Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation

László Szilágyi, Sándor M. Szilágyi, Balázs Benyó, Z. Benyó

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

9 Citations (Scopus)

Abstract

Intensity inhomogeneity or intensity non-uniformity (INU) is an undesired phenomenon that represents the main obstacle for MR image segmentation and registration methods. Various techniques have been proposed to eliminate or compensate the INU, most of which are embedded into clustering algorithms. This paper proposes a hybrid c-means clustering approach to replace the FCM algorithm found in several existing solutions. The novel clustering model is assisted by a pre-filtering technique for Gaussian and impulse noise elimination, and a smoothening filter that helps the c-means algorithm at the estimation of inhomogeneity as a slowly varying additive or multiplicative noise. The slow variance of the estimated INU is assured by a smoothening filter that performs a context dependent averaging, based on a morphological criterion. The experiments using 2-D synthetic phantoms and real MR images show that the proposed method provides more accurate and more efficient segmentation than the FCM based approach. The produced segmentation and fuzzy membership values can serve as excellent support for 3-D registration and segmentation techniques.

Original languageEnglish
Title of host publicationProceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009
Pages105-110
Number of pages6
DOIs
Publication statusPublished - 2009
Event2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009 - Timisoara
Duration: May 28 2009May 29 2009

Other

Other2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009
CityTimisoara
Period5/28/095/29/09

Fingerprint

Image segmentation
Brain
Gaussian noise (electronic)
Impulse noise
Image registration
Clustering algorithms
Experiments
Compensation and Redress

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Information Systems
  • Software

Cite this

Szilágyi, L., Szilágyi, S. M., Benyó, B., & Benyó, Z. (2009). Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation. In Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009 (pp. 105-110). [5136221] https://doi.org/10.1109/SACI.2009.5136221

Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation. / Szilágyi, László; Szilágyi, Sándor M.; Benyó, Balázs; Benyó, Z.

Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009. 2009. p. 105-110 5136221.

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

Szilágyi, L, Szilágyi, SM, Benyó, B & Benyó, Z 2009, Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation. in Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009., 5136221, pp. 105-110, 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009, Timisoara, 5/28/09. https://doi.org/10.1109/SACI.2009.5136221
Szilágyi L, Szilágyi SM, Benyó B, Benyó Z. Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation. In Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009. 2009. p. 105-110. 5136221 https://doi.org/10.1109/SACI.2009.5136221
Szilágyi, László ; Szilágyi, Sándor M. ; Benyó, Balázs ; Benyó, Z. / Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation. Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009. 2009. pp. 105-110
@inproceedings{5da60164289d4de6bdd31d5a3e74ab45,
title = "Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation",
abstract = "Intensity inhomogeneity or intensity non-uniformity (INU) is an undesired phenomenon that represents the main obstacle for MR image segmentation and registration methods. Various techniques have been proposed to eliminate or compensate the INU, most of which are embedded into clustering algorithms. This paper proposes a hybrid c-means clustering approach to replace the FCM algorithm found in several existing solutions. The novel clustering model is assisted by a pre-filtering technique for Gaussian and impulse noise elimination, and a smoothening filter that helps the c-means algorithm at the estimation of inhomogeneity as a slowly varying additive or multiplicative noise. The slow variance of the estimated INU is assured by a smoothening filter that performs a context dependent averaging, based on a morphological criterion. The experiments using 2-D synthetic phantoms and real MR images show that the proposed method provides more accurate and more efficient segmentation than the FCM based approach. The produced segmentation and fuzzy membership values can serve as excellent support for 3-D registration and segmentation techniques.",
author = "L{\'a}szl{\'o} Szil{\'a}gyi and Szil{\'a}gyi, {S{\'a}ndor M.} and Bal{\'a}zs Beny{\'o} and Z. Beny{\'o}",
year = "2009",
doi = "10.1109/SACI.2009.5136221",
language = "English",
isbn = "9781424444786",
pages = "105--110",
booktitle = "Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009",

}

TY - GEN

T1 - Application of hybrid c-means clustering models in inhomogeneity compensation and MR brain image segmentation

AU - Szilágyi, László

AU - Szilágyi, Sándor M.

AU - Benyó, Balázs

AU - Benyó, Z.

PY - 2009

Y1 - 2009

N2 - Intensity inhomogeneity or intensity non-uniformity (INU) is an undesired phenomenon that represents the main obstacle for MR image segmentation and registration methods. Various techniques have been proposed to eliminate or compensate the INU, most of which are embedded into clustering algorithms. This paper proposes a hybrid c-means clustering approach to replace the FCM algorithm found in several existing solutions. The novel clustering model is assisted by a pre-filtering technique for Gaussian and impulse noise elimination, and a smoothening filter that helps the c-means algorithm at the estimation of inhomogeneity as a slowly varying additive or multiplicative noise. The slow variance of the estimated INU is assured by a smoothening filter that performs a context dependent averaging, based on a morphological criterion. The experiments using 2-D synthetic phantoms and real MR images show that the proposed method provides more accurate and more efficient segmentation than the FCM based approach. The produced segmentation and fuzzy membership values can serve as excellent support for 3-D registration and segmentation techniques.

AB - Intensity inhomogeneity or intensity non-uniformity (INU) is an undesired phenomenon that represents the main obstacle for MR image segmentation and registration methods. Various techniques have been proposed to eliminate or compensate the INU, most of which are embedded into clustering algorithms. This paper proposes a hybrid c-means clustering approach to replace the FCM algorithm found in several existing solutions. The novel clustering model is assisted by a pre-filtering technique for Gaussian and impulse noise elimination, and a smoothening filter that helps the c-means algorithm at the estimation of inhomogeneity as a slowly varying additive or multiplicative noise. The slow variance of the estimated INU is assured by a smoothening filter that performs a context dependent averaging, based on a morphological criterion. The experiments using 2-D synthetic phantoms and real MR images show that the proposed method provides more accurate and more efficient segmentation than the FCM based approach. The produced segmentation and fuzzy membership values can serve as excellent support for 3-D registration and segmentation techniques.

UR - http://www.scopus.com/inward/record.url?scp=70350215696&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350215696&partnerID=8YFLogxK

U2 - 10.1109/SACI.2009.5136221

DO - 10.1109/SACI.2009.5136221

M3 - Conference contribution

AN - SCOPUS:70350215696

SN - 9781424444786

SP - 105

EP - 110

BT - Proceedings - 2009 5th International Symposium on Applied Computational Intelligence and Informatics, SACI 2009

ER -