Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Region segmentation using information divergence measures.

Lyndon S Hibbard1

  • 1Research, CMS, Inc., 1145 Corporate Lake Drive, St. Louis, MO 63132, USA. lyn@cmsrtp.com

Medical Image Analysis
|September 29, 2004
PubMed
Summary

Model-free image segmentation using information theory, specifically Jensen-Rényi divergence (JRD), improves contouring accuracy for multiple objects. This method enhances radiotherapy treatment planning by accurately segmenting patient anatomy in X-ray computed tomography.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Anatomy structure creation and editing using 3D implicit surfaces.

Medical physics·2012
Same author

Clinical validation of atlas-based auto-segmentation of multiple target volumes and normal tissue (swallowing/mastication) structures in the head and neck.

International journal of radiation oncology, biology, physics·2010
Same author

Atlas-based auto-segmentation of head and neck CT images.

Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention·2008
See all related articles

Area of Science:

  • Medical imaging analysis
  • Information theory applications
  • Computational anatomy

Background:

  • Traditional image segmentation relies on parametric models, which can be inaccurate for real-world image data.
  • Model-free inference methods offer potential advantages by avoiding assumptions about feature distributions.
  • Relative entropy (RE) is an information-theoretic measure useful for model-free inference and boundary detection.

Purpose of the Study:

  • To introduce and evaluate a generalized relative entropy method, Jensen-Rényi divergence (JRD), for model-free image segmentation.
  • To demonstrate JRD's capability in contouring multiple objects simultaneously and improving segmentation accuracy.
  • To apply JRD-based segmentation to contour patient anatomy in X-ray computed tomography for radiotherapy planning.

Main Methods:

Related Experiment Videos

  • Utilized Jensen-Rényi divergence (JRD), a generalization of relative entropy, for optimal n-way decisions in image segmentation.
  • Developed an edge detector based on JRD combined with multivariate pixel segmentation.
  • Applied the JRD functions to segment patient anatomy from X-ray computed tomography images.

Main Results:

  • The JRD-based edge detector, coupled with multivariate pixel segmentation, generally improved segmentation error.
  • The JRD method demonstrated the ability to contour multiple objects simultaneously without significant contour overlap.
  • Seed regions expanded naturally, facilitating robust object contouring.

Conclusions:

  • Jensen-Rényi divergence provides an effective model-free approach for accurate image segmentation, particularly for multiple objects.
  • This method shows promise for enhancing radiotherapy treatment planning through precise patient anatomy contouring in medical imaging.
  • The JRD approach offers a robust alternative to model-dependent segmentation techniques in complex imaging scenarios.