Dimensionality reduced cortical features and their use in the classification of Alzheimer's disease and mild cognitive impairment

Hyunjin Park, Jin Ju Yang, Jongbum Seo, Jong Min Lee

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

Features defined on the cortical surface derived from magnetic resonance imaging provide important information to distinguish normal controls from Alzheimer's disease (AD) and mild cognitive impairment (MCI). We adopted cortical thickness and sulcal depth, parameterized by three dimensional meshes, as our feature. The cortical feature is high dimensional and direct use of it is problematic in a modern classifier due to small sample size problem. We applied manifold learning to reduce the dimensionality of the feature and then tested the usage of the dimensionality reduced feature with a support vector machine classifier. A leave-one-out cross-validation was adopted for quantifying classifier performance. We chose principal component analysis (PCA) as the manifold learning method. We applied PCA to a region of interest within the cortical surface. Our classification performance was at least on par for the AD/normal and MCI/normal groups and significantly better for the AD/MCI groups compared to recent studies. Our approach was tested using 25 AD, 25 MCI, and 50 normal control patients from the OASIS database.

Original languageEnglish
Pages (from-to)123-127
Number of pages5
JournalNeuroscience Letters
Volume529
Issue number2
DOIs
Publication statusPublished - 2012 Nov 7

Fingerprint

Alzheimer Disease
Principal Component Analysis
Learning
Sample Size
Magnetic Resonance Imaging
Databases
Cognitive Dysfunction

All Science Journal Classification (ASJC) codes

  • Neuroscience(all)

Cite this

@article{4bc1c4e19fdd4677abab896f4ef82ad2,
title = "Dimensionality reduced cortical features and their use in the classification of Alzheimer's disease and mild cognitive impairment",
abstract = "Features defined on the cortical surface derived from magnetic resonance imaging provide important information to distinguish normal controls from Alzheimer's disease (AD) and mild cognitive impairment (MCI). We adopted cortical thickness and sulcal depth, parameterized by three dimensional meshes, as our feature. The cortical feature is high dimensional and direct use of it is problematic in a modern classifier due to small sample size problem. We applied manifold learning to reduce the dimensionality of the feature and then tested the usage of the dimensionality reduced feature with a support vector machine classifier. A leave-one-out cross-validation was adopted for quantifying classifier performance. We chose principal component analysis (PCA) as the manifold learning method. We applied PCA to a region of interest within the cortical surface. Our classification performance was at least on par for the AD/normal and MCI/normal groups and significantly better for the AD/MCI groups compared to recent studies. Our approach was tested using 25 AD, 25 MCI, and 50 normal control patients from the OASIS database.",
author = "Hyunjin Park and Yang, {Jin Ju} and Jongbum Seo and Lee, {Jong Min}",
year = "2012",
month = "11",
day = "7",
doi = "10.1016/j.neulet.2012.09.011",
language = "English",
volume = "529",
pages = "123--127",
journal = "Neuroscience Letters",
issn = "0304-3940",
publisher = "Elsevier Ireland Ltd",
number = "2",

}

Dimensionality reduced cortical features and their use in the classification of Alzheimer's disease and mild cognitive impairment. / Park, Hyunjin; Yang, Jin Ju; Seo, Jongbum; Lee, Jong Min.

In: Neuroscience Letters, Vol. 529, No. 2, 07.11.2012, p. 123-127.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Dimensionality reduced cortical features and their use in the classification of Alzheimer's disease and mild cognitive impairment

AU - Park, Hyunjin

AU - Yang, Jin Ju

AU - Seo, Jongbum

AU - Lee, Jong Min

PY - 2012/11/7

Y1 - 2012/11/7

N2 - Features defined on the cortical surface derived from magnetic resonance imaging provide important information to distinguish normal controls from Alzheimer's disease (AD) and mild cognitive impairment (MCI). We adopted cortical thickness and sulcal depth, parameterized by three dimensional meshes, as our feature. The cortical feature is high dimensional and direct use of it is problematic in a modern classifier due to small sample size problem. We applied manifold learning to reduce the dimensionality of the feature and then tested the usage of the dimensionality reduced feature with a support vector machine classifier. A leave-one-out cross-validation was adopted for quantifying classifier performance. We chose principal component analysis (PCA) as the manifold learning method. We applied PCA to a region of interest within the cortical surface. Our classification performance was at least on par for the AD/normal and MCI/normal groups and significantly better for the AD/MCI groups compared to recent studies. Our approach was tested using 25 AD, 25 MCI, and 50 normal control patients from the OASIS database.

AB - Features defined on the cortical surface derived from magnetic resonance imaging provide important information to distinguish normal controls from Alzheimer's disease (AD) and mild cognitive impairment (MCI). We adopted cortical thickness and sulcal depth, parameterized by three dimensional meshes, as our feature. The cortical feature is high dimensional and direct use of it is problematic in a modern classifier due to small sample size problem. We applied manifold learning to reduce the dimensionality of the feature and then tested the usage of the dimensionality reduced feature with a support vector machine classifier. A leave-one-out cross-validation was adopted for quantifying classifier performance. We chose principal component analysis (PCA) as the manifold learning method. We applied PCA to a region of interest within the cortical surface. Our classification performance was at least on par for the AD/normal and MCI/normal groups and significantly better for the AD/MCI groups compared to recent studies. Our approach was tested using 25 AD, 25 MCI, and 50 normal control patients from the OASIS database.

UR - http://www.scopus.com/inward/record.url?scp=84868000111&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868000111&partnerID=8YFLogxK

U2 - 10.1016/j.neulet.2012.09.011

DO - 10.1016/j.neulet.2012.09.011

M3 - Article

C2 - 23000551

AN - SCOPUS:84868000111

VL - 529

SP - 123

EP - 127

JO - Neuroscience Letters

JF - Neuroscience Letters

SN - 0304-3940

IS - 2

ER -