Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images

Renjie He, Sushmita Datta, Balasrinivasa Rao Sajja, Ponnada A. Narayana

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

An integrated approach for multi-spectral segmentation of MR images is presented. This method is based on the fuzzy c-means (FCM) and includes bias field correction and contextual constraints over spatial intensity distribution and accounts for the non-spherical cluster's shape in the feature space. The bias field is modeled as a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of intensity are added into the FCM cost functions. To reduce the computational complexity, the contextual regularizations are separated from the clustering iterations. Since the feature space is not isotropic, distance measure adopted in Gustafson-Kessel (G-K) algorithm is used instead of the Euclidean distance, to account for the non-spherical shape of the clusters in the feature space. These algorithms are quantitatively evaluated on MR brain images using the similarity measures.

Original languageEnglish (US)
Pages (from-to)353-366
Number of pages14
JournalComputerized Medical Imaging and Graphics
Volume32
Issue number5
DOIs
StatePublished - Jul 1 2008

Fingerprint

Fuzzy clustering
Magnetic resonance
Cluster Analysis
Magnetic Resonance Spectroscopy
Cost functions
Computational complexity
Brain
Polynomials
Costs and Cost Analysis

Keywords

  • Adaptive FCM
  • Contextual constraints
  • G-K algorithm
  • Inhomogeneity field
  • MRI
  • Multi-spectral segmentation
  • Similarity measures

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Computer Vision and Pattern Recognition
  • Health Informatics
  • Computer Graphics and Computer-Aided Design

Cite this

Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images. / He, Renjie; Datta, Sushmita; Sajja, Balasrinivasa Rao; Narayana, Ponnada A.

In: Computerized Medical Imaging and Graphics, Vol. 32, No. 5, 01.07.2008, p. 353-366.

Research output: Contribution to journalArticle

@article{c277ba3ac4a446a0ac02525917e07427,
title = "Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images",
abstract = "An integrated approach for multi-spectral segmentation of MR images is presented. This method is based on the fuzzy c-means (FCM) and includes bias field correction and contextual constraints over spatial intensity distribution and accounts for the non-spherical cluster's shape in the feature space. The bias field is modeled as a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of intensity are added into the FCM cost functions. To reduce the computational complexity, the contextual regularizations are separated from the clustering iterations. Since the feature space is not isotropic, distance measure adopted in Gustafson-Kessel (G-K) algorithm is used instead of the Euclidean distance, to account for the non-spherical shape of the clusters in the feature space. These algorithms are quantitatively evaluated on MR brain images using the similarity measures.",
keywords = "Adaptive FCM, Contextual constraints, G-K algorithm, Inhomogeneity field, MRI, Multi-spectral segmentation, Similarity measures",
author = "Renjie He and Sushmita Datta and Sajja, {Balasrinivasa Rao} and Narayana, {Ponnada A.}",
year = "2008",
month = "7",
day = "1",
doi = "10.1016/j.compmedimag.2008.02.002",
language = "English (US)",
volume = "32",
pages = "353--366",
journal = "Computerized Medical Imaging and Graphics",
issn = "0895-6111",
publisher = "Elsevier Limited",
number = "5",

}

TY - JOUR

T1 - Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images

AU - He, Renjie

AU - Datta, Sushmita

AU - Sajja, Balasrinivasa Rao

AU - Narayana, Ponnada A.

PY - 2008/7/1

Y1 - 2008/7/1

N2 - An integrated approach for multi-spectral segmentation of MR images is presented. This method is based on the fuzzy c-means (FCM) and includes bias field correction and contextual constraints over spatial intensity distribution and accounts for the non-spherical cluster's shape in the feature space. The bias field is modeled as a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of intensity are added into the FCM cost functions. To reduce the computational complexity, the contextual regularizations are separated from the clustering iterations. Since the feature space is not isotropic, distance measure adopted in Gustafson-Kessel (G-K) algorithm is used instead of the Euclidean distance, to account for the non-spherical shape of the clusters in the feature space. These algorithms are quantitatively evaluated on MR brain images using the similarity measures.

AB - An integrated approach for multi-spectral segmentation of MR images is presented. This method is based on the fuzzy c-means (FCM) and includes bias field correction and contextual constraints over spatial intensity distribution and accounts for the non-spherical cluster's shape in the feature space. The bias field is modeled as a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of intensity are added into the FCM cost functions. To reduce the computational complexity, the contextual regularizations are separated from the clustering iterations. Since the feature space is not isotropic, distance measure adopted in Gustafson-Kessel (G-K) algorithm is used instead of the Euclidean distance, to account for the non-spherical shape of the clusters in the feature space. These algorithms are quantitatively evaluated on MR brain images using the similarity measures.

KW - Adaptive FCM

KW - Contextual constraints

KW - G-K algorithm

KW - Inhomogeneity field

KW - MRI

KW - Multi-spectral segmentation

KW - Similarity measures

UR - http://www.scopus.com/inward/record.url?scp=44149104064&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44149104064&partnerID=8YFLogxK

U2 - 10.1016/j.compmedimag.2008.02.002

DO - 10.1016/j.compmedimag.2008.02.002

M3 - Article

C2 - 18387784

AN - SCOPUS:44149104064

VL - 32

SP - 353

EP - 366

JO - Computerized Medical Imaging and Graphics

JF - Computerized Medical Imaging and Graphics

SN - 0895-6111

IS - 5

ER -