Automatic segmentation of neonatal images using convex optimization and coupled level sets

Li Wang, Feng Shi, Weili Lin, John H. Gilmore, Dinggang Shen

Research output: Contribution to journalArticle

90 Citations (Scopus)

Abstract

Accurate segmentation of neonatal brain MR images remains challenging mainly due to their poor spatial resolution, inverted contrast between white matter and gray matter, and high intensity inhomogeneity. Most existing methods for neonatal brain segmentation are atlas-based and voxel-wise. Although active contour/surface models with geometric information constraint have been successfully applied to adult brain segmentation, they are not fully explored in the neonatal image segmentation. In this paper, we propose a novel neonatal image segmentation method by combining local intensity information, atlas spatial prior, and cortical thickness constraint in a single level-set framework. Besides, we also provide a robust and reliable tissue surface initialization for the proposed method by using a convex optimization technique. Thus, tissue segmentation, as well as inner and outer cortical surface reconstruction, can be obtained simultaneously. The proposed method has been tested on a large neonatal dataset, and the validation on 10 neonatal brain images (with manual segmentations) shows very promising results.

Original languageEnglish
Pages (from-to)805-817
Number of pages13
JournalNeuroImage
Volume58
Issue number3
DOIs
Publication statusPublished - 2011 Oct 1
Externally publishedYes

Fingerprint

Atlases
Brain
White Matter
Gray Matter
Datasets

Keywords

  • Atlas-based segmentation
  • Convex optimization
  • Coupled level sets
  • Neonatal tissue segmentation

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

Automatic segmentation of neonatal images using convex optimization and coupled level sets. / Wang, Li; Shi, Feng; Lin, Weili; Gilmore, John H.; Shen, Dinggang.

In: NeuroImage, Vol. 58, No. 3, 01.10.2011, p. 805-817.

Research output: Contribution to journalArticle

Wang, Li ; Shi, Feng ; Lin, Weili ; Gilmore, John H. ; Shen, Dinggang. / Automatic segmentation of neonatal images using convex optimization and coupled level sets. In: NeuroImage. 2011 ; Vol. 58, No. 3. pp. 805-817.
@article{4d8617ef3aa64389a441a213c55475a3,
title = "Automatic segmentation of neonatal images using convex optimization and coupled level sets",
abstract = "Accurate segmentation of neonatal brain MR images remains challenging mainly due to their poor spatial resolution, inverted contrast between white matter and gray matter, and high intensity inhomogeneity. Most existing methods for neonatal brain segmentation are atlas-based and voxel-wise. Although active contour/surface models with geometric information constraint have been successfully applied to adult brain segmentation, they are not fully explored in the neonatal image segmentation. In this paper, we propose a novel neonatal image segmentation method by combining local intensity information, atlas spatial prior, and cortical thickness constraint in a single level-set framework. Besides, we also provide a robust and reliable tissue surface initialization for the proposed method by using a convex optimization technique. Thus, tissue segmentation, as well as inner and outer cortical surface reconstruction, can be obtained simultaneously. The proposed method has been tested on a large neonatal dataset, and the validation on 10 neonatal brain images (with manual segmentations) shows very promising results.",
keywords = "Atlas-based segmentation, Convex optimization, Coupled level sets, Neonatal tissue segmentation",
author = "Li Wang and Feng Shi and Weili Lin and Gilmore, {John H.} and Dinggang Shen",
year = "2011",
month = "10",
day = "1",
doi = "10.1016/j.neuroimage.2011.06.064",
language = "English",
volume = "58",
pages = "805--817",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",
number = "3",

}

TY - JOUR

T1 - Automatic segmentation of neonatal images using convex optimization and coupled level sets

AU - Wang, Li

AU - Shi, Feng

AU - Lin, Weili

AU - Gilmore, John H.

AU - Shen, Dinggang

PY - 2011/10/1

Y1 - 2011/10/1

N2 - Accurate segmentation of neonatal brain MR images remains challenging mainly due to their poor spatial resolution, inverted contrast between white matter and gray matter, and high intensity inhomogeneity. Most existing methods for neonatal brain segmentation are atlas-based and voxel-wise. Although active contour/surface models with geometric information constraint have been successfully applied to adult brain segmentation, they are not fully explored in the neonatal image segmentation. In this paper, we propose a novel neonatal image segmentation method by combining local intensity information, atlas spatial prior, and cortical thickness constraint in a single level-set framework. Besides, we also provide a robust and reliable tissue surface initialization for the proposed method by using a convex optimization technique. Thus, tissue segmentation, as well as inner and outer cortical surface reconstruction, can be obtained simultaneously. The proposed method has been tested on a large neonatal dataset, and the validation on 10 neonatal brain images (with manual segmentations) shows very promising results.

AB - Accurate segmentation of neonatal brain MR images remains challenging mainly due to their poor spatial resolution, inverted contrast between white matter and gray matter, and high intensity inhomogeneity. Most existing methods for neonatal brain segmentation are atlas-based and voxel-wise. Although active contour/surface models with geometric information constraint have been successfully applied to adult brain segmentation, they are not fully explored in the neonatal image segmentation. In this paper, we propose a novel neonatal image segmentation method by combining local intensity information, atlas spatial prior, and cortical thickness constraint in a single level-set framework. Besides, we also provide a robust and reliable tissue surface initialization for the proposed method by using a convex optimization technique. Thus, tissue segmentation, as well as inner and outer cortical surface reconstruction, can be obtained simultaneously. The proposed method has been tested on a large neonatal dataset, and the validation on 10 neonatal brain images (with manual segmentations) shows very promising results.

KW - Atlas-based segmentation

KW - Convex optimization

KW - Coupled level sets

KW - Neonatal tissue segmentation

UR - http://www.scopus.com/inward/record.url?scp=80052159689&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80052159689&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2011.06.064

DO - 10.1016/j.neuroimage.2011.06.064

M3 - Article

VL - 58

SP - 805

EP - 817

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

IS - 3

ER -