Automatic segmentation of neonatal images using convex optimization and coupled level set method

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

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

2 Citations (Scopus)

Abstract

Accurate segmentation of neonatal brain MR images remains challenging mainly due to poor spatial resolution, low tissue contrast, high intensity inhomogeneity. Most existing methods for neonatal brain segmentation are atlas-based and voxel-wise. Although parametric or geometric deformable models have been successfully applied to adult brain segmentation, to the best of our knowledge, they are not explored in neonatal images. In this paper, we propose a novel neonatal image segmentation method, combining local intensity information, atlas spatial prior and cortical thickness constraint, in a level set framework. Besides, we also provide a robust and reliable tissue surfaces initialization for our proposed level set method by using a convex optimization technique. Validation is performed on 10 neonatal brain images with promising results.

Original languageEnglish
Title of host publicationMedical Imaging and Augmented Reality - 5th International Workshop, MIAR 2010, Proceedings
Pages1-10
Number of pages10
DOIs
Publication statusPublished - 2010
Externally publishedYes
Event5th International Workshop on Medical Imaging and Augmented Reality, MIAR 2010 - Beijing, China
Duration: 2010 Sep 192010 Sep 20

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6326 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other5th International Workshop on Medical Imaging and Augmented Reality, MIAR 2010
CountryChina
CityBeijing
Period10/9/1910/9/20

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Automatic segmentation of neonatal images using convex optimization and coupled level set method'. Together they form a unique fingerprint.

Cite this