COMPARE: Classification of morphological patterns using adaptive regional elements

Yong Fan, Dinggang Shen, Ruben C. Gur, Raquel E. Gur, Christos Davatzikos

Research output: Contribution to journalArticle

256 Citations (Scopus)

Abstract

This paper presents a method for classification of structural brain magnetic resonance (MR) images, by using a combination of deformation-based morphometry and machine learning methods. A morphological representation of the anatomy of interest is first obtained using a high-dimensional mass-preserving template warping method, which results in tissue density maps that constitute local tissue volumetric measurements. Regions that display strong correlations between tissue volume and classification (clinical) variables are extracted using a watershed segmentation algorithm, taking into account the regional smoothness of the correlation map which is estimated by a cross-validation strategy to achieve robustness to outliers. A volume increment algorithm is then applied to these regions to extract regional volumetric features, from which a feature selection technique using support vector machine (SVM)-based criteria is used to select the most discriminative features, according to their effect on the upper bound of the leave-one-out generalization error. Finally, SVM-based classification is applied using the best set of features, and it is tested using a leave-one-out cross-validation strategy. The results on MR brain images of healthy controls and schizophrenia patients demonstrate not only high classification accuracy (91.8% for female subjects and 90.8% for male subjects), but also good stability with respect to the number of features selected and the size of SVM kernel used.

Original languageEnglish
Pages (from-to)93-105
Number of pages13
JournalIEEE Transactions on Medical Imaging
Volume26
Issue number1
DOIs
Publication statusPublished - 2007 Jan 1
Externally publishedYes

Fingerprint

Support vector machines
Magnetic resonance
Tissue
Brain
Magnetic Resonance Spectroscopy
Watersheds
Learning systems
Feature extraction
Anatomy
Schizophrenia
Support Vector Machine
Machine Learning

Keywords

  • Feature selection
  • Morphological pattern analysis
  • Pattern classification
  • Regional feature extraction
  • Schizophrenia
  • Structural MRI
  • Support vector machines (SVM)

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

COMPARE : Classification of morphological patterns using adaptive regional elements. / Fan, Yong; Shen, Dinggang; Gur, Ruben C.; Gur, Raquel E.; Davatzikos, Christos.

In: IEEE Transactions on Medical Imaging, Vol. 26, No. 1, 01.01.2007, p. 93-105.

Research output: Contribution to journalArticle

Fan, Yong ; Shen, Dinggang ; Gur, Ruben C. ; Gur, Raquel E. ; Davatzikos, Christos. / COMPARE : Classification of morphological patterns using adaptive regional elements. In: IEEE Transactions on Medical Imaging. 2007 ; Vol. 26, No. 1. pp. 93-105.
@article{4c08832c91c248db9fe7c49ae65dfb24,
title = "COMPARE: Classification of morphological patterns using adaptive regional elements",
abstract = "This paper presents a method for classification of structural brain magnetic resonance (MR) images, by using a combination of deformation-based morphometry and machine learning methods. A morphological representation of the anatomy of interest is first obtained using a high-dimensional mass-preserving template warping method, which results in tissue density maps that constitute local tissue volumetric measurements. Regions that display strong correlations between tissue volume and classification (clinical) variables are extracted using a watershed segmentation algorithm, taking into account the regional smoothness of the correlation map which is estimated by a cross-validation strategy to achieve robustness to outliers. A volume increment algorithm is then applied to these regions to extract regional volumetric features, from which a feature selection technique using support vector machine (SVM)-based criteria is used to select the most discriminative features, according to their effect on the upper bound of the leave-one-out generalization error. Finally, SVM-based classification is applied using the best set of features, and it is tested using a leave-one-out cross-validation strategy. The results on MR brain images of healthy controls and schizophrenia patients demonstrate not only high classification accuracy (91.8{\%} for female subjects and 90.8{\%} for male subjects), but also good stability with respect to the number of features selected and the size of SVM kernel used.",
keywords = "Feature selection, Morphological pattern analysis, Pattern classification, Regional feature extraction, Schizophrenia, Structural MRI, Support vector machines (SVM)",
author = "Yong Fan and Dinggang Shen and Gur, {Ruben C.} and Gur, {Raquel E.} and Christos Davatzikos",
year = "2007",
month = "1",
day = "1",
doi = "10.1109/TMI.2006.886812",
language = "English",
volume = "26",
pages = "93--105",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

TY - JOUR

T1 - COMPARE

T2 - Classification of morphological patterns using adaptive regional elements

AU - Fan, Yong

AU - Shen, Dinggang

AU - Gur, Ruben C.

AU - Gur, Raquel E.

AU - Davatzikos, Christos

PY - 2007/1/1

Y1 - 2007/1/1

N2 - This paper presents a method for classification of structural brain magnetic resonance (MR) images, by using a combination of deformation-based morphometry and machine learning methods. A morphological representation of the anatomy of interest is first obtained using a high-dimensional mass-preserving template warping method, which results in tissue density maps that constitute local tissue volumetric measurements. Regions that display strong correlations between tissue volume and classification (clinical) variables are extracted using a watershed segmentation algorithm, taking into account the regional smoothness of the correlation map which is estimated by a cross-validation strategy to achieve robustness to outliers. A volume increment algorithm is then applied to these regions to extract regional volumetric features, from which a feature selection technique using support vector machine (SVM)-based criteria is used to select the most discriminative features, according to their effect on the upper bound of the leave-one-out generalization error. Finally, SVM-based classification is applied using the best set of features, and it is tested using a leave-one-out cross-validation strategy. The results on MR brain images of healthy controls and schizophrenia patients demonstrate not only high classification accuracy (91.8% for female subjects and 90.8% for male subjects), but also good stability with respect to the number of features selected and the size of SVM kernel used.

AB - This paper presents a method for classification of structural brain magnetic resonance (MR) images, by using a combination of deformation-based morphometry and machine learning methods. A morphological representation of the anatomy of interest is first obtained using a high-dimensional mass-preserving template warping method, which results in tissue density maps that constitute local tissue volumetric measurements. Regions that display strong correlations between tissue volume and classification (clinical) variables are extracted using a watershed segmentation algorithm, taking into account the regional smoothness of the correlation map which is estimated by a cross-validation strategy to achieve robustness to outliers. A volume increment algorithm is then applied to these regions to extract regional volumetric features, from which a feature selection technique using support vector machine (SVM)-based criteria is used to select the most discriminative features, according to their effect on the upper bound of the leave-one-out generalization error. Finally, SVM-based classification is applied using the best set of features, and it is tested using a leave-one-out cross-validation strategy. The results on MR brain images of healthy controls and schizophrenia patients demonstrate not only high classification accuracy (91.8% for female subjects and 90.8% for male subjects), but also good stability with respect to the number of features selected and the size of SVM kernel used.

KW - Feature selection

KW - Morphological pattern analysis

KW - Pattern classification

KW - Regional feature extraction

KW - Schizophrenia

KW - Structural MRI

KW - Support vector machines (SVM)

UR - http://www.scopus.com/inward/record.url?scp=33846264886&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33846264886&partnerID=8YFLogxK

U2 - 10.1109/TMI.2006.886812

DO - 10.1109/TMI.2006.886812

M3 - Article

C2 - 17243588

AN - SCOPUS:33846264886

VL - 26

SP - 93

EP - 105

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 1

ER -