Joint Classification and Regression via Deep Multi-Task Multi-Channel Learning for Alzheimer's Disease Diagnosis

Mingxia Liu, Jun Zhang, Ehsan Adeli, Dinggang Shen

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

In the field of computer-aided Alzheimer&#x0027;s disease (AD) diagnosis, jointly identifying brain diseases and predicting clinical scores using magnetic resonance (MR) imaging have attracted increasing attention since these two tasks are highly correlated. Most of existing joint learning approaches require hand-crafted feature representations for MR images. Since handcrafted features of MRI and classification/regression models may not coordinate well with each other, conventional methods may lead to sub-optimal learning performance. Also, demographic information (e.g., age, gender, and education) of subjects may also be related to brain status, and thus can help improve the diagnostic performance. However, conventional joint learning methods seldom incorporate such demographic information into the learning models. To this end, we propose a deep multi-task multichannel learning (<formula><tex>$DM^{2}L$</tex></formula>) framework for simultaneous brain disease classification and clinical score regression, using MR imaging data and demographic information of subjects. Specifically, we first identify the discriminative anatomical landmarks from MR images in a data-driven manner, and then extract multiple image patches around these detected landmarks. We then propose a deep multi-task multi-channel convolutional neural network for joint classification and regression. Our <formula><tex>$DM^{2}L$</tex></formula> framework can not only automatically learn discriminative features for MR images, but also explicitly incorporate the demographic information of subjects into the learning process. We evaluate the proposed method on four large multi-center cohorts with 1; 984 subjects, and the experimental results demonstrate that <formula><tex>$DM^{2}L$</tex></formula> is superior to several state-of-the-art joint learning methods in both the tasks of disease classification and clinical score regression.

Original languageEnglish
JournalIEEE Transactions on Biomedical Engineering
DOIs
Publication statusAccepted/In press - 2018 Sep 11

Fingerprint

Magnetic resonance
Alzheimer Disease
Joints
Learning
Brain
Demography
Magnetic Resonance Spectroscopy
Brain Diseases
Imaging techniques
Magnetic Resonance Imaging
Magnetic resonance imaging
Education
Neural networks
Hand

Keywords

  • Anatomical landmark
  • Brain disease diagnosis
  • Classification
  • Convolutional neural network
  • Convolutional neural networks
  • Dementia
  • Education
  • Feature extraction
  • Magnetic resonance imaging
  • Regression
  • Task analysis

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

@article{1d92e1a0cca141ccac5dc26b79ecfe53,
title = "Joint Classification and Regression via Deep Multi-Task Multi-Channel Learning for Alzheimer's Disease Diagnosis",
abstract = "In the field of computer-aided Alzheimer's disease (AD) diagnosis, jointly identifying brain diseases and predicting clinical scores using magnetic resonance (MR) imaging have attracted increasing attention since these two tasks are highly correlated. Most of existing joint learning approaches require hand-crafted feature representations for MR images. Since handcrafted features of MRI and classification/regression models may not coordinate well with each other, conventional methods may lead to sub-optimal learning performance. Also, demographic information (e.g., age, gender, and education) of subjects may also be related to brain status, and thus can help improve the diagnostic performance. However, conventional joint learning methods seldom incorporate such demographic information into the learning models. To this end, we propose a deep multi-task multichannel learning ($DM^{2}L$) framework for simultaneous brain disease classification and clinical score regression, using MR imaging data and demographic information of subjects. Specifically, we first identify the discriminative anatomical landmarks from MR images in a data-driven manner, and then extract multiple image patches around these detected landmarks. We then propose a deep multi-task multi-channel convolutional neural network for joint classification and regression. Our $DM^{2}L$ framework can not only automatically learn discriminative features for MR images, but also explicitly incorporate the demographic information of subjects into the learning process. We evaluate the proposed method on four large multi-center cohorts with 1; 984 subjects, and the experimental results demonstrate that $DM^{2}L$ is superior to several state-of-the-art joint learning methods in both the tasks of disease classification and clinical score regression.",
keywords = "Anatomical landmark, Brain disease diagnosis, Classification, Convolutional neural network, Convolutional neural networks, Dementia, Education, Feature extraction, Magnetic resonance imaging, Regression, Task analysis",
author = "Mingxia Liu and Jun Zhang and Ehsan Adeli and Dinggang Shen",
year = "2018",
month = "9",
day = "11",
doi = "10.1109/TBME.2018.2869989",
language = "English",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",

}

TY - JOUR

T1 - Joint Classification and Regression via Deep Multi-Task Multi-Channel Learning for Alzheimer's Disease Diagnosis

AU - Liu, Mingxia

AU - Zhang, Jun

AU - Adeli, Ehsan

AU - Shen, Dinggang

PY - 2018/9/11

Y1 - 2018/9/11

N2 - In the field of computer-aided Alzheimer's disease (AD) diagnosis, jointly identifying brain diseases and predicting clinical scores using magnetic resonance (MR) imaging have attracted increasing attention since these two tasks are highly correlated. Most of existing joint learning approaches require hand-crafted feature representations for MR images. Since handcrafted features of MRI and classification/regression models may not coordinate well with each other, conventional methods may lead to sub-optimal learning performance. Also, demographic information (e.g., age, gender, and education) of subjects may also be related to brain status, and thus can help improve the diagnostic performance. However, conventional joint learning methods seldom incorporate such demographic information into the learning models. To this end, we propose a deep multi-task multichannel learning ($DM^{2}L$) framework for simultaneous brain disease classification and clinical score regression, using MR imaging data and demographic information of subjects. Specifically, we first identify the discriminative anatomical landmarks from MR images in a data-driven manner, and then extract multiple image patches around these detected landmarks. We then propose a deep multi-task multi-channel convolutional neural network for joint classification and regression. Our $DM^{2}L$ framework can not only automatically learn discriminative features for MR images, but also explicitly incorporate the demographic information of subjects into the learning process. We evaluate the proposed method on four large multi-center cohorts with 1; 984 subjects, and the experimental results demonstrate that $DM^{2}L$ is superior to several state-of-the-art joint learning methods in both the tasks of disease classification and clinical score regression.

AB - In the field of computer-aided Alzheimer's disease (AD) diagnosis, jointly identifying brain diseases and predicting clinical scores using magnetic resonance (MR) imaging have attracted increasing attention since these two tasks are highly correlated. Most of existing joint learning approaches require hand-crafted feature representations for MR images. Since handcrafted features of MRI and classification/regression models may not coordinate well with each other, conventional methods may lead to sub-optimal learning performance. Also, demographic information (e.g., age, gender, and education) of subjects may also be related to brain status, and thus can help improve the diagnostic performance. However, conventional joint learning methods seldom incorporate such demographic information into the learning models. To this end, we propose a deep multi-task multichannel learning ($DM^{2}L$) framework for simultaneous brain disease classification and clinical score regression, using MR imaging data and demographic information of subjects. Specifically, we first identify the discriminative anatomical landmarks from MR images in a data-driven manner, and then extract multiple image patches around these detected landmarks. We then propose a deep multi-task multi-channel convolutional neural network for joint classification and regression. Our $DM^{2}L$ framework can not only automatically learn discriminative features for MR images, but also explicitly incorporate the demographic information of subjects into the learning process. We evaluate the proposed method on four large multi-center cohorts with 1; 984 subjects, and the experimental results demonstrate that $DM^{2}L$ is superior to several state-of-the-art joint learning methods in both the tasks of disease classification and clinical score regression.

KW - Anatomical landmark

KW - Brain disease diagnosis

KW - Classification

KW - Convolutional neural network

KW - Convolutional neural networks

KW - Dementia

KW - Education

KW - Feature extraction

KW - Magnetic resonance imaging

KW - Regression

KW - Task analysis

UR - http://www.scopus.com/inward/record.url?scp=85053292802&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053292802&partnerID=8YFLogxK

U2 - 10.1109/TBME.2018.2869989

DO - 10.1109/TBME.2018.2869989

M3 - Article

C2 - 30222548

AN - SCOPUS:85053292802

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

ER -