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

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

9 Zitate (Scopus)

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.

OriginalspracheEnglisch
TitelFunctional Imaging and Modeling of the Heart - 6th International Conference, FIMH 2011, Proceedings
Seiten63-70
Seitenumfang8
DOIs
PublikationsstatusVeröffentlicht - 2011
Extern publiziertJa
Veranstaltung6th International Conference on Functional Imaging and Modeling of the Heart, FIMH 2011 - New York City, NY, USA/Vereinigte Staaten
Dauer: 25 Mai 201127 Mai 2011

Publikationsreihe

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band6666 LNCS
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz

Konferenz6th International Conference on Functional Imaging and Modeling of the Heart, FIMH 2011
Land/GebietUSA/Vereinigte Staaten
OrtNew York City, NY
Zeitraum25/05/1127/05/11

Fingerprint

Untersuchen Sie die Forschungsthemen von „Automatic segmentation of left atrial scar from delayed-enhancement magnetic resonance imaging“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren