Automatic segmentation of left atrial scar from delayed-enhancement magnetic resonance imaging

Rashed Karim, Aruna Arujuna, Alex Brazier, Jaswinder Gill, C. Aldo Rinaldi, Mark O'Neill, Reza Razavi, Tobias Schaeffter, Daniel Rueckert, Kawal S. Rhode

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

9 Scopus citations

Abstract

Delayed-enhancement magnetic resonance imaging is an effective technique for imaging left atrial (LA) scars both pre- and post- radio-frequency ablation for the treatment of atrial fibrillation. Existing techniques for LA scar segmentation require expert manual interaction making them tedious and prone to high observer variability. In this paper, we propose a novel automatic segmentation algorithm for segmenting LA scar based on a probabilistic tissue intensity model. This is implemented as a Markov random field-based energy formulation and solved using graph-cuts. It was evaluated against an existing semi-automatic approach and expert manual segmentations using 9 patient data sets. Surface representations were used to compare the methods. The segmented LA scar was expressed as a percentage of the total LA surface. Statistical analysis showed that the novel algorithm was not significantly different to the manual method and that it compared more favorably with this than the semi-automatic approach.

Original languageEnglish
Title of host publicationFunctional Imaging and Modeling of the Heart - 6th International Conference, FIMH 2011, Proceedings
Pages63-70
Number of pages8
DOIs
StatePublished - 2011
Externally publishedYes
Event6th International Conference on Functional Imaging and Modeling of the Heart, FIMH 2011 - New York City, NY, United States
Duration: 25 May 201127 May 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6666 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference6th International Conference on Functional Imaging and Modeling of the Heart, FIMH 2011
Country/TerritoryUnited States
CityNew York City, NY
Period25/05/1127/05/11

Keywords

  • Markov random fields
  • atrial fibrillation
  • delayed enhancement MRI
  • graph-cuts
  • scar segmentation

Fingerprint

Dive into the research topics of 'Automatic segmentation of left atrial scar from delayed-enhancement magnetic resonance imaging'. Together they form a unique fingerprint.

Cite this