MCI identification by joint learning on multiple MRI data

Yue Gao, Chong Yaw Wee, Minjeong Kim, Panteleimon Giannakopoulos, Marie Louise Montandon, Sven Haller, Dinggang Shen

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

13 Citations (Scopus)

Abstract

The identification of subtle brain changes that are associated with mild cognitive impairment (MCI), the at-risk stage of Alzheimer’s disease, is still a challenging task. Different from existing works, which employ multimodal data (e.g., MRI, PET or CSF) to identify MCI subjects from normal elderly controls, we use four MRI sequences, including T1-weighted MRI (T1), Diffusion Tensor Imaging (DTI), Resting-State functional MRI (RS-fMRI) and Arterial Spin Labeling (ASL) perfusion imaging. Since these MRI sequences simultaneously capture various aspects of brain structure and function during clinical routine scan, it simplifies finding the relationship between subjects by incorporating the mutual information among them. To this end, we devise a hypergraph-based semisupervised learning algorithm. In particular, we first construct a hypergraph for each of MRI sequences separately using a star expansion method with both the training and testing data. A centralized learning is then performed to model the optimal relevance between subjects by incorporating mutual information between different MRI sequences. We then combine all centralized hypergraphs by learning the optimal weight of each hypergraph based on the minimum Laplacian. We apply our proposed method on a cohort of 41 consecutive MCI subjects and 63 age-and-gender matched controls with four MRI sequences. Our method achieves at least a 7.61% improvement in classification accuracy compared to state-of-theart methods using multiple MRI data.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages78-85
Number of pages8
Volume9350
ISBN (Print)9783319245706, 9783319245706, 9783319245706
DOIs
Publication statusPublished - 2015
Externally publishedYes
Event18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015 - Munich, Germany
Duration: 2015 Oct 52015 Oct 9

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9350
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015
CountryGermany
CityMunich
Period15/10/515/10/9

Fingerprint

Magnetic resonance imaging
Hypergraph
Mutual Information
Imaging
Semi-supervised Learning
Alzheimer's Disease
Brain
Labeling
Diffusion tensor imaging
Consecutive
Learning Algorithm
Star
Simplify
Tensor
Learning
Testing
Learning algorithms
Stars
Imaging techniques
Model

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Gao, Y., Wee, C. Y., Kim, M., Giannakopoulos, P., Montandon, M. L., Haller, S., & Shen, D. (2015). MCI identification by joint learning on multiple MRI data. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9350, pp. 78-85). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9350). Springer Verlag. https://doi.org/10.1007/978-3-319-24571-3_10

MCI identification by joint learning on multiple MRI data. / Gao, Yue; Wee, Chong Yaw; Kim, Minjeong; Giannakopoulos, Panteleimon; Montandon, Marie Louise; Haller, Sven; Shen, Dinggang.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9350 Springer Verlag, 2015. p. 78-85 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9350).

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

Gao, Y, Wee, CY, Kim, M, Giannakopoulos, P, Montandon, ML, Haller, S & Shen, D 2015, MCI identification by joint learning on multiple MRI data. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 9350, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9350, Springer Verlag, pp. 78-85, 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015, Munich, Germany, 15/10/5. https://doi.org/10.1007/978-3-319-24571-3_10
Gao Y, Wee CY, Kim M, Giannakopoulos P, Montandon ML, Haller S et al. MCI identification by joint learning on multiple MRI data. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9350. Springer Verlag. 2015. p. 78-85. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-24571-3_10
Gao, Yue ; Wee, Chong Yaw ; Kim, Minjeong ; Giannakopoulos, Panteleimon ; Montandon, Marie Louise ; Haller, Sven ; Shen, Dinggang. / MCI identification by joint learning on multiple MRI data. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9350 Springer Verlag, 2015. pp. 78-85 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{70e9db8596eb4089a96833db099d16f0,
title = "MCI identification by joint learning on multiple MRI data",
abstract = "The identification of subtle brain changes that are associated with mild cognitive impairment (MCI), the at-risk stage of Alzheimer’s disease, is still a challenging task. Different from existing works, which employ multimodal data (e.g., MRI, PET or CSF) to identify MCI subjects from normal elderly controls, we use four MRI sequences, including T1-weighted MRI (T1), Diffusion Tensor Imaging (DTI), Resting-State functional MRI (RS-fMRI) and Arterial Spin Labeling (ASL) perfusion imaging. Since these MRI sequences simultaneously capture various aspects of brain structure and function during clinical routine scan, it simplifies finding the relationship between subjects by incorporating the mutual information among them. To this end, we devise a hypergraph-based semisupervised learning algorithm. In particular, we first construct a hypergraph for each of MRI sequences separately using a star expansion method with both the training and testing data. A centralized learning is then performed to model the optimal relevance between subjects by incorporating mutual information between different MRI sequences. We then combine all centralized hypergraphs by learning the optimal weight of each hypergraph based on the minimum Laplacian. We apply our proposed method on a cohort of 41 consecutive MCI subjects and 63 age-and-gender matched controls with four MRI sequences. Our method achieves at least a 7.61{\%} improvement in classification accuracy compared to state-of-theart methods using multiple MRI data.",
author = "Yue Gao and Wee, {Chong Yaw} and Minjeong Kim and Panteleimon Giannakopoulos and Montandon, {Marie Louise} and Sven Haller and Dinggang Shen",
year = "2015",
doi = "10.1007/978-3-319-24571-3_10",
language = "English",
isbn = "9783319245706",
volume = "9350",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "78--85",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - MCI identification by joint learning on multiple MRI data

AU - Gao, Yue

AU - Wee, Chong Yaw

AU - Kim, Minjeong

AU - Giannakopoulos, Panteleimon

AU - Montandon, Marie Louise

AU - Haller, Sven

AU - Shen, Dinggang

PY - 2015

Y1 - 2015

N2 - The identification of subtle brain changes that are associated with mild cognitive impairment (MCI), the at-risk stage of Alzheimer’s disease, is still a challenging task. Different from existing works, which employ multimodal data (e.g., MRI, PET or CSF) to identify MCI subjects from normal elderly controls, we use four MRI sequences, including T1-weighted MRI (T1), Diffusion Tensor Imaging (DTI), Resting-State functional MRI (RS-fMRI) and Arterial Spin Labeling (ASL) perfusion imaging. Since these MRI sequences simultaneously capture various aspects of brain structure and function during clinical routine scan, it simplifies finding the relationship between subjects by incorporating the mutual information among them. To this end, we devise a hypergraph-based semisupervised learning algorithm. In particular, we first construct a hypergraph for each of MRI sequences separately using a star expansion method with both the training and testing data. A centralized learning is then performed to model the optimal relevance between subjects by incorporating mutual information between different MRI sequences. We then combine all centralized hypergraphs by learning the optimal weight of each hypergraph based on the minimum Laplacian. We apply our proposed method on a cohort of 41 consecutive MCI subjects and 63 age-and-gender matched controls with four MRI sequences. Our method achieves at least a 7.61% improvement in classification accuracy compared to state-of-theart methods using multiple MRI data.

AB - The identification of subtle brain changes that are associated with mild cognitive impairment (MCI), the at-risk stage of Alzheimer’s disease, is still a challenging task. Different from existing works, which employ multimodal data (e.g., MRI, PET or CSF) to identify MCI subjects from normal elderly controls, we use four MRI sequences, including T1-weighted MRI (T1), Diffusion Tensor Imaging (DTI), Resting-State functional MRI (RS-fMRI) and Arterial Spin Labeling (ASL) perfusion imaging. Since these MRI sequences simultaneously capture various aspects of brain structure and function during clinical routine scan, it simplifies finding the relationship between subjects by incorporating the mutual information among them. To this end, we devise a hypergraph-based semisupervised learning algorithm. In particular, we first construct a hypergraph for each of MRI sequences separately using a star expansion method with both the training and testing data. A centralized learning is then performed to model the optimal relevance between subjects by incorporating mutual information between different MRI sequences. We then combine all centralized hypergraphs by learning the optimal weight of each hypergraph based on the minimum Laplacian. We apply our proposed method on a cohort of 41 consecutive MCI subjects and 63 age-and-gender matched controls with four MRI sequences. Our method achieves at least a 7.61% improvement in classification accuracy compared to state-of-theart methods using multiple MRI data.

UR - http://www.scopus.com/inward/record.url?scp=84951147140&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84951147140&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-24571-3_10

DO - 10.1007/978-3-319-24571-3_10

M3 - Conference contribution

AN - SCOPUS:84951147140

SN - 9783319245706

SN - 9783319245706

SN - 9783319245706

VL - 9350

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

SP - 78

EP - 85

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

PB - Springer Verlag

ER -