3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network

Dong Yang, Bo Liu, Leon Axel, Dimitri Metaxas

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Cardiac magnetic resonance imaging (MRI) is one of the most useful techniques to understand and measure cardiac functions. Given the MR image data, segmentation of the left ventricle (LV) myocardium is the most common task to be addressed for recovering and studying LV wall motion. However, most of segmentation methods heavily rely on the imaging appearance for extracting the myocardial contours (epi- and endo-cardium). These methods cannot guarantee a consistent volume of the heart wall during cardiac cycle in reconstructed 3D LV wall models, which is contradictory to the assumption of approximately constant myocardial tissue. In the paper, we propose a probability-based segmentation method to estimate the probabilities of boundary pixels in the trabeculated region near the solid wall belonging to blood or muscle. It helps avoid artifactually moving the endocardium boundary inward during systole, as commonly happens with simple threshold-based segmentation methods. Our method takes s stack of 2D cine MRI slices as input, and produces the 3D probabilistic segmentation of the heart wall using a generative adversarial network (GAN). Based on numerical experiments, our proposed method outperformed the baseline method in terms of evaluation metrics on a synthetic dataset. Moreover, we achieved very good quality reconstructed results with on a real 2D cine MRI dataset (there is no truly independent 3D ground truth). The proposed approach helps to achieve better understanding cardiovascular motion. Moreover, it is the first attempt to use probabilistic segmentation of LV myocardium for 3D heart wall reconstruction from 2D cardiac cine MRI data, to the best of our knowledge.

Original languageEnglish (US)
Title of host publicationStatistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers
EditorsShuo Li, Tommaso Mansi, Kristin McLeod, Mihaela Pop, Jichao Zhao, Maxime Sermesant, Alistair Young, Kawal Rhode
PublisherSpringer Verlag
Pages181-190
Number of pages10
ISBN (Print)9783030120283
DOIs
StatePublished - Jan 1 2019
Event9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018 - Granada, Spain
Duration: Sep 16 2018Sep 16 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11395 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018
CountrySpain
CityGranada
Period9/16/189/16/18

Fingerprint

Left Ventricle
Magnetic Resonance Imaging
Magnetic resonance
Cardiac
Segmentation
Imaging techniques
Myocardium
Systole
Muscle
Motion
Blood
Pixels
Tissue
Slice
Baseline
Pixel
Imaging
Numerical Experiment
Cycle
Metric

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Yang, D., Liu, B., Axel, L., & Metaxas, D. (2019). 3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network. In S. Li, T. Mansi, K. McLeod, M. Pop, J. Zhao, M. Sermesant, A. Young, ... K. Rhode (Eds.), Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers (pp. 181-190). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11395 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-12029-0_20
Yang, Dong ; Liu, Bo ; Axel, Leon ; Metaxas, Dimitri. / 3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network. Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. editor / Shuo Li ; Tommaso Mansi ; Kristin McLeod ; Mihaela Pop ; Jichao Zhao ; Maxime Sermesant ; Alistair Young ; Kawal Rhode. Springer Verlag, 2019. pp. 181-190 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{ee6400b5c8f34dab8a77dc4188767f01,
title = "3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network",
abstract = "Cardiac magnetic resonance imaging (MRI) is one of the most useful techniques to understand and measure cardiac functions. Given the MR image data, segmentation of the left ventricle (LV) myocardium is the most common task to be addressed for recovering and studying LV wall motion. However, most of segmentation methods heavily rely on the imaging appearance for extracting the myocardial contours (epi- and endo-cardium). These methods cannot guarantee a consistent volume of the heart wall during cardiac cycle in reconstructed 3D LV wall models, which is contradictory to the assumption of approximately constant myocardial tissue. In the paper, we propose a probability-based segmentation method to estimate the probabilities of boundary pixels in the trabeculated region near the solid wall belonging to blood or muscle. It helps avoid artifactually moving the endocardium boundary inward during systole, as commonly happens with simple threshold-based segmentation methods. Our method takes s stack of 2D cine MRI slices as input, and produces the 3D probabilistic segmentation of the heart wall using a generative adversarial network (GAN). Based on numerical experiments, our proposed method outperformed the baseline method in terms of evaluation metrics on a synthetic dataset. Moreover, we achieved very good quality reconstructed results with on a real 2D cine MRI dataset (there is no truly independent 3D ground truth). The proposed approach helps to achieve better understanding cardiovascular motion. Moreover, it is the first attempt to use probabilistic segmentation of LV myocardium for 3D heart wall reconstruction from 2D cardiac cine MRI data, to the best of our knowledge.",
author = "Dong Yang and Bo Liu and Leon Axel and Dimitri Metaxas",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-12029-0_20",
language = "English (US)",
isbn = "9783030120283",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "181--190",
editor = "Shuo Li and Tommaso Mansi and Kristin McLeod and Mihaela Pop and Jichao Zhao and Maxime Sermesant and Alistair Young and Kawal Rhode",
booktitle = "Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers",
address = "Germany",

}

Yang, D, Liu, B, Axel, L & Metaxas, D 2019, 3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network. in S Li, T Mansi, K McLeod, M Pop, J Zhao, M Sermesant, A Young & K Rhode (eds), Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11395 LNCS, Springer Verlag, pp. 181-190, 9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018, Granada, Spain, 9/16/18. https://doi.org/10.1007/978-3-030-12029-0_20

3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network. / Yang, Dong; Liu, Bo; Axel, Leon; Metaxas, Dimitri.

Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. ed. / Shuo Li; Tommaso Mansi; Kristin McLeod; Mihaela Pop; Jichao Zhao; Maxime Sermesant; Alistair Young; Kawal Rhode. Springer Verlag, 2019. p. 181-190 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11395 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - 3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network

AU - Yang, Dong

AU - Liu, Bo

AU - Axel, Leon

AU - Metaxas, Dimitri

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Cardiac magnetic resonance imaging (MRI) is one of the most useful techniques to understand and measure cardiac functions. Given the MR image data, segmentation of the left ventricle (LV) myocardium is the most common task to be addressed for recovering and studying LV wall motion. However, most of segmentation methods heavily rely on the imaging appearance for extracting the myocardial contours (epi- and endo-cardium). These methods cannot guarantee a consistent volume of the heart wall during cardiac cycle in reconstructed 3D LV wall models, which is contradictory to the assumption of approximately constant myocardial tissue. In the paper, we propose a probability-based segmentation method to estimate the probabilities of boundary pixels in the trabeculated region near the solid wall belonging to blood or muscle. It helps avoid artifactually moving the endocardium boundary inward during systole, as commonly happens with simple threshold-based segmentation methods. Our method takes s stack of 2D cine MRI slices as input, and produces the 3D probabilistic segmentation of the heart wall using a generative adversarial network (GAN). Based on numerical experiments, our proposed method outperformed the baseline method in terms of evaluation metrics on a synthetic dataset. Moreover, we achieved very good quality reconstructed results with on a real 2D cine MRI dataset (there is no truly independent 3D ground truth). The proposed approach helps to achieve better understanding cardiovascular motion. Moreover, it is the first attempt to use probabilistic segmentation of LV myocardium for 3D heart wall reconstruction from 2D cardiac cine MRI data, to the best of our knowledge.

AB - Cardiac magnetic resonance imaging (MRI) is one of the most useful techniques to understand and measure cardiac functions. Given the MR image data, segmentation of the left ventricle (LV) myocardium is the most common task to be addressed for recovering and studying LV wall motion. However, most of segmentation methods heavily rely on the imaging appearance for extracting the myocardial contours (epi- and endo-cardium). These methods cannot guarantee a consistent volume of the heart wall during cardiac cycle in reconstructed 3D LV wall models, which is contradictory to the assumption of approximately constant myocardial tissue. In the paper, we propose a probability-based segmentation method to estimate the probabilities of boundary pixels in the trabeculated region near the solid wall belonging to blood or muscle. It helps avoid artifactually moving the endocardium boundary inward during systole, as commonly happens with simple threshold-based segmentation methods. Our method takes s stack of 2D cine MRI slices as input, and produces the 3D probabilistic segmentation of the heart wall using a generative adversarial network (GAN). Based on numerical experiments, our proposed method outperformed the baseline method in terms of evaluation metrics on a synthetic dataset. Moreover, we achieved very good quality reconstructed results with on a real 2D cine MRI dataset (there is no truly independent 3D ground truth). The proposed approach helps to achieve better understanding cardiovascular motion. Moreover, it is the first attempt to use probabilistic segmentation of LV myocardium for 3D heart wall reconstruction from 2D cardiac cine MRI data, to the best of our knowledge.

UR - http://www.scopus.com/inward/record.url?scp=85064048751&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064048751&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-12029-0_20

DO - 10.1007/978-3-030-12029-0_20

M3 - Conference contribution

SN - 9783030120283

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 181

EP - 190

BT - Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers

A2 - Li, Shuo

A2 - Mansi, Tommaso

A2 - McLeod, Kristin

A2 - Pop, Mihaela

A2 - Zhao, Jichao

A2 - Sermesant, Maxime

A2 - Young, Alistair

A2 - Rhode, Kawal

PB - Springer Verlag

ER -

Yang D, Liu B, Axel L, Metaxas D. 3D LV Probabilistic Segmentation in Cardiac MRI Using Generative Adversarial Network. In Li S, Mansi T, McLeod K, Pop M, Zhao J, Sermesant M, Young A, Rhode K, editors, Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. Springer Verlag. 2019. p. 181-190. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-12029-0_20