<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Am 21.10.2010 19:17, schrieb Torsten Rohlfing:
<blockquote cite="mid:4CC075AB.6050601@synapse.sri.com" type="cite">
<meta http-equiv="content-type" content="text/html;
charset=ISO-8859-1">
Hi guys -<br>
<br>
This is more of a question than a suggestion: does ITK currently
support "pixel padding"? That is, can we currently define a value
in the image data that serves as a marker for pixels that aren't
really there? Think of it as a binary mask coded into the image
itself.<br>
<br>
In my experience, when you register partial objects to complete
objects (also interesting to register skull-stripped brains to
full head images, for example), the key is to be able to exclude
pixels from the metric computation altogether, rather than just
setting them to zero.<br>
<br>
Alternatively, because we cannot always find a pixel value that
isn't actually used by the data already, one could implement this
behaviour by providing the registration framework with optional
explicit binary masks for fixed and moving images.<br>
<br>
Best,<br>
Torsten<br>
<br>
<blockquote type="cite">
<pre>Hello Erik
You may want to try changing the number of resolutions employed in the
multi-resolution pyramid used to perform the registration. In
addition, you may want to try switching the role of the fixed/moving
image. We are currently revising the registration framework so now is
a good time to bring these types of issues up and work with us on
improving registration performance.
Brian
On Thu, Oct 21, 2010 at 4:35 AM, Erik Tuerke <<a moz-do-not-send="true" href="http://www.itk.org/mailman/listinfo/insight-users">tuerke at cbs.mpg.de</a>> wrote:
><i> Hi!
</i>><i>
</i>><i> I want to register one hemisphere (actually 30 slices of 0.75 mm) to a
</i>><i> fullbrain scan (30 slices of 0.75mm, but shifted to superior for about 5 mm)
</i>><i> . Currently i am using a MattesMutualInformationImageToImage metric, a
</i>><i> RegularStepGradientDescentOptimizer and a VersorRigid3DTransform.
</i>><i>
</i>><i> Prior to the registration the hemisphere is roughly aligned to the fullbrain
</i>><i> (only needs translation of few mm and little rotation). If i start the
</i>><i> registration the optimizer finishes with a rotation of about 90 deg x and y.
</i>><i> I also tried a VersorRigid3DOptimizer and a NormalizedCorrelation with
</i>><i> almost the same result.
</i>><i>
</i>><i> The scalefactors of my optimizer are:
</i>><i> optimizerScaleRegularStepGradient[0] = 1.0;
</i>><i> optimizerScaleRegularStepGradient[1] = 1.0;
</i>><i> optimizerScaleRegularStepGradient[2] = 1.0;
</i>><i> optimizerScaleRegularStepGradient[3] = 1.0/1000.0;
</i>><i> optimizerScaleRegularStepGradient[4] = 1.0/1000.0;
</i>><i> optimizerScaleRegularStepGradient[5] = 1.0/1000.0;
</i>><i>
</i>><i> I guess the itk examples are proposing the same values.
</i>><i> Is there any trick to apply such a sub-brain registration to a full-brain
</i>><i> registration. The worrying thing is that SPM gets it right :-(
</i>><i>
</i>><i> Thanks and best regards!!
</i>><i>
</i>><i> --
</i>><i> Erik Tuerke
</i>><i> Department of Neurophysics Max-Planck-Institute for Human Cognitive and
</i>><i> Brain Sciences
</i>><i> Stephanstrasse 1A
</i>><i> 04103 Leipzig
</i>><i> Germany Tel: +49 341 99 40-2440
</i>><i> Email: <a moz-do-not-send="true" href="http://www.itk.org/mailman/listinfo/insight-users">tuerke at cbs.mpg.de</a>
</i>><i> <a moz-do-not-send="true" class="moz-txt-link-abbreviated" href="http://www.cbs.mpg.de">www.cbs.mpg.de</a>
</i>><i>
</i>><i>
</i>><i> _____________________________________
</i>><i> Powered by <a moz-do-not-send="true" class="moz-txt-link-abbreviated" href="http://www.kitware.com">www.kitware.com</a>
</i>><i>
</i>><i> Visit other Kitware open-source projects at
</i>><i> <a moz-do-not-send="true" href="http://www.kitware.com/opensource/opensource.html">http://www.kitware.com/opensource/opensource.html</a>
</i>><i>
</i>><i> Kitware offers ITK Training Courses, for more information visit:
</i>><i> <a moz-do-not-send="true" href="http://www.kitware.com/products/protraining.html">http://www.kitware.com/products/protraining.html</a>
</i>><i>
</i>><i> Please keep messages on-topic and check the ITK FAQ at:
</i>><i> <a moz-do-not-send="true" href="http://www.itk.org/Wiki/ITK_FAQ">http://www.itk.org/Wiki/ITK_FAQ</a>
</i>><i>
</i>><i> Follow this link to subscribe/unsubscribe:
</i>><i> <a moz-do-not-send="true" href="http://www.itk.org/mailman/listinfo/insight-users">http://www.itk.org/mailman/listinfo/insight-users</a>
</i>><i>
</i>
</pre>
</blockquote>
<br>
<pre class="moz-signature" cols="72">--
Torsten Rohlfing, PhD SRI International, Neuroscience Program
Senior Research Scientist 333 Ravenswood Ave, Menlo Park, CA 94025
Phone: ++1 (650) 859-3379 Fax: ++1 (650) 859-2743
<a moz-do-not-send="true" class="moz-txt-link-abbreviated" href="mailto:torsten@synapse.sri.com">torsten@synapse.sri.com</a> <a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://www.stanford.edu/%7Erohlfing/">http://www.stanford.edu/~rohlfing/</a>
"Though this be madness, yet there is a method in't"
</pre>
<pre wrap="">
<fieldset class="mimeAttachmentHeader"></fieldset>
_____________________________________
Powered by <a class="moz-txt-link-abbreviated" href="http://www.kitware.com">www.kitware.com</a>
Visit other Kitware open-source projects at
<a class="moz-txt-link-freetext" href="http://www.kitware.com/opensource/opensource.html">http://www.kitware.com/opensource/opensource.html</a>
Kitware offers ITK Training Courses, for more information visit:
<a class="moz-txt-link-freetext" href="http://www.kitware.com/products/protraining.html">http://www.kitware.com/products/protraining.html</a>
Please keep messages on-topic and check the ITK FAQ at:
<a class="moz-txt-link-freetext" href="http://www.itk.org/Wiki/ITK_FAQ">http://www.itk.org/Wiki/ITK_FAQ</a>
Follow this link to subscribe/unsubscribe:
<a class="moz-txt-link-freetext" href="http://www.itk.org/mailman/listinfo/insight-users">http://www.itk.org/mailman/listinfo/insight-users</a>
</pre>
</blockquote>
Hi!<br>
<br>
Well i think extracting a region of interest which should be
considered during the registration is possible by passing a mask to
the image metric. So the challenge is to define this mask
automatically. I am first thinking of an Otsu filter to define such
a mask.<br>
A second possibility would be to use the "<a class="el"
href="http://www.itk.org/Doxygen318/html/classitk_1_1ImageToImageMetric.html#ae56734d7cc4bb45a6ba462e56f260546">SetUseFixedImageSamplesIntensityThreshold</a>"
method or the "SetUseSequentialSampling" method of the
MattesMutualInformationImageToImageMetric. But since these both
methods are only implemented in the review code i did not use one of
them.<br>
<br>
Best regards so far!<br>
<br>
<pre class="moz-signature" cols="72">--
Erik Türke
Department of Neurophysics
Max-Planck-Institute for Human Cognitive and Brain Sciences
Stephanstrasse 1A
04103 Leipzig
Germany
Tel: +49 341 99 40-2440
Email: <a class="moz-txt-link-abbreviated" href="mailto:tuerke@cbs.mpg.de">tuerke@cbs.mpg.de</a>
<a class="moz-txt-link-abbreviated" href="http://www.cbs.mpg.de">www.cbs.mpg.de</a> </pre>
</body>
</html>