A surface-based approach for classification of 3D neuroanatomic structures

Li Shen, James Ford, Fillia Makedon, Andrew Saykin

Research output: Contribution to journalArticle

62 Citations (Scopus)

Abstract

We present a new framework for 3D surface object classification that combines a powerful shape description method with suitable pattern classification techniques. Spherical harmonic parameterization and normalization techniques are used to describe a surface shape and derive a dual high dimensional landmark representation. A point distribution model is applied to reduce the dimensionality. Fisher's linear discriminants and support vector machines are used for classification. Several feature selection schemes are proposed for learning better classifiers. After showing the effectiveness of this framework using simulated shape data, we apply it to real hippocampal data in schizophrenia and perform extensive experimental studies by examining different combinations of techniques. We achieve best leave-one-out cross-validation accuracies of 93% (whole set, N = 56) and 90% (right-handed males, N = 39), respectively, which are competitive with the best results in previous studies using different techniques on similar types of data. Furthermore, to help medical diagnosis in practice, we employ a threshold-free receiver operating characteristic (ROC) approach as an alternative evaluation of classification results as well as propose a new method for visualizing discriminative patterns.

Original languageEnglish (US)
Pages (from-to)519-542
Number of pages24
JournalIntelligent Data Analysis
Volume8
Issue number6
StatePublished - 2004
Externally publishedYes

Fingerprint

Parameterization
Object Classification
Pattern recognition
Support vector machines
Right handed
Feature extraction
Pattern Classification
Operating Characteristics
Spherical Harmonics
Classifiers
Landmarks
Discriminant
Cross-validation
Feature Selection
Normalization
Dimensionality
Experimental Study
Support Vector Machine
High-dimensional
Receiver

Keywords

  • classification
  • feature selection
  • shape analysis
  • statistical pattern recognition

ASJC Scopus subject areas

  • Artificial Intelligence
  • Theoretical Computer Science
  • Computer Vision and Pattern Recognition

Cite this

A surface-based approach for classification of 3D neuroanatomic structures. / Shen, Li; Ford, James; Makedon, Fillia; Saykin, Andrew.

In: Intelligent Data Analysis, Vol. 8, No. 6, 2004, p. 519-542.

Research output: Contribution to journalArticle

@article{a11fd8681b5a4c4aa7f4afa246bb58fa,
title = "A surface-based approach for classification of 3D neuroanatomic structures",
abstract = "We present a new framework for 3D surface object classification that combines a powerful shape description method with suitable pattern classification techniques. Spherical harmonic parameterization and normalization techniques are used to describe a surface shape and derive a dual high dimensional landmark representation. A point distribution model is applied to reduce the dimensionality. Fisher's linear discriminants and support vector machines are used for classification. Several feature selection schemes are proposed for learning better classifiers. After showing the effectiveness of this framework using simulated shape data, we apply it to real hippocampal data in schizophrenia and perform extensive experimental studies by examining different combinations of techniques. We achieve best leave-one-out cross-validation accuracies of 93{\%} (whole set, N = 56) and 90{\%} (right-handed males, N = 39), respectively, which are competitive with the best results in previous studies using different techniques on similar types of data. Furthermore, to help medical diagnosis in practice, we employ a threshold-free receiver operating characteristic (ROC) approach as an alternative evaluation of classification results as well as propose a new method for visualizing discriminative patterns.",
keywords = "classification, feature selection, shape analysis, statistical pattern recognition",
author = "Li Shen and James Ford and Fillia Makedon and Andrew Saykin",
year = "2004",
language = "English (US)",
volume = "8",
pages = "519--542",
journal = "Intelligent Data Analysis",
issn = "1088-467X",
publisher = "IOS Press",
number = "6",

}

TY - JOUR

T1 - A surface-based approach for classification of 3D neuroanatomic structures

AU - Shen, Li

AU - Ford, James

AU - Makedon, Fillia

AU - Saykin, Andrew

PY - 2004

Y1 - 2004

N2 - We present a new framework for 3D surface object classification that combines a powerful shape description method with suitable pattern classification techniques. Spherical harmonic parameterization and normalization techniques are used to describe a surface shape and derive a dual high dimensional landmark representation. A point distribution model is applied to reduce the dimensionality. Fisher's linear discriminants and support vector machines are used for classification. Several feature selection schemes are proposed for learning better classifiers. After showing the effectiveness of this framework using simulated shape data, we apply it to real hippocampal data in schizophrenia and perform extensive experimental studies by examining different combinations of techniques. We achieve best leave-one-out cross-validation accuracies of 93% (whole set, N = 56) and 90% (right-handed males, N = 39), respectively, which are competitive with the best results in previous studies using different techniques on similar types of data. Furthermore, to help medical diagnosis in practice, we employ a threshold-free receiver operating characteristic (ROC) approach as an alternative evaluation of classification results as well as propose a new method for visualizing discriminative patterns.

AB - We present a new framework for 3D surface object classification that combines a powerful shape description method with suitable pattern classification techniques. Spherical harmonic parameterization and normalization techniques are used to describe a surface shape and derive a dual high dimensional landmark representation. A point distribution model is applied to reduce the dimensionality. Fisher's linear discriminants and support vector machines are used for classification. Several feature selection schemes are proposed for learning better classifiers. After showing the effectiveness of this framework using simulated shape data, we apply it to real hippocampal data in schizophrenia and perform extensive experimental studies by examining different combinations of techniques. We achieve best leave-one-out cross-validation accuracies of 93% (whole set, N = 56) and 90% (right-handed males, N = 39), respectively, which are competitive with the best results in previous studies using different techniques on similar types of data. Furthermore, to help medical diagnosis in practice, we employ a threshold-free receiver operating characteristic (ROC) approach as an alternative evaluation of classification results as well as propose a new method for visualizing discriminative patterns.

KW - classification

KW - feature selection

KW - shape analysis

KW - statistical pattern recognition

UR - http://www.scopus.com/inward/record.url?scp=3042586437&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=3042586437&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:3042586437

VL - 8

SP - 519

EP - 542

JO - Intelligent Data Analysis

JF - Intelligent Data Analysis

SN - 1088-467X

IS - 6

ER -