Fully automated contour detection of the ascending aorta in cardiac 2D phase-contrast MRI

Marina Codari, Marco Scarabello, Francesco Secchi, Chiarella Sforza, Giuseppe Baselli, Francesco Sardanelli

Research output: Contribution to journalArticle

Abstract

Purpose In this study we proposed a fully automated method for localizing and segmenting the ascending aortic lumen with phase-contrast magnetic resonance imaging (PC-MRI). Material and methods Twenty-five phase-contrast series were randomly selected out of a large population dataset of patients whose cardiac MRI examination, performed from September 2008 to October 2013, was unremarkable. The local Ethical Committee approved this retrospective study. The ascending aorta was automatically identified on each phase of the cardiac cycle using a priori knowledge of aortic geometry. The frame that maximized the area, eccentricity, and solidity parameters was chosen for unsupervised initialization. Aortic segmentation was performed on each frame using active contouring without edges techniques. The entire algorithm was developed using Matlab R2016b. To validate the proposed method, the manual segmentation performed by a highly experienced operator was used. Dice similarity coefficient, Bland-Altman analysis, and Pearson's correlation coefficient were used as performance metrics. Results Comparing automated and manual segmentation of the aortic lumen on 714 images, Bland-Altman analysis showed a bias of − 6.68 mm2, a coefficient of repeatability of 91.22 mm2, a mean area measurement of 581.40 mm2, and a reproducibility of 85%. Automated and manual segmentation were highly correlated (R = 0.98). The Dice similarity coefficient versus the manual reference standard was 94.6 ± 2.1% (mean ± standard deviation). Conclusion A fully automated and robust method for identification and segmentation of ascending aorta on PC-MRI was developed. Its application on patients with a variety of pathologic conditions is advisable.

Original languageEnglish
Pages (from-to)77-82
Number of pages6
JournalMagnetic Resonance Imaging
Volume47
DOIs
Publication statusPublished - Apr 1 2018

Fingerprint

Magnetic resonance
Magnetic resonance imaging
Aorta
Imaging techniques
Magnetic Resonance Imaging
Geometry
Retrospective Studies
Population

Keywords

  • Aorta
  • Computer-assisted image processing
  • Magnetic resonance imaging

ASJC Scopus subject areas

  • Biophysics
  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging

Cite this

Fully automated contour detection of the ascending aorta in cardiac 2D phase-contrast MRI. / Codari, Marina; Scarabello, Marco; Secchi, Francesco; Sforza, Chiarella; Baselli, Giuseppe; Sardanelli, Francesco.

In: Magnetic Resonance Imaging, Vol. 47, 01.04.2018, p. 77-82.

Research output: Contribution to journalArticle

@article{bf9fc437761b4a67b2cea9f94124ab0d,
title = "Fully automated contour detection of the ascending aorta in cardiac 2D phase-contrast MRI",
abstract = "Purpose In this study we proposed a fully automated method for localizing and segmenting the ascending aortic lumen with phase-contrast magnetic resonance imaging (PC-MRI). Material and methods Twenty-five phase-contrast series were randomly selected out of a large population dataset of patients whose cardiac MRI examination, performed from September 2008 to October 2013, was unremarkable. The local Ethical Committee approved this retrospective study. The ascending aorta was automatically identified on each phase of the cardiac cycle using a priori knowledge of aortic geometry. The frame that maximized the area, eccentricity, and solidity parameters was chosen for unsupervised initialization. Aortic segmentation was performed on each frame using active contouring without edges techniques. The entire algorithm was developed using Matlab R2016b. To validate the proposed method, the manual segmentation performed by a highly experienced operator was used. Dice similarity coefficient, Bland-Altman analysis, and Pearson's correlation coefficient were used as performance metrics. Results Comparing automated and manual segmentation of the aortic lumen on 714 images, Bland-Altman analysis showed a bias of − 6.68 mm2, a coefficient of repeatability of 91.22 mm2, a mean area measurement of 581.40 mm2, and a reproducibility of 85{\%}. Automated and manual segmentation were highly correlated (R = 0.98). The Dice similarity coefficient versus the manual reference standard was 94.6 ± 2.1{\%} (mean ± standard deviation). Conclusion A fully automated and robust method for identification and segmentation of ascending aorta on PC-MRI was developed. Its application on patients with a variety of pathologic conditions is advisable.",
keywords = "Aorta, Computer-assisted image processing, Magnetic resonance imaging",
author = "Marina Codari and Marco Scarabello and Francesco Secchi and Chiarella Sforza and Giuseppe Baselli and Francesco Sardanelli",
year = "2018",
month = "4",
day = "1",
doi = "10.1016/j.mri.2017.11.010",
language = "English",
volume = "47",
pages = "77--82",
journal = "Magnetic Resonance Imaging",
issn = "0730-725X",
publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - Fully automated contour detection of the ascending aorta in cardiac 2D phase-contrast MRI

AU - Codari, Marina

AU - Scarabello, Marco

AU - Secchi, Francesco

AU - Sforza, Chiarella

AU - Baselli, Giuseppe

AU - Sardanelli, Francesco

PY - 2018/4/1

Y1 - 2018/4/1

N2 - Purpose In this study we proposed a fully automated method for localizing and segmenting the ascending aortic lumen with phase-contrast magnetic resonance imaging (PC-MRI). Material and methods Twenty-five phase-contrast series were randomly selected out of a large population dataset of patients whose cardiac MRI examination, performed from September 2008 to October 2013, was unremarkable. The local Ethical Committee approved this retrospective study. The ascending aorta was automatically identified on each phase of the cardiac cycle using a priori knowledge of aortic geometry. The frame that maximized the area, eccentricity, and solidity parameters was chosen for unsupervised initialization. Aortic segmentation was performed on each frame using active contouring without edges techniques. The entire algorithm was developed using Matlab R2016b. To validate the proposed method, the manual segmentation performed by a highly experienced operator was used. Dice similarity coefficient, Bland-Altman analysis, and Pearson's correlation coefficient were used as performance metrics. Results Comparing automated and manual segmentation of the aortic lumen on 714 images, Bland-Altman analysis showed a bias of − 6.68 mm2, a coefficient of repeatability of 91.22 mm2, a mean area measurement of 581.40 mm2, and a reproducibility of 85%. Automated and manual segmentation were highly correlated (R = 0.98). The Dice similarity coefficient versus the manual reference standard was 94.6 ± 2.1% (mean ± standard deviation). Conclusion A fully automated and robust method for identification and segmentation of ascending aorta on PC-MRI was developed. Its application on patients with a variety of pathologic conditions is advisable.

AB - Purpose In this study we proposed a fully automated method for localizing and segmenting the ascending aortic lumen with phase-contrast magnetic resonance imaging (PC-MRI). Material and methods Twenty-five phase-contrast series were randomly selected out of a large population dataset of patients whose cardiac MRI examination, performed from September 2008 to October 2013, was unremarkable. The local Ethical Committee approved this retrospective study. The ascending aorta was automatically identified on each phase of the cardiac cycle using a priori knowledge of aortic geometry. The frame that maximized the area, eccentricity, and solidity parameters was chosen for unsupervised initialization. Aortic segmentation was performed on each frame using active contouring without edges techniques. The entire algorithm was developed using Matlab R2016b. To validate the proposed method, the manual segmentation performed by a highly experienced operator was used. Dice similarity coefficient, Bland-Altman analysis, and Pearson's correlation coefficient were used as performance metrics. Results Comparing automated and manual segmentation of the aortic lumen on 714 images, Bland-Altman analysis showed a bias of − 6.68 mm2, a coefficient of repeatability of 91.22 mm2, a mean area measurement of 581.40 mm2, and a reproducibility of 85%. Automated and manual segmentation were highly correlated (R = 0.98). The Dice similarity coefficient versus the manual reference standard was 94.6 ± 2.1% (mean ± standard deviation). Conclusion A fully automated and robust method for identification and segmentation of ascending aorta on PC-MRI was developed. Its application on patients with a variety of pathologic conditions is advisable.

KW - Aorta

KW - Computer-assisted image processing

KW - Magnetic resonance imaging

UR - http://www.scopus.com/inward/record.url?scp=85036671064&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85036671064&partnerID=8YFLogxK

U2 - 10.1016/j.mri.2017.11.010

DO - 10.1016/j.mri.2017.11.010

M3 - Article

C2 - 29180100

AN - SCOPUS:85036671064

VL - 47

SP - 77

EP - 82

JO - Magnetic Resonance Imaging

JF - Magnetic Resonance Imaging

SN - 0730-725X

ER -