Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme

Evangelia I. Zacharaki, Sumei Wang, Sanjeev Chawla, Dong Soo Yoo, Ronald Wolf, Elias R. Melhem, Christos Davatzikos

Research output: Contribution to journalArticle

Abstract

The objective of this study is to investigate the use of pattern classification methods for distinguishing different types of brain tumors, such as primary gliomas from metastases, and also for grading of gliomas. The availability of an automated computer analysis tool that is more objective than human readers can potentially lead to more reliable and reproducible brain tumor diagnostic procedures. A computer-assisted classification method combining conventional MRI and perfusion MRI is developed and used for differential diagnosis. The proposed scheme consists of several steps including region-of-interest definition, feature extraction, feature selection, and classification. The extracted features include tumor shape and intensity characteristics, as well as rotation invariant texture features. Feature subset selection is performed using support vector machines with recursive feature elimination. The method was applied on a population of 102 brain tumors histologically diagnosed as metastasis (24), meningiomas (4), gliomas World Health Organization grade II (22), gliomas World Health Organization grade III (18), and glioblastomas (34). The binary support vector machine classification accuracy, sensitivity, and specificity, assessed by leave-one-out cross-validation, were, respectively, 85%, 87%, and 79% for discrimination of metastases from gliomas and 88%, 85%, and 96% for discrimination of high-grade (grades III and IV) from low-grade (grade II) neoplasms. Multiclass classification was also performed via a one-vs-all voting scheme.

Original languageEnglish (US)
Pages (from-to)1609-1618
Number of pages10
JournalMagnetic Resonance in Medicine
Volume62
Issue number6
DOIs
StatePublished - Dec 2009
Externally publishedYes

Fingerprint

Brain Neoplasms
Glioma
Neoplasm Metastasis
Meningioma
Politics
Glioblastoma
Neoplasms
Differential Diagnosis
Perfusion
Machine Learning
Sensitivity and Specificity
Population
Support Vector Machine

Keywords

  • Brain tumor
  • Classification
  • Feature selection
  • MRI
  • SVM
  • Texture
  • Tumor grade

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this

Zacharaki, E. I., Wang, S., Chawla, S., Yoo, D. S., Wolf, R., Melhem, E. R., & Davatzikos, C. (2009). Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme. Magnetic Resonance in Medicine, 62(6), 1609-1618. https://doi.org/10.1002/mrm.22147

Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme. / Zacharaki, Evangelia I.; Wang, Sumei; Chawla, Sanjeev; Yoo, Dong Soo; Wolf, Ronald; Melhem, Elias R.; Davatzikos, Christos.

In: Magnetic Resonance in Medicine, Vol. 62, No. 6, 12.2009, p. 1609-1618.

Research output: Contribution to journalArticle

Zacharaki, EI, Wang, S, Chawla, S, Yoo, DS, Wolf, R, Melhem, ER & Davatzikos, C 2009, 'Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme', Magnetic Resonance in Medicine, vol. 62, no. 6, pp. 1609-1618. https://doi.org/10.1002/mrm.22147
Zacharaki, Evangelia I. ; Wang, Sumei ; Chawla, Sanjeev ; Yoo, Dong Soo ; Wolf, Ronald ; Melhem, Elias R. ; Davatzikos, Christos. / Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme. In: Magnetic Resonance in Medicine. 2009 ; Vol. 62, No. 6. pp. 1609-1618.
@article{8ba1ad34862b4324a6e5fb8124291969,
title = "Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme",
abstract = "The objective of this study is to investigate the use of pattern classification methods for distinguishing different types of brain tumors, such as primary gliomas from metastases, and also for grading of gliomas. The availability of an automated computer analysis tool that is more objective than human readers can potentially lead to more reliable and reproducible brain tumor diagnostic procedures. A computer-assisted classification method combining conventional MRI and perfusion MRI is developed and used for differential diagnosis. The proposed scheme consists of several steps including region-of-interest definition, feature extraction, feature selection, and classification. The extracted features include tumor shape and intensity characteristics, as well as rotation invariant texture features. Feature subset selection is performed using support vector machines with recursive feature elimination. The method was applied on a population of 102 brain tumors histologically diagnosed as metastasis (24), meningiomas (4), gliomas World Health Organization grade II (22), gliomas World Health Organization grade III (18), and glioblastomas (34). The binary support vector machine classification accuracy, sensitivity, and specificity, assessed by leave-one-out cross-validation, were, respectively, 85{\%}, 87{\%}, and 79{\%} for discrimination of metastases from gliomas and 88{\%}, 85{\%}, and 96{\%} for discrimination of high-grade (grades III and IV) from low-grade (grade II) neoplasms. Multiclass classification was also performed via a one-vs-all voting scheme.",
keywords = "Brain tumor, Classification, Feature selection, MRI, SVM, Texture, Tumor grade",
author = "Zacharaki, {Evangelia I.} and Sumei Wang and Sanjeev Chawla and Yoo, {Dong Soo} and Ronald Wolf and Melhem, {Elias R.} and Christos Davatzikos",
year = "2009",
month = "12",
doi = "10.1002/mrm.22147",
language = "English (US)",
volume = "62",
pages = "1609--1618",
journal = "Magnetic Resonance in Medicine",
issn = "0740-3194",
publisher = "John Wiley and Sons Inc.",
number = "6",

}

TY - JOUR

T1 - Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme

AU - Zacharaki, Evangelia I.

AU - Wang, Sumei

AU - Chawla, Sanjeev

AU - Yoo, Dong Soo

AU - Wolf, Ronald

AU - Melhem, Elias R.

AU - Davatzikos, Christos

PY - 2009/12

Y1 - 2009/12

N2 - The objective of this study is to investigate the use of pattern classification methods for distinguishing different types of brain tumors, such as primary gliomas from metastases, and also for grading of gliomas. The availability of an automated computer analysis tool that is more objective than human readers can potentially lead to more reliable and reproducible brain tumor diagnostic procedures. A computer-assisted classification method combining conventional MRI and perfusion MRI is developed and used for differential diagnosis. The proposed scheme consists of several steps including region-of-interest definition, feature extraction, feature selection, and classification. The extracted features include tumor shape and intensity characteristics, as well as rotation invariant texture features. Feature subset selection is performed using support vector machines with recursive feature elimination. The method was applied on a population of 102 brain tumors histologically diagnosed as metastasis (24), meningiomas (4), gliomas World Health Organization grade II (22), gliomas World Health Organization grade III (18), and glioblastomas (34). The binary support vector machine classification accuracy, sensitivity, and specificity, assessed by leave-one-out cross-validation, were, respectively, 85%, 87%, and 79% for discrimination of metastases from gliomas and 88%, 85%, and 96% for discrimination of high-grade (grades III and IV) from low-grade (grade II) neoplasms. Multiclass classification was also performed via a one-vs-all voting scheme.

AB - The objective of this study is to investigate the use of pattern classification methods for distinguishing different types of brain tumors, such as primary gliomas from metastases, and also for grading of gliomas. The availability of an automated computer analysis tool that is more objective than human readers can potentially lead to more reliable and reproducible brain tumor diagnostic procedures. A computer-assisted classification method combining conventional MRI and perfusion MRI is developed and used for differential diagnosis. The proposed scheme consists of several steps including region-of-interest definition, feature extraction, feature selection, and classification. The extracted features include tumor shape and intensity characteristics, as well as rotation invariant texture features. Feature subset selection is performed using support vector machines with recursive feature elimination. The method was applied on a population of 102 brain tumors histologically diagnosed as metastasis (24), meningiomas (4), gliomas World Health Organization grade II (22), gliomas World Health Organization grade III (18), and glioblastomas (34). The binary support vector machine classification accuracy, sensitivity, and specificity, assessed by leave-one-out cross-validation, were, respectively, 85%, 87%, and 79% for discrimination of metastases from gliomas and 88%, 85%, and 96% for discrimination of high-grade (grades III and IV) from low-grade (grade II) neoplasms. Multiclass classification was also performed via a one-vs-all voting scheme.

KW - Brain tumor

KW - Classification

KW - Feature selection

KW - MRI

KW - SVM

KW - Texture

KW - Tumor grade

UR - http://www.scopus.com/inward/record.url?scp=73149086838&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=73149086838&partnerID=8YFLogxK

U2 - 10.1002/mrm.22147

DO - 10.1002/mrm.22147

M3 - Article

VL - 62

SP - 1609

EP - 1618

JO - Magnetic Resonance in Medicine

JF - Magnetic Resonance in Medicine

SN - 0740-3194

IS - 6

ER -