Graph-guided joint prediction of class label and clinical scores for the Alzheimer’s disease

Guan Yu, Yufeng Liu, Dinggang Shen

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Accurate diagnosis of Alzheimer’s disease and its prodromal stage, i.e., mild cognitive impairment, is very important for early treatment. Over the last decade, various machine learning methods have been proposed to predict disease status and clinical scores from brain images. It is worth noting that many features extracted from brain images are correlated significantly. In this case, feature selection combined with the additional correlation information among features can effectively improve classification/regression performance. Typically, the correlation information among features can be modeled by the connectivity of an undirected graph, where each node represents one feature and each edge indicates that the two involved features are correlated significantly. In this paper, we propose a new graph-guided multi-task learning method incorporating this undirected graph information to predict multiple response variables (i.e., class label and clinical scores) jointly. Specifically, based on the sparse undirected feature graph, we utilize a new latent group Lasso penalty to encourage the correlated features to be selected together. Furthermore, this new penalty also encourages the intrinsic correlated tasks to share a common feature subset. To validate our method, we have performed many numerical studies using simulated datasets and the Alzheimer’s Disease Neuroimaging Initiative dataset. Compared with the other methods, our proposed method has very promising performance.

Original languageEnglish
JournalBrain Structure and Function
DOIs
Publication statusAccepted/In press - 2015 Oct 17

Fingerprint

Alzheimer Disease
Joints
Prodromal Symptoms
Brain
Neuroimaging
Learning
Datasets

Keywords

  • Alzheimer’s disease
  • Group Lasso
  • Magnetic resonance imaging (MRI)
  • Multi-task learning
  • Partial correlation
  • Positron emission tomography (PET)
  • Undirected graph

ASJC Scopus subject areas

  • Neuroscience(all)
  • Anatomy
  • Histology

Cite this

Graph-guided joint prediction of class label and clinical scores for the Alzheimer’s disease. / Yu, Guan; Liu, Yufeng; Shen, Dinggang.

In: Brain Structure and Function, 17.10.2015.

Research output: Contribution to journalArticle

@article{147ca01c828f4abd91de7f4b8c395e61,
title = "Graph-guided joint prediction of class label and clinical scores for the Alzheimer’s disease",
abstract = "Accurate diagnosis of Alzheimer’s disease and its prodromal stage, i.e., mild cognitive impairment, is very important for early treatment. Over the last decade, various machine learning methods have been proposed to predict disease status and clinical scores from brain images. It is worth noting that many features extracted from brain images are correlated significantly. In this case, feature selection combined with the additional correlation information among features can effectively improve classification/regression performance. Typically, the correlation information among features can be modeled by the connectivity of an undirected graph, where each node represents one feature and each edge indicates that the two involved features are correlated significantly. In this paper, we propose a new graph-guided multi-task learning method incorporating this undirected graph information to predict multiple response variables (i.e., class label and clinical scores) jointly. Specifically, based on the sparse undirected feature graph, we utilize a new latent group Lasso penalty to encourage the correlated features to be selected together. Furthermore, this new penalty also encourages the intrinsic correlated tasks to share a common feature subset. To validate our method, we have performed many numerical studies using simulated datasets and the Alzheimer’s Disease Neuroimaging Initiative dataset. Compared with the other methods, our proposed method has very promising performance.",
keywords = "Alzheimer’s disease, Group Lasso, Magnetic resonance imaging (MRI), Multi-task learning, Partial correlation, Positron emission tomography (PET), Undirected graph",
author = "Guan Yu and Yufeng Liu and Dinggang Shen",
year = "2015",
month = "10",
day = "17",
doi = "10.1007/s00429-015-1132-6",
language = "English",
journal = "Anatomische Hefte",
issn = "0177-5154",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Graph-guided joint prediction of class label and clinical scores for the Alzheimer’s disease

AU - Yu, Guan

AU - Liu, Yufeng

AU - Shen, Dinggang

PY - 2015/10/17

Y1 - 2015/10/17

N2 - Accurate diagnosis of Alzheimer’s disease and its prodromal stage, i.e., mild cognitive impairment, is very important for early treatment. Over the last decade, various machine learning methods have been proposed to predict disease status and clinical scores from brain images. It is worth noting that many features extracted from brain images are correlated significantly. In this case, feature selection combined with the additional correlation information among features can effectively improve classification/regression performance. Typically, the correlation information among features can be modeled by the connectivity of an undirected graph, where each node represents one feature and each edge indicates that the two involved features are correlated significantly. In this paper, we propose a new graph-guided multi-task learning method incorporating this undirected graph information to predict multiple response variables (i.e., class label and clinical scores) jointly. Specifically, based on the sparse undirected feature graph, we utilize a new latent group Lasso penalty to encourage the correlated features to be selected together. Furthermore, this new penalty also encourages the intrinsic correlated tasks to share a common feature subset. To validate our method, we have performed many numerical studies using simulated datasets and the Alzheimer’s Disease Neuroimaging Initiative dataset. Compared with the other methods, our proposed method has very promising performance.

AB - Accurate diagnosis of Alzheimer’s disease and its prodromal stage, i.e., mild cognitive impairment, is very important for early treatment. Over the last decade, various machine learning methods have been proposed to predict disease status and clinical scores from brain images. It is worth noting that many features extracted from brain images are correlated significantly. In this case, feature selection combined with the additional correlation information among features can effectively improve classification/regression performance. Typically, the correlation information among features can be modeled by the connectivity of an undirected graph, where each node represents one feature and each edge indicates that the two involved features are correlated significantly. In this paper, we propose a new graph-guided multi-task learning method incorporating this undirected graph information to predict multiple response variables (i.e., class label and clinical scores) jointly. Specifically, based on the sparse undirected feature graph, we utilize a new latent group Lasso penalty to encourage the correlated features to be selected together. Furthermore, this new penalty also encourages the intrinsic correlated tasks to share a common feature subset. To validate our method, we have performed many numerical studies using simulated datasets and the Alzheimer’s Disease Neuroimaging Initiative dataset. Compared with the other methods, our proposed method has very promising performance.

KW - Alzheimer’s disease

KW - Group Lasso

KW - Magnetic resonance imaging (MRI)

KW - Multi-task learning

KW - Partial correlation

KW - Positron emission tomography (PET)

KW - Undirected graph

UR - http://www.scopus.com/inward/record.url?scp=84945232623&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84945232623&partnerID=8YFLogxK

U2 - 10.1007/s00429-015-1132-6

DO - 10.1007/s00429-015-1132-6

M3 - Article

JO - Anatomische Hefte

JF - Anatomische Hefte

SN - 0177-5154

ER -