Wed, Nov 2, 2011 at 4:02 PM, Nick Schmansky
> <ni...@nmr.mgh.harvard.edu> wrote:
> joshua,
>
> try adding -use-mritotal to the end of your recon-all
> command. it will
> use an alternate way to compute the talairach.xfm which works
> better on
> some data.
>
> the change from v5.0 to v5.1 in the talairach stage was that
> in v5.1,
> the talairach.xfm is needed to compute nu.mgz, so the
> talairach stage
> needs to precede the nu_correct stage. this means the
> talairach stage
> can no longer use nu.mgz as input, which seems to work better
> than
> orig.mgz for some data (and is the case for your data).
>
> in the next release of freesurfer, recon-all will
> automatically try the
> alternate method (mritotal) if the default talairach_avi
> method fails
> the tal-check.
>
> n.
>
>
>
> On Wed, 2011-11-02 at 08:45 -0400, Bruce Fischl wrote:
> > can you send us an example that works in 5.0 and fails in
> 5.1?
> >
> > thanks
> > Bruce
> > On Wed, 2 Nov
> > 2011, Joshua Lee wrote:
> >
> > > Small Update: Error Disappears if Using Freesurfer version
> 5.0 instead of
> > > 5.1. However, since I am keenly interested in the
> hippocampal-subfield
> > > segmentation capabilities of the most recent version, I am
> loathe to regress
> > > to an earlier version of FreeSurfer
> > > -
> > > Joshua Lee
> > > Graduate Student
> > > Center for Mind and Brain &
> > > Department of Psychology
> > > University of California, Davis
> > > 530.747.3805
> > >
> > >
> > > On Wed, Nov 2, 2011 at 2:29 AM, Joshua Lee
> <jki...@ucdavis.edu> wrote:
> > > #@# Talairach Failure Detection Wed Nov 2 02:03:32
> EDT 2011
> > > /home/virtualuser/freesurfer/subjects/test005/mri
> > >
> > > talairach_afd -T 0.005 -xfm
> transforms/talairach.xfm
> > >
> > > I am receiving the error (pasted below) on about
> half of my
> > > scans, despite all cans using same protocol. I
> detect no obvious
> > > issue with the quality of the images. I have
> original dicoms,
> > > and am using mri_convert to save into a single .mgz
> (or .nii,
> > > doesn't neem to make a difference), and the
> unpredictability of
> > > the error makes diagnosis problematic. TKMedit2
> reveals
> > > absolutely wacky registrations. I am processing on
> Freesurfer
> > > 5.1, and have replicated issue on two different
> installations. I
> > > am designing a new longitudinal study inheriting the
> same
> > > imaging protocol and I do not want to get in a
> situation
> > > demanding manual editing of the talairach. Can
> anyone help me
> > > out? Thanks, Joshua
> > >
> > > ERROR: talairach_afd: Talairach Transform:
> > > transforms/talairach.xfm ***FAILED*** (p=0.0030,
> pval=0.0000 <
> > > threshold=0.0050)
> > > Manual Talairach alignment may be necessary, or
> > > include the -notal-check flag to skip this test,
> > > making sure the -notal-check flag follows -all
> > > or -autorecon1 in the command string.
> > > See
> > >
> http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/Talairach
> > > Linux FreeSurfer 2.6.28-11-generic #42-Ubuntu SMP
> Fri Apr 17
> > > 01:57:59 UTC 2009 i686 GNU/Linux
> > >
> > > recon-all -s test005 exited with ERRORS at Wed Nov
> 2 02:03:32
> > > EDT 2011
> > >
> > > For more details, see the log file
> > >
> /home/virtualuser/freesurfer/subjects/test005/scripts/recon-all.log
> > > To report a problem, see
> > >
> http://surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
> > >
> > > -
> > > Joshua Lee
> > > Graduate Student
> > > Center for Mind and Brain &
> > > Department of Psychology
> > > University of California, Davis
> > > 530.747.3805
> > >
> > >
> > >
> > >
>
> > _______________________________________________ Freesurfer
> mailing list Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>
>
>
>
> The information in this e-mail is intended only for the person
> to whom it is
> addressed. If you believe this e-mail was sent to you in error
> and the e-mail
> contains patient information, please contact the Partners
> Compliance HelpLine at
> http://www.partners.org/complianceline . If the e-mail was
> sent to you in error
> but does not contain patient information, please contact the
> sender and properly
> dispose of the e-mail.
>
>
>
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.
