Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease

Daoqiang Zhang, Dinggang Shen

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

5 Citations (Scopus)

Abstract

One recent interest in computer-aided diagnosis of neurological diseases is to predict the clinical scores from brain images. Most existing methods usually estimate multiple clinical variables separately, without considering the useful correlation information among them. On the other hand, nearly all methods use only one modality of data (mostly structural MRI) for regression, and thus ignore the complementary information among different modalities. To address these issues, in this paper, we present a general methodology, namely Multi-Modal Multi-Task (M3T) learning, to jointly predict multiple variables from multi-modal data. Our method contains three major subsequent steps: (1) a multi-task feature selection which selects the common subset of relevant features for the related multiple clinical variables from each modality; (2) a kernel-based multimodal data fusion which fuses the above-selected features from all modalities; (3) a support vector regression which predicts multiple clinical variables based on the previously learnt mixed kernel. Experimental results on ADNI dataset with both imaging modalities (MRI and PET) and biological modality (CSF) validate the efficacy of the proposed M3T learning method.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages60-67
Number of pages8
Volume7012 LNCS
DOIs
Publication statusPublished - 2011 Oct 11
Externally publishedYes
Event1st International Workshop on Multimodal Brain Image Analysis, MBIA 2011, in Conjunction with the 14th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2011 - Toronto, ON, Canada
Duration: 2011 Sep 182011 Sep 18

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7012 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other1st International Workshop on Multimodal Brain Image Analysis, MBIA 2011, in Conjunction with the 14th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2011
CountryCanada
CityToronto, ON
Period11/9/1811/9/18

Fingerprint

Multi-task Learning
Alzheimer's Disease
Magnetic resonance imaging
Modality
Computer aided diagnosis
Prediction
Data fusion
Electric fuses
Feature extraction
Brain
Imaging techniques
Predict
kernel
Computer-aided Diagnosis
Support Vector Regression
Data Fusion
Feature Selection
Efficacy
Regression
Imaging

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Zhang, D., & Shen, D. (2011). Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7012 LNCS, pp. 60-67). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7012 LNCS). https://doi.org/10.1007/978-3-642-24446-9_8

Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease. / Zhang, Daoqiang; Shen, Dinggang.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7012 LNCS 2011. p. 60-67 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7012 LNCS).

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

Zhang, D & Shen, D 2011, Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7012 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7012 LNCS, pp. 60-67, 1st International Workshop on Multimodal Brain Image Analysis, MBIA 2011, in Conjunction with the 14th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2011, Toronto, ON, Canada, 11/9/18. https://doi.org/10.1007/978-3-642-24446-9_8
Zhang D, Shen D. Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7012 LNCS. 2011. p. 60-67. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-24446-9_8
Zhang, Daoqiang ; Shen, Dinggang. / Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7012 LNCS 2011. pp. 60-67 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{41a637ed7be342e890f19579857fecc8,
title = "Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease",
abstract = "One recent interest in computer-aided diagnosis of neurological diseases is to predict the clinical scores from brain images. Most existing methods usually estimate multiple clinical variables separately, without considering the useful correlation information among them. On the other hand, nearly all methods use only one modality of data (mostly structural MRI) for regression, and thus ignore the complementary information among different modalities. To address these issues, in this paper, we present a general methodology, namely Multi-Modal Multi-Task (M3T) learning, to jointly predict multiple variables from multi-modal data. Our method contains three major subsequent steps: (1) a multi-task feature selection which selects the common subset of relevant features for the related multiple clinical variables from each modality; (2) a kernel-based multimodal data fusion which fuses the above-selected features from all modalities; (3) a support vector regression which predicts multiple clinical variables based on the previously learnt mixed kernel. Experimental results on ADNI dataset with both imaging modalities (MRI and PET) and biological modality (CSF) validate the efficacy of the proposed M3T learning method.",
author = "Daoqiang Zhang and Dinggang Shen",
year = "2011",
month = "10",
day = "11",
doi = "10.1007/978-3-642-24446-9_8",
language = "English",
isbn = "9783642244452",
volume = "7012 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "60--67",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Multi-modal multi-task learning for joint prediction of clinical scores in Alzheimer's disease

AU - Zhang, Daoqiang

AU - Shen, Dinggang

PY - 2011/10/11

Y1 - 2011/10/11

N2 - One recent interest in computer-aided diagnosis of neurological diseases is to predict the clinical scores from brain images. Most existing methods usually estimate multiple clinical variables separately, without considering the useful correlation information among them. On the other hand, nearly all methods use only one modality of data (mostly structural MRI) for regression, and thus ignore the complementary information among different modalities. To address these issues, in this paper, we present a general methodology, namely Multi-Modal Multi-Task (M3T) learning, to jointly predict multiple variables from multi-modal data. Our method contains three major subsequent steps: (1) a multi-task feature selection which selects the common subset of relevant features for the related multiple clinical variables from each modality; (2) a kernel-based multimodal data fusion which fuses the above-selected features from all modalities; (3) a support vector regression which predicts multiple clinical variables based on the previously learnt mixed kernel. Experimental results on ADNI dataset with both imaging modalities (MRI and PET) and biological modality (CSF) validate the efficacy of the proposed M3T learning method.

AB - One recent interest in computer-aided diagnosis of neurological diseases is to predict the clinical scores from brain images. Most existing methods usually estimate multiple clinical variables separately, without considering the useful correlation information among them. On the other hand, nearly all methods use only one modality of data (mostly structural MRI) for regression, and thus ignore the complementary information among different modalities. To address these issues, in this paper, we present a general methodology, namely Multi-Modal Multi-Task (M3T) learning, to jointly predict multiple variables from multi-modal data. Our method contains three major subsequent steps: (1) a multi-task feature selection which selects the common subset of relevant features for the related multiple clinical variables from each modality; (2) a kernel-based multimodal data fusion which fuses the above-selected features from all modalities; (3) a support vector regression which predicts multiple clinical variables based on the previously learnt mixed kernel. Experimental results on ADNI dataset with both imaging modalities (MRI and PET) and biological modality (CSF) validate the efficacy of the proposed M3T learning method.

UR - http://www.scopus.com/inward/record.url?scp=80053553966&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80053553966&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-24446-9_8

DO - 10.1007/978-3-642-24446-9_8

M3 - Conference contribution

AN - SCOPUS:80053553966

SN - 9783642244452

VL - 7012 LNCS

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

SP - 60

EP - 67

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

ER -