A low cost approach for brain tumor segmentation based on intensity modeling and 3D Random Walker

Vasileios G. Kanas, Evangelia I. Zacharaki, Christos Davatzikos, Kyriakos N. Sgarbas, Vasileios Megalooikonomou

Research output: Contribution to journalArticle

Abstract

Abstract Objective Magnetic resonance imaging (MRI) is the primary imaging technique for evaluation of the brain tumor progression before and after radiotherapy or surgery. The purpose of the current study is to exploit conventional MR modalities in order to identify and segment brain images with neoplasms. Methods Four conventional MR sequences, namely, T1-weighted, gadolinium-enhanced T1-weighted, T2-weighted and fluid attenuation inversion recovery, are combined with machine learning techniques to extract global and local information of brain tissues and model the healthy and neoplastic imaging profiles. Healthy tissue clustering, outlier detection and geometric and spatial constraints are applied to perform a first segmentation which is further improved by a modified multiparametric Random Walker segmentation method. The proposed framework is applied on clinical data from 57 brain tumor patients (acquired by different scanners and acquisition parameters) and on 25 synthetic MR images with tumors. Assessment is performed against expert-defined tissue masks and is based on sensitivity analysis and Dice coefficient. Results The results demonstrate that the proposed multiparametric framework differentiates neoplastic tissues with accuracy similar to most current approaches while it achieves lower computational cost and higher degree of automation. Conclusion This study might provide a decision-support tool for neoplastic tissue segmentation, which can assist in treatment planning for tumor resection or focused radiotherapy.

Original languageEnglish (US)
Article number698
Pages (from-to)19-30
Number of pages12
JournalBiomedical Signal Processing and Control
Volume22
DOIs
StatePublished - Jul 13 2015
Externally publishedYes

Fingerprint

Brain Neoplasms
Tumors
Brain
Tissue
Costs and Cost Analysis
Radiotherapy
Costs
Imaging techniques
Neoplasms
Gadolinium
Automation
Masks
Magnetic resonance imaging
Surgery
Sensitivity analysis
Cluster Analysis
Learning systems
Magnetic Resonance Imaging
Planning
Recovery

Keywords

  • Brain neoplasms
  • Magnetic resonance imaging (MRI)
  • Outlier detection
  • Random walks
  • Tumor segmentation

ASJC Scopus subject areas

  • Health Informatics
  • Signal Processing

Cite this

A low cost approach for brain tumor segmentation based on intensity modeling and 3D Random Walker. / Kanas, Vasileios G.; Zacharaki, Evangelia I.; Davatzikos, Christos; Sgarbas, Kyriakos N.; Megalooikonomou, Vasileios.

In: Biomedical Signal Processing and Control, Vol. 22, 698, 13.07.2015, p. 19-30.

Research output: Contribution to journalArticle

Kanas, Vasileios G. ; Zacharaki, Evangelia I. ; Davatzikos, Christos ; Sgarbas, Kyriakos N. ; Megalooikonomou, Vasileios. / A low cost approach for brain tumor segmentation based on intensity modeling and 3D Random Walker. In: Biomedical Signal Processing and Control. 2015 ; Vol. 22. pp. 19-30.
@article{6b5d8cb40590432eb0991f371371f615,
title = "A low cost approach for brain tumor segmentation based on intensity modeling and 3D Random Walker",
abstract = "Abstract Objective Magnetic resonance imaging (MRI) is the primary imaging technique for evaluation of the brain tumor progression before and after radiotherapy or surgery. The purpose of the current study is to exploit conventional MR modalities in order to identify and segment brain images with neoplasms. Methods Four conventional MR sequences, namely, T1-weighted, gadolinium-enhanced T1-weighted, T2-weighted and fluid attenuation inversion recovery, are combined with machine learning techniques to extract global and local information of brain tissues and model the healthy and neoplastic imaging profiles. Healthy tissue clustering, outlier detection and geometric and spatial constraints are applied to perform a first segmentation which is further improved by a modified multiparametric Random Walker segmentation method. The proposed framework is applied on clinical data from 57 brain tumor patients (acquired by different scanners and acquisition parameters) and on 25 synthetic MR images with tumors. Assessment is performed against expert-defined tissue masks and is based on sensitivity analysis and Dice coefficient. Results The results demonstrate that the proposed multiparametric framework differentiates neoplastic tissues with accuracy similar to most current approaches while it achieves lower computational cost and higher degree of automation. Conclusion This study might provide a decision-support tool for neoplastic tissue segmentation, which can assist in treatment planning for tumor resection or focused radiotherapy.",
keywords = "Brain neoplasms, Magnetic resonance imaging (MRI), Outlier detection, Random walks, Tumor segmentation",
author = "Kanas, {Vasileios G.} and Zacharaki, {Evangelia I.} and Christos Davatzikos and Sgarbas, {Kyriakos N.} and Vasileios Megalooikonomou",
year = "2015",
month = "7",
day = "13",
doi = "10.1016/j.bspc.2015.06.004",
language = "English (US)",
volume = "22",
pages = "19--30",
journal = "Biomedical Signal Processing and Control",
issn = "1746-8094",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - A low cost approach for brain tumor segmentation based on intensity modeling and 3D Random Walker

AU - Kanas, Vasileios G.

AU - Zacharaki, Evangelia I.

AU - Davatzikos, Christos

AU - Sgarbas, Kyriakos N.

AU - Megalooikonomou, Vasileios

PY - 2015/7/13

Y1 - 2015/7/13

N2 - Abstract Objective Magnetic resonance imaging (MRI) is the primary imaging technique for evaluation of the brain tumor progression before and after radiotherapy or surgery. The purpose of the current study is to exploit conventional MR modalities in order to identify and segment brain images with neoplasms. Methods Four conventional MR sequences, namely, T1-weighted, gadolinium-enhanced T1-weighted, T2-weighted and fluid attenuation inversion recovery, are combined with machine learning techniques to extract global and local information of brain tissues and model the healthy and neoplastic imaging profiles. Healthy tissue clustering, outlier detection and geometric and spatial constraints are applied to perform a first segmentation which is further improved by a modified multiparametric Random Walker segmentation method. The proposed framework is applied on clinical data from 57 brain tumor patients (acquired by different scanners and acquisition parameters) and on 25 synthetic MR images with tumors. Assessment is performed against expert-defined tissue masks and is based on sensitivity analysis and Dice coefficient. Results The results demonstrate that the proposed multiparametric framework differentiates neoplastic tissues with accuracy similar to most current approaches while it achieves lower computational cost and higher degree of automation. Conclusion This study might provide a decision-support tool for neoplastic tissue segmentation, which can assist in treatment planning for tumor resection or focused radiotherapy.

AB - Abstract Objective Magnetic resonance imaging (MRI) is the primary imaging technique for evaluation of the brain tumor progression before and after radiotherapy or surgery. The purpose of the current study is to exploit conventional MR modalities in order to identify and segment brain images with neoplasms. Methods Four conventional MR sequences, namely, T1-weighted, gadolinium-enhanced T1-weighted, T2-weighted and fluid attenuation inversion recovery, are combined with machine learning techniques to extract global and local information of brain tissues and model the healthy and neoplastic imaging profiles. Healthy tissue clustering, outlier detection and geometric and spatial constraints are applied to perform a first segmentation which is further improved by a modified multiparametric Random Walker segmentation method. The proposed framework is applied on clinical data from 57 brain tumor patients (acquired by different scanners and acquisition parameters) and on 25 synthetic MR images with tumors. Assessment is performed against expert-defined tissue masks and is based on sensitivity analysis and Dice coefficient. Results The results demonstrate that the proposed multiparametric framework differentiates neoplastic tissues with accuracy similar to most current approaches while it achieves lower computational cost and higher degree of automation. Conclusion This study might provide a decision-support tool for neoplastic tissue segmentation, which can assist in treatment planning for tumor resection or focused radiotherapy.

KW - Brain neoplasms

KW - Magnetic resonance imaging (MRI)

KW - Outlier detection

KW - Random walks

KW - Tumor segmentation

UR - http://www.scopus.com/inward/record.url?scp=84938541046&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938541046&partnerID=8YFLogxK

U2 - 10.1016/j.bspc.2015.06.004

DO - 10.1016/j.bspc.2015.06.004

M3 - Article

AN - SCOPUS:84938541046

VL - 22

SP - 19

EP - 30

JO - Biomedical Signal Processing and Control

JF - Biomedical Signal Processing and Control

SN - 1746-8094

M1 - 698

ER -