Hi all,
I've searched the archives a bit, and found only one truly similar error which seems like it was never answered (that I could see, anyway).
The case we're trying to process stopped during mris_fix_topology at this point:
CORRECTING DEFECT 50 (vertices=22858, convex hull=4870) normal vector of length zero at vertex 153186 with 0 faces vertex 153186 has 0 face Numerical result out of range
and then recon-all exits with errors.
I checked the euler number, which seems high: eno=-988 (nv=200584, nf=403144, ne=604716)
and the surface does report a very large number of defects: 166466 ambiguous faces found in tessellation segmenting defects... 185 defects found, arbitrating ambiguous regions...
but from what I can see on orig.nofix, nothing seems really grossly wrong (i.e. no major skullstripping errors); from the defects labels, most of the defects seem to be in the sulci.
I can send along the full logs if that would be helpful. Any advice?
I'm using freesurfer-Linux-rh9-stable-pub-v4.2.0
Thank you, Diandra
wow, that is a ton of defects. The surface also looks pretty big. Is the cerebellum or skull attached? What kind of scan is it?
Bruce
On Fri, 19 Nov 2010, Diandra Lucia wrote:
Hi all,
I've searched the archives a bit, and found only one truly similar error which seems like it was never answered (that I could see, anyway).
The case we're trying to process stopped during mris_fix_topology at this point:
CORRECTING DEFECT 50 (vertices=22858, convex hull=4870) normal vector of length zero at vertex 153186 with 0 faces vertex 153186 has 0 face Numerical result out of range
and then recon-all exits with errors.
I checked the euler number, which seems high: eno=-988 (nv=200584, nf=403144, ne=604716)
and the surface does report a very large number of defects: 166466 ambiguous faces found in tessellation segmenting defects... 185 defects found, arbitrating ambiguous regions...
but from what I can see on orig.nofix, nothing seems really grossly wrong (i.e. no major skullstripping errors); from the defects labels, most of the defects seem to be in the sulci.
I can send along the full logs if that would be helpful. Any advice?
I'm using freesurfer-Linux-rh9-stable-pub-v4.2.0
Thank you, Diandra
Hi Bruce,
No cerebellum attached in the *.nofix files; maybe a tiny bit of skull or dura at the top in one place, but it's not big at all. This is a 3T MPRAGE scan.
- Diandra
On Fri, Nov 19, 2010 at 5:01 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote:
wow, that is a ton of defects. The surface also looks pretty big. Is the cerebellum or skull attached? What kind of scan is it?
Bruce
On Fri, 19 Nov 2010, Diandra Lucia wrote:
Hi all,
I've searched the archives a bit, and found only one truly similar error which seems like it was never answered (that I could see, anyway).
The case we're trying to process stopped during mris_fix_topology at this point:
CORRECTING DEFECT 50 (vertices=22858, convex hull=4870) normal vector of length zero at vertex 153186 with 0 faces vertex 153186 has 0 face Numerical result out of range
and then recon-all exits with errors.
I checked the euler number, which seems high: eno=-988 (nv=200584, nf=403144, ne=604716)
and the surface does report a very large number of defects: 166466 ambiguous faces found in tessellation segmenting defects... 185 defects found, arbitrating ambiguous regions...
but from what I can see on orig.nofix, nothing seems really grossly wrong (i.e. no major skullstripping errors); from the defects labels, most of the defects seem to be in the sulci.
I can send along the full logs if that would be helpful. Any advice?
I'm using freesurfer-Linux-rh9-stable-pub-v4.2.0
Thank you, Diandra
do you want to send us this one to look at?
Bruce On Mon, 22 Nov 2010, Diandra Lucia wrote:
Hi Bruce,
No cerebellum attached in the *.nofix files; maybe a tiny bit of skull or dura at the top in one place, but it's not big at all. This is a 3T MPRAGE scan.
- Diandra
On Fri, Nov 19, 2010 at 5:01 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote:
wow, that is a ton of defects. The surface also looks pretty big. Is the cerebellum or skull attached? What kind of scan is it?
Bruce
On Fri, 19 Nov 2010, Diandra Lucia wrote:
Hi all,
I've searched the archives a bit, and found only one truly similar error which seems like it was never answered (that I could see, anyway).
The case we're trying to process stopped during mris_fix_topology at this point:
CORRECTING DEFECT 50 (vertices=22858, convex hull=4870) normal vector of length zero at vertex 153186 with 0 faces vertex 153186 has 0 face Numerical result out of range
and then recon-all exits with errors.
I checked the euler number, which seems high: eno=-988 (nv=200584, nf=403144, ne=604716)
and the surface does report a very large number of defects: 166466 ambiguous faces found in tessellation segmenting defects... 185 defects found, arbitrating ambiguous regions...
but from what I can see on orig.nofix, nothing seems really grossly wrong (i.e. no major skullstripping errors); from the defects labels, most of the defects seem to be in the sulci.
I can send along the full logs if that would be helpful. Any advice?
I'm using freesurfer-Linux-rh9-stable-pub-v4.2.0
Thank you, Diandra
Sure, will do. Thanks! - Diandra
On Mon, Nov 22, 2010 at 12:06 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote:
do you want to send us this one to look at?
Bruce On Mon, 22 Nov 2010, Diandra Lucia wrote:
Hi Bruce,
No cerebellum attached in the *.nofix files; maybe a tiny bit of skull or dura at the top in one place, but it's not big at all. This is a 3T MPRAGE scan.
- Diandra
On Fri, Nov 19, 2010 at 5:01 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote:
wow, that is a ton of defects. The surface also looks pretty big. Is the cerebellum or skull attached? What kind of scan is it?
Bruce
On Fri, 19 Nov 2010, Diandra Lucia wrote:
Hi all,
I've searched the archives a bit, and found only one truly similar error which seems like it was never answered (that I could see, anyway).
The case we're trying to process stopped during mris_fix_topology at this point:
CORRECTING DEFECT 50 (vertices=22858, convex hull=4870) normal vector of length zero at vertex 153186 with 0 faces vertex 153186 has 0 face Numerical result out of range
and then recon-all exits with errors.
I checked the euler number, which seems high: eno=-988 (nv=200584, nf=403144, ne=604716)
and the surface does report a very large number of defects: 166466 ambiguous faces found in tessellation segmenting defects... 185 defects found, arbitrating ambiguous regions...
but from what I can see on orig.nofix, nothing seems really grossly wrong (i.e. no major skullstripping errors); from the defects labels, most of the defects seem to be in the sulci.
I can send along the full logs if that would be helpful. Any advice?
I'm using freesurfer-Linux-rh9-stable-pub-v4.2.0
Thank you, Diandra
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Diandra Lucia