Hi Michael
thanks for the info, we'll look into this. We are aware of some of the
issues with running multiple times, but need to look into it further when
we find the time
Bruce
On Wed, 22 Jun 2011, Michael Scheel wrote:
Dear Freesurfer experts,
we are testing a grid computer infrastructure for running freesurfer
analysis and are evaluating consistency and reliability of freesurfer
results across platforms.
Before we start our tests across platforms we conducted some freesurfer
analysis on the same computer and had some unexpected results.
Subsequent executions of the freesurfer pipeline ("New invocation of
recon-all") on a certain subject - without any manual editing in between
- give different results. We have invoked recon-all up to four times and
all results are different. We tested in particular the aseg.stats
results, where we found different volume estimations (up to 4%)
depending on the number of processing loops.
Intermediate results have also been found to be different
The effect is reproducible, meaning the nth invocation of recon-all on
the subject gives always the nth result.
We suspect that the cause is (following the pipeline documentation:
http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable) the twofold
generation of "talairach_with_skull.lta", once in the autorecon-1 and a
second time in the autorecon-2 phase:
autorecon1:
mri_em_register -skull nu.mgz
$FREESURFER_HOME/average/RB_all_withskull_2008-03-26.gca
transforms/talairach_with_skull.lta
autorecon2:
mri_em_register -skull -t transforms/talairach.lta nu_noneck.mgz
$FREESURFER_HOME/average/RB_all_withskull_2008-03-26.gca
transforms/talairach_with_skull.lta
and it seems that the step in autorecon1 is skipped if the file
"talairach_with_skull.lta" is already present.
We were wondering if this behaviour is intended, so are the FreeSurfer
results expected to improve with multiple invocations, even without
manual intervention? Shall they converge after sufficient re-runs?
Any help or comments are highly appreciated.
Best, Michael
_______________________________________________
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.
