Your message dated Thu, 15 Jul 2010 09:47:38 -0400
with message-id <20100715134738.gr16...@onerussian.com>
has caused the report #589166,
regarding ants: --rigid-affine has unexpected effects
to be marked as having been forwarded to the upstream software
author(s) Brian Avants <stn...@gmail.com>
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact ow...@bugs.debian.org
immediately.)
--
589166: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=589166
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Hi Brian,
Could you have a look at the weirdness Michael spotted?
do you think current trunk could for some reason be more correct? or may
be invocation is somewhat incorrect?
actual images are available from original bug report page:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=589166
Thanks in advance!
----- Forwarded message from Michael Hanke <michael.ha...@gmail.com> -----
Date: Thu, 15 Jul 2010 09:36:24 -0400
From: Michael Hanke <michael.ha...@gmail.com>
To: Debian Bug Tracking System <sub...@bugs.debian.org>
Subject: Bug#589166: ants: --rigid-affine has unexpected effects
X-CRM114-Status: Good ( pR: 39.4955 )
Package: ants
Version: 1.9+svn532-2~squeeze.nd1
Severity: normal
Hi,
I'm attempting an affine transformation between two MR images of the
same brain -- moving: EPI; fixed: T1_anatomical.
I run somthing like this:
ANTS 3 -m MI[ants_fixed.nii.gz,ants_epi.nii.gz,1,64] \
-x ants_fixed_mask.nii.gz \
-i 0 \
-o ants_something
It yields a result that is somewhat aligned (brain outline matches,
etc.), but is overall pretty bad. Since these are images of the same
brain, I wanted to try a rigid affine transformation to see whether that
helps. That made it even worse. The call stays the same, but I added
--rigid-affine true
BTW: what is the difference to --do-rigid? Looks like an alias, but
isn't labeled as such...
The supposedly rigid body transformation yields a substantial offset in
comparison to the previous affine and also in comparison to the target
image. The brain outline doesn't match at all. I'm attaching a
screenshot to illustrate that.
gray -- fixed image, barely visible
yellow -- mask of the fixed image
red -- moving image after applying affine transformation (somewhat
matches to shape of the fixed image)
green -- moving image after applying rigid body transformation
(significant offset)
I suspect I'm doing something wrong. I played with other settings
mentioned in the documentation, such as
- switching metric to PR
- --MI-option 64x64000
- --number-of-affine-iterations 10000x10000x10000
In all combinations I tried I get the offset once I include
--rigid-affine true in the call.
Attached are the two affine transformations the were used to create the
screenshot and the screenshot itself. I can provide the actual images if
that helps.
Thanks,
Michael
--
GPG key: 1024D/3144BE0F Michael Hanke
http://mih.voxindeserto.de
#Insight Transform File V1.0
# Transform 0
Transform: MatrixOffsetTransformBase_double_3_3
Parameters: 1.07637 -0.0374557 0.0104446 0.0534896 1.04295 0.0614291
-0.00548789 -0.0154716 1.05312 -0.0521075 -10.5199 16.5958
FixedParameters: 3.56699 6.79132 -5.53247
#Insight Transform File V1.0
# Transform 0
Transform: MatrixOffsetTransformBase_double_3_3
Parameters: 0.999612 -0.0148503 0.0235775 0.0113349 0.989688 0.142791
-0.0254549 -0.142469 0.989472 -0.46153 18.5183 15.7682
FixedParameters: 3.56699 6.79132 -5.53247
----- End forwarded message -----
--
Yaroslav O. Halchenko
Postdoctoral Fellow, Department of Psychological and Brain Sciences
Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755
Phone: +1 (603) 646-9834 Fax: +1 (603) 646-1419
WWW: http://www.linkedin.com/in/yarik
--- End Message ---
