Segmentation Integrating Watershed and Shape Priors Applied to Cardiac Delayed Enhancement MR Images

Abstract : Background: In recent years, there has been a rapid rise in the use of shape priors applied to segmentation process of medical images. Previous approaches on left ventricle segmentation from Delayed-Enhancement Magnetic Resonance Imaging (DE-MRI) have focused on the extraction of myocardium or just diseased region in short axis orientation. However these studies did not take into account the segmentation of non-diseased myocardium from DE-MRI. The segmentation of non-diseased myocardium from DE-MRI, has some useful applications. For instance it can simplify the PET-MR registration process. Methods: This paper presents a novel semi-automatic segmentation method of non-diseased myocardium contours from DE-MRI based on watershed algorithm with shape priors application. The segmentation process was performed on long and short axis DE-MR images, acquired from patients with different cardiac diseases. Results: Segmented images were compared with gold standard contours performed by an experienced user. To assess our results the Dice Coefficient (DC) and Root Mean Square Distance (RMSD) were computed. The best value of these parameters was obtained for four cavity images (RMSD = 1.68 +/- 0.47 mm, DC = 0.78), and the computed value for two cavity images (RMSD = 1.93 +/- 0.38 mm, DC = 0.73) and short axis images (RMSD = 1.89 +/- 0.47 mm, DC = 0.71) were slightly lower. Conclusion: In conclusion, We have proposed a novel solution for non-diseased myocardium segmentation from DE-MRI, which brought very promising results. (C) 2017 AGBM. Published by Elsevier Masson SAS. All rights reserved.
Liste complète des métadonnées
Contributeur : Le2i - Université de Bourgogne <>
Soumis le : mardi 31 octobre 2017 - 17:06:51
Dernière modification le : jeudi 11 janvier 2018 - 06:28:16




Dominika Kruk, A. Boucher, A. Lalande, A. Cochet, T. Sliwa. Segmentation Integrating Watershed and Shape Priors Applied to Cardiac Delayed Enhancement MR Images. IRBM, Elsevier Masson, 2017, 38 (4), pp.224 - 227. 〈〉. 〈10.1016/j.irbm.2017.06.004〉. 〈hal-01627049〉



Consultations de la notice