Clustering-induced multi-task learning for AD/MCI classification.

Heung Ii Suk, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

In this work, we formulate a clustering-induced multi-task learning method for feature selection in Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI) diagnosis. Unlike the previous methods that often assumed a unimodal data distribution, we take into account the underlying multipeak distribution of classes. The rationale for our approach is that it is likely for neuroimaging data to have multiple peaks or modes in distribution due to the inter-subject variability. In this regard, we use a clustering method to discover the multipeak distributional characteristics and define subclasses based on the clustering results, in which each cluster covers a peak. We then encode the respective subclasses, i.e., clusters, with their unique codes by imposing the subclasses of the same original class close to each other and those of different original classes L2,1-penalized regression framework by taking the codes as new label vectors of our training samples, through which we select features for classification. In our experimental results on the ADNI dataset, we validated the effectiveness of the proposed method by achieving the maximal classification accuracies of 95.18% (AD/Normal Control: NC), 79.52% (MCI/NC), and 72.02% (MCI converter/MCl non-converter), outperforming the competing single-task learning method.

Original languageEnglish
Title of host publicationMedical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
Pages393-400
Number of pages8
Volume17
EditionPt 3
Publication statusPublished - 2014 Jan 1

Fingerprint

Cluster Analysis
Alzheimer Disease
Learning
Neuroimaging
Cognitive Dysfunction

ASJC Scopus subject areas

  • Medicine(all)

Cite this

Suk, H. I., & Shen, D. (2014). Clustering-induced multi-task learning for AD/MCI classification. In Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention (Pt 3 ed., Vol. 17, pp. 393-400)

Clustering-induced multi-task learning for AD/MCI classification. / Suk, Heung Ii; Shen, Dinggang.

Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Vol. 17 Pt 3. ed. 2014. p. 393-400.

Research output: Chapter in Book/Report/Conference proceedingChapter

Suk, HI & Shen, D 2014, Clustering-induced multi-task learning for AD/MCI classification. in Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Pt 3 edn, vol. 17, pp. 393-400.
Suk HI, Shen D. Clustering-induced multi-task learning for AD/MCI classification. In Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Pt 3 ed. Vol. 17. 2014. p. 393-400
Suk, Heung Ii ; Shen, Dinggang. / Clustering-induced multi-task learning for AD/MCI classification. Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention. Vol. 17 Pt 3. ed. 2014. pp. 393-400
@inbook{832c9458c11f4afaa5d14f2e85093dff,
title = "Clustering-induced multi-task learning for AD/MCI classification.",
abstract = "In this work, we formulate a clustering-induced multi-task learning method for feature selection in Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI) diagnosis. Unlike the previous methods that often assumed a unimodal data distribution, we take into account the underlying multipeak distribution of classes. The rationale for our approach is that it is likely for neuroimaging data to have multiple peaks or modes in distribution due to the inter-subject variability. In this regard, we use a clustering method to discover the multipeak distributional characteristics and define subclasses based on the clustering results, in which each cluster covers a peak. We then encode the respective subclasses, i.e., clusters, with their unique codes by imposing the subclasses of the same original class close to each other and those of different original classes L2,1-penalized regression framework by taking the codes as new label vectors of our training samples, through which we select features for classification. In our experimental results on the ADNI dataset, we validated the effectiveness of the proposed method by achieving the maximal classification accuracies of 95.18{\%} (AD/Normal Control: NC), 79.52{\%} (MCI/NC), and 72.02{\%} (MCI converter/MCl non-converter), outperforming the competing single-task learning method.",
author = "Suk, {Heung Ii} and Dinggang Shen",
year = "2014",
month = "1",
day = "1",
language = "English",
volume = "17",
pages = "393--400",
booktitle = "Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention",
edition = "Pt 3",

}

TY - CHAP

T1 - Clustering-induced multi-task learning for AD/MCI classification.

AU - Suk, Heung Ii

AU - Shen, Dinggang

PY - 2014/1/1

Y1 - 2014/1/1

N2 - In this work, we formulate a clustering-induced multi-task learning method for feature selection in Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI) diagnosis. Unlike the previous methods that often assumed a unimodal data distribution, we take into account the underlying multipeak distribution of classes. The rationale for our approach is that it is likely for neuroimaging data to have multiple peaks or modes in distribution due to the inter-subject variability. In this regard, we use a clustering method to discover the multipeak distributional characteristics and define subclasses based on the clustering results, in which each cluster covers a peak. We then encode the respective subclasses, i.e., clusters, with their unique codes by imposing the subclasses of the same original class close to each other and those of different original classes L2,1-penalized regression framework by taking the codes as new label vectors of our training samples, through which we select features for classification. In our experimental results on the ADNI dataset, we validated the effectiveness of the proposed method by achieving the maximal classification accuracies of 95.18% (AD/Normal Control: NC), 79.52% (MCI/NC), and 72.02% (MCI converter/MCl non-converter), outperforming the competing single-task learning method.

AB - In this work, we formulate a clustering-induced multi-task learning method for feature selection in Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI) diagnosis. Unlike the previous methods that often assumed a unimodal data distribution, we take into account the underlying multipeak distribution of classes. The rationale for our approach is that it is likely for neuroimaging data to have multiple peaks or modes in distribution due to the inter-subject variability. In this regard, we use a clustering method to discover the multipeak distributional characteristics and define subclasses based on the clustering results, in which each cluster covers a peak. We then encode the respective subclasses, i.e., clusters, with their unique codes by imposing the subclasses of the same original class close to each other and those of different original classes L2,1-penalized regression framework by taking the codes as new label vectors of our training samples, through which we select features for classification. In our experimental results on the ADNI dataset, we validated the effectiveness of the proposed method by achieving the maximal classification accuracies of 95.18% (AD/Normal Control: NC), 79.52% (MCI/NC), and 72.02% (MCI converter/MCl non-converter), outperforming the competing single-task learning method.

UR - http://www.scopus.com/inward/record.url?scp=84909599282&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84909599282&partnerID=8YFLogxK

M3 - Chapter

C2 - 25320824

VL - 17

SP - 393

EP - 400

BT - Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention

ER -