Multimodal manifold-regularized transfer learning for MCI conversion prediction

Bo Cheng, Mingxia Liu, Heung-Il Suk, Dinggang Shen, Daoqiang Zhang

Research output: Contribution to journalArticle

34 Citations (Scopus)

Abstract

As the early stage of Alzheimer’s disease (AD), mild cognitive impairment (MCI) has high chance to convert to AD. Effective prediction of such conversion from MCI to AD is of great importance for early diagnosis of AD and also for evaluating AD risk pre-symptomatically. Unlike most previous methods that used only the samples from a target domain to train a classifier, in this paper, we propose a novel multimodal manifold-regularized transfer learning (M2TL) method that jointly utilizes samples from another domain (e.g., AD vs. normal controls (NC)) as well as unlabeled samples to boost the performance of the MCI conversion prediction. Specifically, the proposed M2TL method includes two key components. The first one is a kernel-based maximum mean discrepancy criterion, which helps eliminate the potential negative effect induced by the distributional difference between the auxiliary domain (i.e., AD and NC) and the target domain (i.e., MCI converters (MCI-C) and MCI non-converters (MCI-NC)). The second one is a semi-supervised multimodal manifold-regularized least squares classification method, where the target-domain samples, the auxiliary-domain samples, and the unlabeled samples can be jointly used for training our classifier. Furthermore, with the integration of a group sparsity constraint into our objective function, the proposed M2TL has a capability of selecting the informative samples to build a robust classifier. Experimental results on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database validate the effectiveness of the proposed method by significantly improving the classification accuracy of 80.1 % for MCI conversion prediction, and also outperforming the state-of-the-art methods.

Original languageEnglish
Pages (from-to)913-926
Number of pages14
JournalBrain Imaging and Behavior
Volume9
Issue number4
DOIs
Publication statusPublished - 2015 Dec 1

Fingerprint

Alzheimer Disease
Cognitive Dysfunction
Transfer (Psychology)
Least-Squares Analysis
Neuroimaging
Early Diagnosis
Databases

Keywords

  • Manifold regularization
  • Mild cognitive impairment conversion
  • Multimodal classification
  • Sample selection
  • Semi-supervised learning
  • Transfer learning

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Behavioral Neuroscience
  • Cellular and Molecular Neuroscience
  • Cognitive Neuroscience
  • Neurology
  • Psychiatry and Mental health
  • Clinical Neurology

Cite this

Multimodal manifold-regularized transfer learning for MCI conversion prediction. / Cheng, Bo; Liu, Mingxia; Suk, Heung-Il; Shen, Dinggang; Zhang, Daoqiang.

In: Brain Imaging and Behavior, Vol. 9, No. 4, 01.12.2015, p. 913-926.

Research output: Contribution to journalArticle

@article{204f3a843e3d4f24b87f081a54a25096,
title = "Multimodal manifold-regularized transfer learning for MCI conversion prediction",
abstract = "As the early stage of Alzheimer’s disease (AD), mild cognitive impairment (MCI) has high chance to convert to AD. Effective prediction of such conversion from MCI to AD is of great importance for early diagnosis of AD and also for evaluating AD risk pre-symptomatically. Unlike most previous methods that used only the samples from a target domain to train a classifier, in this paper, we propose a novel multimodal manifold-regularized transfer learning (M2TL) method that jointly utilizes samples from another domain (e.g., AD vs. normal controls (NC)) as well as unlabeled samples to boost the performance of the MCI conversion prediction. Specifically, the proposed M2TL method includes two key components. The first one is a kernel-based maximum mean discrepancy criterion, which helps eliminate the potential negative effect induced by the distributional difference between the auxiliary domain (i.e., AD and NC) and the target domain (i.e., MCI converters (MCI-C) and MCI non-converters (MCI-NC)). The second one is a semi-supervised multimodal manifold-regularized least squares classification method, where the target-domain samples, the auxiliary-domain samples, and the unlabeled samples can be jointly used for training our classifier. Furthermore, with the integration of a group sparsity constraint into our objective function, the proposed M2TL has a capability of selecting the informative samples to build a robust classifier. Experimental results on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database validate the effectiveness of the proposed method by significantly improving the classification accuracy of 80.1 {\%} for MCI conversion prediction, and also outperforming the state-of-the-art methods.",
keywords = "Manifold regularization, Mild cognitive impairment conversion, Multimodal classification, Sample selection, Semi-supervised learning, Transfer learning",
author = "Bo Cheng and Mingxia Liu and Heung-Il Suk and Dinggang Shen and Daoqiang Zhang",
year = "2015",
month = "12",
day = "1",
doi = "10.1007/s11682-015-9356-x",
language = "English",
volume = "9",
pages = "913--926",
journal = "Brain Imaging and Behavior",
issn = "1931-7557",
publisher = "Springer New York",
number = "4",

}

TY - JOUR

T1 - Multimodal manifold-regularized transfer learning for MCI conversion prediction

AU - Cheng, Bo

AU - Liu, Mingxia

AU - Suk, Heung-Il

AU - Shen, Dinggang

AU - Zhang, Daoqiang

PY - 2015/12/1

Y1 - 2015/12/1

N2 - As the early stage of Alzheimer’s disease (AD), mild cognitive impairment (MCI) has high chance to convert to AD. Effective prediction of such conversion from MCI to AD is of great importance for early diagnosis of AD and also for evaluating AD risk pre-symptomatically. Unlike most previous methods that used only the samples from a target domain to train a classifier, in this paper, we propose a novel multimodal manifold-regularized transfer learning (M2TL) method that jointly utilizes samples from another domain (e.g., AD vs. normal controls (NC)) as well as unlabeled samples to boost the performance of the MCI conversion prediction. Specifically, the proposed M2TL method includes two key components. The first one is a kernel-based maximum mean discrepancy criterion, which helps eliminate the potential negative effect induced by the distributional difference between the auxiliary domain (i.e., AD and NC) and the target domain (i.e., MCI converters (MCI-C) and MCI non-converters (MCI-NC)). The second one is a semi-supervised multimodal manifold-regularized least squares classification method, where the target-domain samples, the auxiliary-domain samples, and the unlabeled samples can be jointly used for training our classifier. Furthermore, with the integration of a group sparsity constraint into our objective function, the proposed M2TL has a capability of selecting the informative samples to build a robust classifier. Experimental results on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database validate the effectiveness of the proposed method by significantly improving the classification accuracy of 80.1 % for MCI conversion prediction, and also outperforming the state-of-the-art methods.

AB - As the early stage of Alzheimer’s disease (AD), mild cognitive impairment (MCI) has high chance to convert to AD. Effective prediction of such conversion from MCI to AD is of great importance for early diagnosis of AD and also for evaluating AD risk pre-symptomatically. Unlike most previous methods that used only the samples from a target domain to train a classifier, in this paper, we propose a novel multimodal manifold-regularized transfer learning (M2TL) method that jointly utilizes samples from another domain (e.g., AD vs. normal controls (NC)) as well as unlabeled samples to boost the performance of the MCI conversion prediction. Specifically, the proposed M2TL method includes two key components. The first one is a kernel-based maximum mean discrepancy criterion, which helps eliminate the potential negative effect induced by the distributional difference between the auxiliary domain (i.e., AD and NC) and the target domain (i.e., MCI converters (MCI-C) and MCI non-converters (MCI-NC)). The second one is a semi-supervised multimodal manifold-regularized least squares classification method, where the target-domain samples, the auxiliary-domain samples, and the unlabeled samples can be jointly used for training our classifier. Furthermore, with the integration of a group sparsity constraint into our objective function, the proposed M2TL has a capability of selecting the informative samples to build a robust classifier. Experimental results on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database validate the effectiveness of the proposed method by significantly improving the classification accuracy of 80.1 % for MCI conversion prediction, and also outperforming the state-of-the-art methods.

KW - Manifold regularization

KW - Mild cognitive impairment conversion

KW - Multimodal classification

KW - Sample selection

KW - Semi-supervised learning

KW - Transfer learning

UR - http://www.scopus.com/inward/record.url?scp=84948567283&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84948567283&partnerID=8YFLogxK

U2 - 10.1007/s11682-015-9356-x

DO - 10.1007/s11682-015-9356-x

M3 - Article

C2 - 25702248

AN - SCOPUS:84948567283

VL - 9

SP - 913

EP - 926

JO - Brain Imaging and Behavior

JF - Brain Imaging and Behavior

SN - 1931-7557

IS - 4

ER -