[ITK] Question about the use of LevelSetsv4

Cerrolaza, Juan JCerrola at childrensnational.org
Thu Oct 23 17:53:24 EDT 2014


Dear ITK community,
I recently started to use the new framework ITK provides for Level Sets (LevelSetsv4). It seems to be a pretty nice framework and an ideal way to combine different terms from Level Sets theory (advection terms, curvature-based terms, propagation terms, etc). However, I'm still a bit confused about how to use and configure it.

I would really appreciate if anyone of you could help me to clarify any of these questions.


i)                    Looking into the description of these terms, for instance the curvature term:  'LevelSetEquationCurvatureTerm' (though this is this is also valid for any other term such as the advection term 'LevelSetEquationAdvectionTerm', or the propagation term 'LevelSetEquationPropagationTerm'), it seems like the user has the option to define a specific curvature image (according to the literature, this terms is typically defined as an the result of using any edge detector to the image we are trying to segment. If no curvature image is set, it casts the input image and uses it in the term contribution calculation. My question is how does ITK calculate this term actually. Does it use any edge detector over the input image (such as 1/(1 + G(I)), where I is the input image, and G(.) is the Gaussian filter)?

What happened if both, the input image and the curvature image are defined? Does ITK ignore the first one?



ii)                   The second question is about how to control the propagation direction of the level set. In the example, the level set is initialized as a big square area that contains all the elements to segment. In this scenario, the evolution of the level set should be inward. What if we initialize the level set inside the target structure? In this case the evolution of the level set should be outward. How can we control it? My guess was that the sign of the coefficients that define the weight of each term would control it, but after making different test, I'm not completely sure about it.
Thank you very much in advance.
Juan


Juan J. Cerrolaza PhD

Sheikh Zayed Institute for Pediatric Surgical Innovation
Children's National Health System
111 Michigan Avenue, NW
Washington, DC 20010

E-mail: JCerrola at cnmc.org<mailto:JCerrola at cnmc.org>, jjcerromar at gmail.com<mailto:jjcerromar at gmail.com>
Phone: +1<tel:%2B1-202-476-3059> (202) 476-1272
We stand for children.
www.ChildrensNational.org<http://www.ChildrensNational.org>



,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
Confidentiality Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/community/attachments/20141023/fe34523e/attachment-0002.html>


More information about the Community mailing list