I just added the relevant stuff to recon-all and mri_nu_correct.mni. I've put the new versions at:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve
along with a sample xopts file.
Would you test this out for us? Make sure to make backups of your current versions of recon-all and mri_nu_correct.mni :).
doug
Sebastian Moeller wrote:
Hi List,
On 3. Nov 2006, at 19:48 Uhr, freesurfer-request@nmr.mgh.harvard.edu wrote:
Message: 10 Date: Fri, 03 Nov 2006 13:48:18 -0500 From: Nick Schmansky nicks@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] parameters for N3 correction To: Chacko Cherian chacko_cherian@yahoo.com Cc: Freesurfer Mailing List freesurfer@nmr.mgh.harvard.edu Message-ID: 1162579698.27296.26.camel@minerva.nmr.mgh.harvard.edu Content-Type: text/plain
also, the command:
recon-all -s <subjid> -nuintensitycor
will run just that stage.
Something relted on the topic of N3. Using the default settingswith surface coils (RX and TX through the same coil in our case) ends up looking ugly no matter the number of -nuiterations specified in recon-all. Hacking mri_nu_correct.mni to include "-iterations 150 -stop 0.0001" in the invocation of nu_correct, ends up with better results. Just as the N3 documentation recommends for surface coils ;). Now would it be possible to get a "-surfacecoil" argument for mri_nu_correct.mni (or even better recon_all via the experts options mechanism?) included into the distribution. If yes I would be willing to hack something up (at least including this properly into mri_nu_correct.mni). This might be helpful for other surface coil users out there...
ahoi & sorry