Segmentation of the left ventricle in cardiac MRI using a hierarchical extreme learning machine model

Yang Luo, Benqiang Yang, Lisheng Xu, Liling Hao, Jun Liu, Yang Yao, Frans van de Vosse

Research output: Contribution to journalArticleAcademicpeer-review

3 Citations (Scopus)
2 Downloads (Pure)

Abstract

Segmentation of the left ventricle (LV) from cardiac magnetic resonance imaging (MRI) images is an essential step for calculation of clinical indices such as stroke volume, ejection fraction. In this paper, a new automatic LV segmentation method combines a Hierarchical Extreme Learning Machine (H-ELM) and a new location method is developed. An H-ELM can achieve more compact and meaningful feature representations and learn the segmentation task from the ground truth. A new automatic LV location method is integrated to improve the accuracy of classification and reduce the cost of segmentation. Experimental results (including 30 cases, 10 cases for training, 20 cases for testing) show that the mean absolute deviation of images segmented by our proposed method is about 67.9, 81.3 and 98.7% of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean maximum absolute deviation of images segmented by our proposed method is about 63.5, 77.3 and 98.0% of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean dice similarity coefficient of images segmented by our proposed method is about 13.7, 9.3 and 0.5% higher than that of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean speed of our proposed method is about 38.3, 6.7 and 23.8 times faster than that of the level set, the SVM and Hu’s method, respectively. The standard deviation of our proposed method is the lowest among four methods. The results validate that our proposed method is efficient and satisfactory for the LV segmentation.

Original languageEnglish
Pages (from-to)1741-1751
Number of pages11
JournalInternational Journal of Machine Learning and Cybernetics
Volume9
Issue number10
DOIs
Publication statusPublished - 1 Oct 2018

Fingerprint

Magnetic resonance
Learning systems
Imaging techniques
Testing
Costs

Keywords

  • Hierarchical extreme learning machine
  • Image segmentation
  • Left ventricle
  • Magnetic resonance imaging

Cite this

Luo, Yang ; Yang, Benqiang ; Xu, Lisheng ; Hao, Liling ; Liu, Jun ; Yao, Yang ; Vosse, Frans van de. / Segmentation of the left ventricle in cardiac MRI using a hierarchical extreme learning machine model. In: International Journal of Machine Learning and Cybernetics. 2018 ; Vol. 9, No. 10. pp. 1741-1751.
@article{5c62b63e397e42aa8943e7b882f115dd,
title = "Segmentation of the left ventricle in cardiac MRI using a hierarchical extreme learning machine model",
abstract = "Segmentation of the left ventricle (LV) from cardiac magnetic resonance imaging (MRI) images is an essential step for calculation of clinical indices such as stroke volume, ejection fraction. In this paper, a new automatic LV segmentation method combines a Hierarchical Extreme Learning Machine (H-ELM) and a new location method is developed. An H-ELM can achieve more compact and meaningful feature representations and learn the segmentation task from the ground truth. A new automatic LV location method is integrated to improve the accuracy of classification and reduce the cost of segmentation. Experimental results (including 30 cases, 10 cases for training, 20 cases for testing) show that the mean absolute deviation of images segmented by our proposed method is about 67.9, 81.3 and 98.7{\%} of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean maximum absolute deviation of images segmented by our proposed method is about 63.5, 77.3 and 98.0{\%} of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean dice similarity coefficient of images segmented by our proposed method is about 13.7, 9.3 and 0.5{\%} higher than that of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean speed of our proposed method is about 38.3, 6.7 and 23.8 times faster than that of the level set, the SVM and Hu’s method, respectively. The standard deviation of our proposed method is the lowest among four methods. The results validate that our proposed method is efficient and satisfactory for the LV segmentation.",
keywords = "Hierarchical extreme learning machine, Image segmentation, Left ventricle, Magnetic resonance imaging",
author = "Yang Luo and Benqiang Yang and Lisheng Xu and Liling Hao and Jun Liu and Yang Yao and Vosse, {Frans van de}",
year = "2018",
month = "10",
day = "1",
doi = "10.1007/s13042-017-0678-4",
language = "English",
volume = "9",
pages = "1741--1751",
journal = "International Journal of Machine Learning and Cybernetics",
issn = "1868-8071",
publisher = "Springer",
number = "10",

}

Segmentation of the left ventricle in cardiac MRI using a hierarchical extreme learning machine model. / Luo, Yang; Yang, Benqiang; Xu, Lisheng; Hao, Liling; Liu, Jun; Yao, Yang; Vosse, Frans van de.

In: International Journal of Machine Learning and Cybernetics, Vol. 9, No. 10, 01.10.2018, p. 1741-1751.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Segmentation of the left ventricle in cardiac MRI using a hierarchical extreme learning machine model

AU - Luo, Yang

AU - Yang, Benqiang

AU - Xu, Lisheng

AU - Hao, Liling

AU - Liu, Jun

AU - Yao, Yang

AU - Vosse, Frans van de

PY - 2018/10/1

Y1 - 2018/10/1

N2 - Segmentation of the left ventricle (LV) from cardiac magnetic resonance imaging (MRI) images is an essential step for calculation of clinical indices such as stroke volume, ejection fraction. In this paper, a new automatic LV segmentation method combines a Hierarchical Extreme Learning Machine (H-ELM) and a new location method is developed. An H-ELM can achieve more compact and meaningful feature representations and learn the segmentation task from the ground truth. A new automatic LV location method is integrated to improve the accuracy of classification and reduce the cost of segmentation. Experimental results (including 30 cases, 10 cases for training, 20 cases for testing) show that the mean absolute deviation of images segmented by our proposed method is about 67.9, 81.3 and 98.7% of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean maximum absolute deviation of images segmented by our proposed method is about 63.5, 77.3 and 98.0% of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean dice similarity coefficient of images segmented by our proposed method is about 13.7, 9.3 and 0.5% higher than that of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean speed of our proposed method is about 38.3, 6.7 and 23.8 times faster than that of the level set, the SVM and Hu’s method, respectively. The standard deviation of our proposed method is the lowest among four methods. The results validate that our proposed method is efficient and satisfactory for the LV segmentation.

AB - Segmentation of the left ventricle (LV) from cardiac magnetic resonance imaging (MRI) images is an essential step for calculation of clinical indices such as stroke volume, ejection fraction. In this paper, a new automatic LV segmentation method combines a Hierarchical Extreme Learning Machine (H-ELM) and a new location method is developed. An H-ELM can achieve more compact and meaningful feature representations and learn the segmentation task from the ground truth. A new automatic LV location method is integrated to improve the accuracy of classification and reduce the cost of segmentation. Experimental results (including 30 cases, 10 cases for training, 20 cases for testing) show that the mean absolute deviation of images segmented by our proposed method is about 67.9, 81.3 and 98.7% of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean maximum absolute deviation of images segmented by our proposed method is about 63.5, 77.3 and 98.0% of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean dice similarity coefficient of images segmented by our proposed method is about 13.7, 9.3 and 0.5% higher than that of those images segmented by the level set, the SVM and Hu’s method, respectively. The mean speed of our proposed method is about 38.3, 6.7 and 23.8 times faster than that of the level set, the SVM and Hu’s method, respectively. The standard deviation of our proposed method is the lowest among four methods. The results validate that our proposed method is efficient and satisfactory for the LV segmentation.

KW - Hierarchical extreme learning machine

KW - Image segmentation

KW - Left ventricle

KW - Magnetic resonance imaging

UR - http://www.scopus.com/inward/record.url?scp=85052869025&partnerID=8YFLogxK

U2 - 10.1007/s13042-017-0678-4

DO - 10.1007/s13042-017-0678-4

M3 - Article

AN - SCOPUS:85052869025

VL - 9

SP - 1741

EP - 1751

JO - International Journal of Machine Learning and Cybernetics

JF - International Journal of Machine Learning and Cybernetics

SN - 1868-8071

IS - 10

ER -