Hierarchical active shape models, using the wavelet transform

Christos Davatzikos, Xiaodong Tao, Dinggang Shen

Research output: Contribution to journalArticle

Abstract

Active shape models (ASMs) are often limited by the inability of relatively few eigenvectors to capture the full range of biological shape variability. This paper presents a method that overcomes this limitation, by using a hierarchical formulation of active shape models, using the wavelet transform. The statistical properties of the wavelet transform of a deformable contour are analyzed via principal component analysis, and used as priors in the contour's deformation. Some of these priors reflect relatively global shape characteristics of the object boundaries, whereas, some of them capture local and high-frequency shape characteristics and, thus, serve as local smoothness constraints. This formulation achieves two objectives. First, it is robust when only a limited number of training samples is available. Second, by using local statistics as smoothness constraints, it eliminates the need for adopting ad hoc physical models, such as elasticity or other smoothness models, which do not necessarily reflect true biological variability. Examples on magnetic resonance images of the corpus callosum and hand contours demonstrate that good and fully automated segmentations can be achieved, even with as few as five training samples.

Original languageEnglish (US)
Pages (from-to)414-423
Number of pages10
JournalIEEE Transactions on Medical Imaging
Volume22
Issue number3
DOIs
StatePublished - Mar 2003
Externally publishedYes

Fingerprint

Wavelet Analysis
Wavelet transforms
Corpus Callosum
Elasticity
Principal Component Analysis
Magnetic Resonance Spectroscopy
Hand
Magnetic resonance
Eigenvalues and eigenfunctions
Principal component analysis
Statistics

Keywords

  • Active shape model
  • Deformable contours
  • The wavelet transform

ASJC Scopus subject areas

  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology
  • Electrical and Electronic Engineering
  • Computer Science Applications
  • Computational Theory and Mathematics

Cite this

Hierarchical active shape models, using the wavelet transform. / Davatzikos, Christos; Tao, Xiaodong; Shen, Dinggang.

In: IEEE Transactions on Medical Imaging, Vol. 22, No. 3, 03.2003, p. 414-423.

Research output: Contribution to journalArticle

Davatzikos, Christos ; Tao, Xiaodong ; Shen, Dinggang. / Hierarchical active shape models, using the wavelet transform. In: IEEE Transactions on Medical Imaging. 2003 ; Vol. 22, No. 3. pp. 414-423.
@article{3d5044cb967b48e089a8f2ca62ba6871,
title = "Hierarchical active shape models, using the wavelet transform",
abstract = "Active shape models (ASMs) are often limited by the inability of relatively few eigenvectors to capture the full range of biological shape variability. This paper presents a method that overcomes this limitation, by using a hierarchical formulation of active shape models, using the wavelet transform. The statistical properties of the wavelet transform of a deformable contour are analyzed via principal component analysis, and used as priors in the contour's deformation. Some of these priors reflect relatively global shape characteristics of the object boundaries, whereas, some of them capture local and high-frequency shape characteristics and, thus, serve as local smoothness constraints. This formulation achieves two objectives. First, it is robust when only a limited number of training samples is available. Second, by using local statistics as smoothness constraints, it eliminates the need for adopting ad hoc physical models, such as elasticity or other smoothness models, which do not necessarily reflect true biological variability. Examples on magnetic resonance images of the corpus callosum and hand contours demonstrate that good and fully automated segmentations can be achieved, even with as few as five training samples.",
keywords = "Active shape model, Deformable contours, The wavelet transform",
author = "Christos Davatzikos and Xiaodong Tao and Dinggang Shen",
year = "2003",
month = "3",
doi = "10.1109/TMI.2003.809688",
language = "English (US)",
volume = "22",
pages = "414--423",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Hierarchical active shape models, using the wavelet transform

AU - Davatzikos, Christos

AU - Tao, Xiaodong

AU - Shen, Dinggang

PY - 2003/3

Y1 - 2003/3

N2 - Active shape models (ASMs) are often limited by the inability of relatively few eigenvectors to capture the full range of biological shape variability. This paper presents a method that overcomes this limitation, by using a hierarchical formulation of active shape models, using the wavelet transform. The statistical properties of the wavelet transform of a deformable contour are analyzed via principal component analysis, and used as priors in the contour's deformation. Some of these priors reflect relatively global shape characteristics of the object boundaries, whereas, some of them capture local and high-frequency shape characteristics and, thus, serve as local smoothness constraints. This formulation achieves two objectives. First, it is robust when only a limited number of training samples is available. Second, by using local statistics as smoothness constraints, it eliminates the need for adopting ad hoc physical models, such as elasticity or other smoothness models, which do not necessarily reflect true biological variability. Examples on magnetic resonance images of the corpus callosum and hand contours demonstrate that good and fully automated segmentations can be achieved, even with as few as five training samples.

AB - Active shape models (ASMs) are often limited by the inability of relatively few eigenvectors to capture the full range of biological shape variability. This paper presents a method that overcomes this limitation, by using a hierarchical formulation of active shape models, using the wavelet transform. The statistical properties of the wavelet transform of a deformable contour are analyzed via principal component analysis, and used as priors in the contour's deformation. Some of these priors reflect relatively global shape characteristics of the object boundaries, whereas, some of them capture local and high-frequency shape characteristics and, thus, serve as local smoothness constraints. This formulation achieves two objectives. First, it is robust when only a limited number of training samples is available. Second, by using local statistics as smoothness constraints, it eliminates the need for adopting ad hoc physical models, such as elasticity or other smoothness models, which do not necessarily reflect true biological variability. Examples on magnetic resonance images of the corpus callosum and hand contours demonstrate that good and fully automated segmentations can be achieved, even with as few as five training samples.

KW - Active shape model

KW - Deformable contours

KW - The wavelet transform

UR - http://www.scopus.com/inward/record.url?scp=0038321577&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0038321577&partnerID=8YFLogxK

U2 - 10.1109/TMI.2003.809688

DO - 10.1109/TMI.2003.809688

M3 - Article

VL - 22

SP - 414

EP - 423

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 3

ER -