Skip to main navigation Skip to search Skip to main content

Optimizing computed tomographic angiography image segmentation using Fitness Based Partitioning

  • Jeroen Eggermont*
  • , Rui Li
  • , Ernst G.P. Bovenkamp
  • , Henk Marquering
  • , Michael T.M. Emmerich
  • , Aad Van Der Lugt
  • , Thomas Bäck
  • , Jouke Dijkstra
  • , Johan H.C. Reiber
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Computed Tomographic Angiography (CTA) has become a popular image modality for the evaluation of arteries and the detection of narrowings. For an objective and reproducible assessment of objects in CTA images, automated segmentation is very important. However, because of the complexity of CTA images it is not possible to find a single parameter setting that results in an optimal segmentation for each possible image of each possible patient. Therefore, we want to find optimal parameter settings for different CTA images. In this paper we investigate the use of Fitness Based Partitioning to find groups of images that require a similar parameter setting for the segmentation algorithm while at the same time evolving optimal parameter settings for these groups. The results show that Fitness Based Partitioning results in better image segmentation than the original default parameter solutions or a single parameter solution evolved for all images.

Original languageEnglish
Title of host publicationApplications of Evolutionary Computing - EvoWorkshops 2008
Subtitle of host publicationEvoCOMNET, EvoFIN, EvoHOT, EvoIASP, EvoMUSART, EvoNUM, EvoSTOC, and EvoTransLog, Proceedings
Pages275-284
Number of pages10
DOIs
Publication statusPublished - 2008
EventEuropean Workshops on the Theory and Applications of Evolutionary Computation, EvoWorkshops 2008: EvoCOMNET, EvoFIN, EvoHOT, EvoIASP, EvoMUSART, EvoNUM, EvoSTOC, and EvoTransLog - Naples, Italy
Duration: 26 Mar 200828 Mar 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4974 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceEuropean Workshops on the Theory and Applications of Evolutionary Computation, EvoWorkshops 2008: EvoCOMNET, EvoFIN, EvoHOT, EvoIASP, EvoMUSART, EvoNUM, EvoSTOC, and EvoTransLog
Country/TerritoryItaly
CityNaples
Period26/03/200828/03/2008

Fingerprint

Dive into the research topics of 'Optimizing computed tomographic angiography image segmentation using Fitness Based Partitioning'. Together they form a unique fingerprint.

Cite this