POSTER PRESENTATION Open Access

A software tool for segmentation of themyocardium in CMR exams using level-setalgorithmAzadeh Nazemoroaya, Abbas Nasiraei-Moghaddam*

From 19th Annual SCMR Scientific SessionsLos Angeles, CA, USA. 27-30 January 2016

BackgroundIn many Cardiac MR Exams, we need to extract the myo-cardium area precisely, which is a labor-intensive task ifperformed manually. Furthermore, the manual segmenta-tion is a source of error since it highly depends on theuser experience. Several methods, therefore, have beenintroduced to facilitate and expedite this time-consumingprocedure [1-3].We have implemented a software tool inprocessing of cardiac tagging images and calculation ofstrain, which is sensitive to segmentation especially inradial direction. Compared to other methods, such asactive contours and snakes, “level-set” is usually less sen-sitive to the initialization and also can be splitted auto-matically and quickly to detect more than one connectedcomponent [4]. Hence, we used level-set as the core ofour user-friendly software to make it robust to initializa-tion and segmentation of MR images. The possibility ofintensity inhomogeneity was also considered.

MethodsThe level-set framework has been exploited in order todo curve evolution to extract endocardium and epicar-dium. The Algorithm includes definition of an energy ofthe level-set functions and a bias field that accounts forthe intensity inhomogeneity (ƒ (�, c, b)). Segmentationis performed by minimizing the energy based on gradi-ent flow equation, while updated ci (distinct constantsin disjoint regions of image Ωi) and b (bias field) aregiven as input parameters from previous iteration. Inthis implementation user acts as supervisor for the algo-rithm only for the first frame by applying some minorchanges on the calculated contours. Thereafter bordersof the myocardium in subsequent frames are calculated

using the anatomical and dynamical information deter-mined in the previous frame. To examine the accuracyand robustness of the software, the myocardium wasextracted both manually and automatically in short axisview of 120 balanced SSFP MR images related to 4 ser-ies of data from healthy people.

ResultsFigure 1 shows a sample output of the software. Weevaluated the algorithm by calculating the mean sensi-tivity and the mean specificity, which were 0.918 and0.959, respectively. The total time needed for extractingmyocardium for each dataset was dramatically decreasedby a factor of 7. The software coding was carried outentirely in MATLAB 2013a environment and the pro-cessor of used computer was 5th generation Intel®

Core™ i5 processors.

ConclusionsA software tool for semi-automatic segmentation of themyocardium in MR images was developed, which needsno user interaction except for minor changes in the firstframe. Validation has been performed on real data fromhealthy people. A study of segmentation errors by calcu-lating the specificity and sensitivity shows satisfactoryresults. Also the time needed for extracting myocardiumreduced dramatically.

Published: 27 January 2016

References1. Niessen, et al: IEEE Int Symp Biomedical Imaging, Washington DC,USA 2002.2. Marwa, et al: AJBME 2012.3. Cocosco, et al: JMRI 2008.4. Li Chunming, et al: IEEE TIP 2011.

Amirkabir University of Technology, Tehran, Iran (the Islamic Republic of)

Nazemoroaya and Nasiraei-Moghaddam Journal of CardiovascularMagnetic Resonance 2016, 18(Suppl 1):P113http://www.jcmr-online.com/content/18/S1/P113

© 2016 Nazemoroaya and Nasiraei-Moghaddam This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

doi:10.1186/1532-429X-18-S1-P113Cite this article as: Nazemoroaya and Nasiraei-Moghaddam: A softwaretool for segmentation of the myocardium in CMR exams using level-set algorithm. Journal of Cardiovascular Magnetic Resonance 2016 18(Suppl 1):P113.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

Figure 1 A sample output of the software for the phase 1 out of 30.

Nazemoroaya and Nasiraei-Moghaddam Journal of CardiovascularMagnetic Resonance 2016, 18(Suppl 1):P113http://www.jcmr-online.com/content/18/S1/P113

Page 2 of 2