On Thu, Jul 12, 2012 at 4:07 PM, Peter McNally <pmcnally@haverford.edu> wrote:

Hi Bruce,

It appears as if the ?h.orig.nofix file has properly delimited the white matter surface, where the ?h.orig file has not. What makes me dubious of there being a topological issue at hand is that I have already tried out the typical fixes for topological issues, and they've had no positive impact on the problem. Also, what makes me particularly confused is that this issue has cropped up on many edited files that I've run autorecon2 and autorecon3 on. To be honest, I don't have a good idea of where to go next....

I've uploaded an example subject (look at the anterior section of the temporal lobes for the worst areas) as FI137_PreFixes.tar.gz and FI137_PostFixes.tar.gz (pre and post fixes respectively).

Best,

On Thu, Jul 12, 2012 at 3:29 PM, Bruce Fischl <fischl@nmr.mgh.harvard.edu> wrote:
Hi Peter

because you are adding control points to voxels that are already accurately segmented (before you added them) I think your problem is a topological defect that is being fixed incorrectly. You can see this is you load the ?h.orig and it doesn't follow the wm.mgz. Since you don't know what the source of the defect is, it's not clear that making things brighter will help. If it is an erroneous connection then it will actually hurt (which may be what is happening). You need to track down the defect and diagnose it's source.

cheers
Bruce

On Thu, 12 Jul 2012, Peter McNally wrote:

[cleardot.gif]

Hi Bruce,

It is pretty erratic what happens after adding them. Sometimes the wm.mgz is
more accurate and the wm surface is less accurate, but more often they are
both less accurate. The points themselves do typically get brighter (on
wm.mgz).

I've included a screenshot of an example of what's going on as an attachment
(.pdf)

Best,
Peter[cleardot.gif]

On Thu, Jul 12, 2012 at 3:10 PM, Bruce Fischl <fischl@nmr.mgh.harvard.edu>
wrote:
Hi Peter and Mike

yes, 5.2 will undo those changes, although I still wouldn't
expect this behavior. Peter: are the points brighter (closer to
110) after adding them? Is the wm.mgz more accurate?

cheers
Bruce

On Thu, 12 Jul 2012, Michael Harms wrote:

Hi Peter,
This sounds very familiar (including the effect on
the opposing
hemisphere) to something I've observed previously
under FS v5.1 -- see
the posts titled "control point guidance" from late
Jan 2012. This may
be related to a change in how CP's were used in FS
v5.1, which I believe
they are planning on backing out in the next
release.

I've personally always found it much easier to
predict what would occur
using direct edits to the wm.mgz, rather than
control points. And in
this case, wm edits are what we did.

cheers,
-MH

On Thu, 2012-07-12 at 13:15 -0400, Peter McNally
wrote:
Dear FreeSurfer,

Hello, I am currently working on a
project using FreeSurfer (version
5.1.0) at the Kable Labs at the
University of Pennsylvania, and I have
come across a problem when using control
points to add incorporate
more of the white matter into the white
matter surface. After
highlighting the parts of the white
matter that have yet to be brought
into the white matter surface with CPs,
it will sometimes do what I
intend it to but often will not. The
strange part is that when it
fails to incorporate more wm into the wm
surfs it often actually
removes a considerable amount the white
matter both from the wm.mgz
volume and from the white matter
surface. Once it even did so on the
opposite hemisphere as where I had put
down control points. This
problem seems to occur exclusively in
the anterior end of the temporal
lobe (everywhere else my use of control
points has presented me with
no problems), and it happens both when
using control points sparingly
and aggressively (I did a bit of
experimenting, and the quantity of
CPs that I used seemed to make no
difference). Do you have any advice
for how I might be able to get around
this dilemma and/or why it
likely is occurring? Thank you very much
for your time.

Best,
Peter McNally
_______________________________________________
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