A novel multi-relation regularization method for regression and classification in AD diagnosis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

21 Citations (Scopus)

Abstract

In this paper, we consider the joint regression and classification in Alzheimer's disease diagnosis and propose a novel multi-relation regularization method that exploits the relational information inherent in the observations and then combines it with an ℓ2,1-norm within a least square regression framework for feature selection. Specifically, we use three kinds of relationships: feature-feature relation, response-response relation, and sample-sample relation. By imposing these three relational characteristics along with the ℓ2,1-norm on the weight coefficients, we formulate a new objective function. After feature selection based on the optimal weight coefficients, we train two support vector regression models to predict the clinical scores of Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-Cog) and Mini-Mental State Examination (MMSE), respectively, and a support vector classification model to identify the clinical label. We conducted clinical score prediction and disease status identification jointly on the Alzheimer's Disease Neuroimaging Initiative dataset. The experimental results showed that the proposed regularization method outperforms the state-of-the-art methods, in the metrics of correlation coefficient and root mean squared error in regression and classification accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve in classification.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages401-408
Number of pages8
Volume8675 LNCS
EditionPART 3
ISBN (Print)9783319104423
DOIs
Publication statusPublished - 2014 Jan 1
Externally publishedYes
Event17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014 - Boston, MA, United States
Duration: 2014 Sep 142014 Sep 18

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume8675 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014
CountryUnited States
CityBoston, MA
Period14/9/1414/9/18

Fingerprint

Regularization Method
Alzheimer's Disease
Regression
Weight Coefficient
Feature Selection
Feature extraction
Neuroimaging
Norm
Least Squares Regression
Support Vector
Support Vector Regression
Receiver Operating Characteristic Curve
Mean Squared Error
Correlation coefficient
Specificity
Labels
Regression Model
Objective function
Roots
Metric

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Zhu, X., Suk, H-I., & Shen, D. (2014). A novel multi-relation regularization method for regression and classification in AD diagnosis. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (PART 3 ed., Vol. 8675 LNCS, pp. 401-408). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8675 LNCS, No. PART 3). Springer Verlag. https://doi.org/10.1007/978-3-319-10443-0_51

A novel multi-relation regularization method for regression and classification in AD diagnosis. / Zhu, Xiaofeng; Suk, Heung-Il; Shen, Dinggang.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8675 LNCS PART 3. ed. Springer Verlag, 2014. p. 401-408 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8675 LNCS, No. PART 3).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zhu, X, Suk, H-I & Shen, D 2014, A novel multi-relation regularization method for regression and classification in AD diagnosis. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 edn, vol. 8675 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 8675 LNCS, Springer Verlag, pp. 401-408, 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014, Boston, MA, United States, 14/9/14. https://doi.org/10.1007/978-3-319-10443-0_51
Zhu X, Suk H-I, Shen D. A novel multi-relation regularization method for regression and classification in AD diagnosis. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 ed. Vol. 8675 LNCS. Springer Verlag. 2014. p. 401-408. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3). https://doi.org/10.1007/978-3-319-10443-0_51
Zhu, Xiaofeng ; Suk, Heung-Il ; Shen, Dinggang. / A novel multi-relation regularization method for regression and classification in AD diagnosis. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8675 LNCS PART 3. ed. Springer Verlag, 2014. pp. 401-408 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
@inproceedings{61c0c069594447ec9b8e0fc68c92dda5,
title = "A novel multi-relation regularization method for regression and classification in AD diagnosis",
abstract = "In this paper, we consider the joint regression and classification in Alzheimer's disease diagnosis and propose a novel multi-relation regularization method that exploits the relational information inherent in the observations and then combines it with an ℓ2,1-norm within a least square regression framework for feature selection. Specifically, we use three kinds of relationships: feature-feature relation, response-response relation, and sample-sample relation. By imposing these three relational characteristics along with the ℓ2,1-norm on the weight coefficients, we formulate a new objective function. After feature selection based on the optimal weight coefficients, we train two support vector regression models to predict the clinical scores of Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-Cog) and Mini-Mental State Examination (MMSE), respectively, and a support vector classification model to identify the clinical label. We conducted clinical score prediction and disease status identification jointly on the Alzheimer's Disease Neuroimaging Initiative dataset. The experimental results showed that the proposed regularization method outperforms the state-of-the-art methods, in the metrics of correlation coefficient and root mean squared error in regression and classification accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve in classification.",
keywords = "Alzheimer's disease, feature selection, manifold learning, MCI conversion, sparse coding",
author = "Xiaofeng Zhu and Heung-Il Suk and Dinggang Shen",
year = "2014",
month = "1",
day = "1",
doi = "10.1007/978-3-319-10443-0_51",
language = "English",
isbn = "9783319104423",
volume = "8675 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
number = "PART 3",
pages = "401--408",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
edition = "PART 3",

}

TY - GEN

T1 - A novel multi-relation regularization method for regression and classification in AD diagnosis

AU - Zhu, Xiaofeng

AU - Suk, Heung-Il

AU - Shen, Dinggang

PY - 2014/1/1

Y1 - 2014/1/1

N2 - In this paper, we consider the joint regression and classification in Alzheimer's disease diagnosis and propose a novel multi-relation regularization method that exploits the relational information inherent in the observations and then combines it with an ℓ2,1-norm within a least square regression framework for feature selection. Specifically, we use three kinds of relationships: feature-feature relation, response-response relation, and sample-sample relation. By imposing these three relational characteristics along with the ℓ2,1-norm on the weight coefficients, we formulate a new objective function. After feature selection based on the optimal weight coefficients, we train two support vector regression models to predict the clinical scores of Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-Cog) and Mini-Mental State Examination (MMSE), respectively, and a support vector classification model to identify the clinical label. We conducted clinical score prediction and disease status identification jointly on the Alzheimer's Disease Neuroimaging Initiative dataset. The experimental results showed that the proposed regularization method outperforms the state-of-the-art methods, in the metrics of correlation coefficient and root mean squared error in regression and classification accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve in classification.

AB - In this paper, we consider the joint regression and classification in Alzheimer's disease diagnosis and propose a novel multi-relation regularization method that exploits the relational information inherent in the observations and then combines it with an ℓ2,1-norm within a least square regression framework for feature selection. Specifically, we use three kinds of relationships: feature-feature relation, response-response relation, and sample-sample relation. By imposing these three relational characteristics along with the ℓ2,1-norm on the weight coefficients, we formulate a new objective function. After feature selection based on the optimal weight coefficients, we train two support vector regression models to predict the clinical scores of Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-Cog) and Mini-Mental State Examination (MMSE), respectively, and a support vector classification model to identify the clinical label. We conducted clinical score prediction and disease status identification jointly on the Alzheimer's Disease Neuroimaging Initiative dataset. The experimental results showed that the proposed regularization method outperforms the state-of-the-art methods, in the metrics of correlation coefficient and root mean squared error in regression and classification accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve in classification.

KW - Alzheimer's disease

KW - feature selection

KW - manifold learning

KW - MCI conversion

KW - sparse coding

UR - http://www.scopus.com/inward/record.url?scp=84906975457&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84906975457&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-10443-0_51

DO - 10.1007/978-3-319-10443-0_51

M3 - Conference contribution

C2 - 25320825

AN - SCOPUS:84906975457

SN - 9783319104423

VL - 8675 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 401

EP - 408

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -