Multi-phase Volume Segmentation with Tetrahedral Mesh

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

200 Downloads (Pure)

Abstract

In life science and material science, it is often desirable to segment a volumetric data set in such a way that multiple materials (phases) are segmented and a tetrahedral mesh representation is obtained for each segment for downstream applications. Unfortunately, obtaining a mesh, typically from CT or MRI scan, is challenging, especially in 3D. This paper proposes an novel approach for volume segmentation using a tetrahedral mesh. Our method employs a deformable model that minimizes the Mumford-Shah energy function. We apply our method to several CT data sets in order to demonstrate its advantages: multi-phase support, robustness to noise, and adaptive resolution outputs. Our method is based on the Deformable Simplicial Complex (DSC) method for tracking deformable interfaces which is designed specifically to deal with topology changes.
Original languageEnglish
Title of host publicationProceedings of British Machine Vision Conference
Number of pages13
PublisherBMVA Press
Publication date2018
Publication statusPublished - 2018
Event29TH British Machine Vision Conference
- Northumbria University, Newcastle upon Tyne , United Kingdom
Duration: 3 Sep 20186 Sep 2018
Conference number: 29

Conference

Conference29TH British Machine Vision Conference
Number29
LocationNorthumbria University
Country/TerritoryUnited Kingdom
CityNewcastle upon Tyne
Period03/09/201806/09/2018
SponsorAmazon.com, Inc., Microsoft USA, Nvidia, SCANs, SCAPE

Fingerprint

Dive into the research topics of 'Multi-phase Volume Segmentation with Tetrahedral Mesh'. Together they form a unique fingerprint.

Cite this