That one now automatically does subcortical segmentation, which will take about 24 hours. Add -nosubcortseg to the command-line (AFTER -stage1) to disable it.

doug

Timothy O'Keefe wrote:

Also, I am using the OSX Tiger snapshot of FreeSurfer with a new recon-all script courtesy of Doug Greve.

On Jul 18, 2005, at 10:03 AM, Timothy O'Keefe wrote:

...

I ran recon-all stage1 and it took a very long time. Well, at least longer than people document. I didn't time it, but it must have taken close to 10 hours.

Also I received this error during mri_fill(): ... using +/- offset search region where offset is 39..... using +/- offset search region where offset is 42..... using +/- offset search region where offset is 45..... using +/- offset search region where offset is 48..... mri_fill: could not find pons No such file or directory INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst talairach cc position changed to (0.90, -7.51, 23.75) INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst find_cutting_plane:seed point not in structure! Searching neighborhood... find_cutting_plane:seed point not in structure! Searching neighborhood... ERROR: mri_fill exited with non-zero status recon-all exited with errors at Sat Jul 16 01:13:52 EDT 2005

Are there any suggestions? I am using a single T1 (1x1x1) image acquired from a Siemens 1.5T machine. I realize I should use multiple images for averaging, however I am just trying to get a feel for the software.

-T _______________________________________________ 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

Hi Ray,

you shouldn't compare across datasets that use the subcortical segmentation (aseg) and those that don't. The aseg fills the entire ventricle, and thus changes the effective folding pattern on the medial wall, requiring a different atlas and parcellation targets. The advantage is that it is quite a bit more robust and fully automated in 90-95% of the cases we've tested (including elderly and AD), plus you get the whole subcortical segmentation for either structural or functional analysis.

cheers, Bruce

p.s. the subcortical does take a long (>15 hours) time, but reduces person time as most datasets go through without any manual fixing.

On Mon, 18 Jul 2005, Ray Fix wrote:

Hello Doug,

Thanks for the info. I tried out the Tiger install a couple of weeks ago and was very impressed but when I tried to process bert, it took so long that I finally gave up.

What is the advantage of doing this subcortical segmentation step? For example, will it potentially improve the results of later stages?

In the README file in freesurfer/subjects it states that the sample should be processed with

recon-all-nmr -all -subjid bert

Do you recommend adding -nosubcortseg to the recon-all-nmr script?

Previously, I used to use recon-all. It looks like it that is a grepped / stripped subset of the -nmr version. Is recon-all deprecated?

When I run recon-all with bert I get an error about DoCANormalize not being defined.

Will I be able to correctly compare samples that I have run in previously releases using recon-all?

Thank you. Yours confused,

Ray Fix

On Jul 18, 2005, at 10:26 AM, Doug Greve wrote:

...

That one now automatically does subcortical segmentation, which will take about 24 hours. Add -nosubcortseg to the command-line (AFTER -stage1) to disable it.

doug

Timothy O'Keefe wrote:

...

Also, I am using the OSX Tiger snapshot of FreeSurfer with a new recon-all script courtesy of Doug Greve.

On Jul 18, 2005, at 10:03 AM, Timothy O'Keefe wrote:

...

I ran recon-all stage1 and it took a very long time. Well, at least longer than people document. I didn't time it, but it must have taken close to 10 hours.

Also I received this error during mri_fill(): ... using +/- offset search region where offset is 39..... using +/- offset search region where offset is 42..... using +/- offset search region where offset is 45..... using +/- offset search region where offset is 48..... mri_fill: could not find pons No such file or directory INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst talairach cc position changed to (0.90, -7.51, 23.75) INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst find_cutting_plane:seed point not in structure! Searching neighborhood... find_cutting_plane:seed point not in structure! Searching neighborhood... ERROR: mri_fill exited with non-zero status recon-all exited with errors at Sat Jul 16 01:13:52 EDT 2005

Are there any suggestions? I am using a single T1 (1x1x1) image acquired from a Siemens 1.5T machine. I realize I should use multiple images for averaging, however I am just trying to get a feel for the software.

-T _______________________________________________ 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

-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422

---

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