Branching medial models for cardiac shape representation

Hui Sun*, Brian B. Avants, Alejandro F. Frangi, Sebastian Ordas, James C. Gee, Paul A. Yushkevich

*Corresponding author for this work

Research output: Chapter in Book/Conference proceedingConference contributionpeer-review

Abstract

The cm-rep (continuous medial representation) is a powerful shape representation method that models a 3D object by describing its medial axis (skeleton) and boundary as continuous parametric manifolds. It provides parametrization of the entire interior of the object, which can be used for combined statistical analysis of shape and appearance. This paper extends the cm-rep to more complex shapes with multi-figures, i.e., shapes whose skeletons have branches. Along the branching curves, the equality constraints enforced by the medial geometry are implemented as soft penalties in the deformable model. The remaining small violations are corrected by local adjustments. As a proof of concept, the branching continuous medial representation is applied to a 2-chamber heart model data set consisting of 428 cardiac shapes from 90 subjects. The results show that our model can capture the heart shape accurately.

Original languageEnglish
Title of host publication2008 5th IEEE International Symposium on Biomedical Imaging
Subtitle of host publicationFrom Nano to Macro, Proceedings, ISBI
Pages1485-1488
Number of pages4
DOIs
Publication statusPublished - 2008
Event2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI - Paris, France
Duration: 14 May 200817 May 2008

Publication series

Name2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, Proceedings, ISBI

Conference

Conference2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI
Country/TerritoryFrance
CityParis
Period14/05/0817/05/08

Keywords

  • 2-chamber heart model
  • Branching Medial Model
  • Cardiac Thickness Analysis
  • Medial Axis

Fingerprint

Dive into the research topics of 'Branching medial models for cardiac shape representation'. Together they form a unique fingerprint.

Cite this