Deep sparse multi-task learning for feature selection in Alzheimer’s disease diagnosis

Heung-Il Suk, Seong Whan Lee, Dinggang Shen, Alzheimer’S Disease Neuroimaging Initiative The Alzheimer’S Disease Neuroimaging Initiative

Research output: Contribution to journalArticle

34 Citations (Scopus)

Abstract

Recently, neuroimaging-based Alzheimer’s disease (AD) or mild cognitive impairment (MCI) diagnosis has attracted researchers in the field, due to the increasing prevalence of the diseases. Unfortunately, the unfavorable high-dimensional nature of neuroimaging data, but a limited small number of samples available, makes it challenging to build a robust computer-aided diagnosis system. Machine learning techniques have been considered as a useful tool in this respect and, among various methods, sparse regression has shown its validity in the literature. However, to our best knowledge, the existing sparse regression methods mostly try to select features based on the optimal regression coefficients in one step. We argue that since the training feature vectors are composed of both informative and uninformative or less informative features, the resulting optimal regression coefficients are inevidently affected by the uninformative or less informative features. To this end, we first propose a novel deep architecture to recursively discard uninformative features by performing sparse multi-task learning in a hierarchical fashion. We further hypothesize that the optimal regression coefficients reflect the relative importance of features in representing the target response variables. In this regard, we use the optimal regression coefficients learned in one hierarchy as feature weighting factors in the following hierarchy, and formulate a weighted sparse multi-task learning method. Lastly, we also take into account the distributional characteristics of samples per class and use clustering-induced subclass label vectors as target response values in our sparse regression model. In our experiments on the ADNI cohort, we performed both binary and multi-class classification tasks in AD/MCI diagnosis and showed the superiority of the proposed method by comparing with the state-of-the-art methods.

Original languageEnglish
JournalBrain Structure and Function
DOIs
Publication statusAccepted/In press - 2015 May 21

Fingerprint

Alzheimer Disease
Learning
Neuroimaging
Cluster Analysis
Research Personnel
Cognitive Dysfunction

Keywords

  • Alzheimer’s disease (AD)
  • Deep architecture
  • Feature selection
  • Magnetic resonance imaging (MRI)
  • Mild cognitive impairment (MCI)
  • Multi-task learning
  • Positron emission topography (PET)
  • Sparse least squared regression

ASJC Scopus subject areas

  • Neuroscience(all)
  • Anatomy
  • Histology

Cite this

Deep sparse multi-task learning for feature selection in Alzheimer’s disease diagnosis. / Suk, Heung-Il; Lee, Seong Whan; Shen, Dinggang; The Alzheimer’S Disease Neuroimaging Initiative, Alzheimer’S Disease Neuroimaging Initiative.

In: Brain Structure and Function, 21.05.2015.

Research output: Contribution to journalArticle

Suk, Heung-Il ; Lee, Seong Whan ; Shen, Dinggang ; The Alzheimer’S Disease Neuroimaging Initiative, Alzheimer’S Disease Neuroimaging Initiative. / Deep sparse multi-task learning for feature selection in Alzheimer’s disease diagnosis. In: Brain Structure and Function. 2015.
@article{c9767828baea4b9e8d91a356e93d601c,
title = "Deep sparse multi-task learning for feature selection in Alzheimer’s disease diagnosis",
abstract = "Recently, neuroimaging-based Alzheimer’s disease (AD) or mild cognitive impairment (MCI) diagnosis has attracted researchers in the field, due to the increasing prevalence of the diseases. Unfortunately, the unfavorable high-dimensional nature of neuroimaging data, but a limited small number of samples available, makes it challenging to build a robust computer-aided diagnosis system. Machine learning techniques have been considered as a useful tool in this respect and, among various methods, sparse regression has shown its validity in the literature. However, to our best knowledge, the existing sparse regression methods mostly try to select features based on the optimal regression coefficients in one step. We argue that since the training feature vectors are composed of both informative and uninformative or less informative features, the resulting optimal regression coefficients are inevidently affected by the uninformative or less informative features. To this end, we first propose a novel deep architecture to recursively discard uninformative features by performing sparse multi-task learning in a hierarchical fashion. We further hypothesize that the optimal regression coefficients reflect the relative importance of features in representing the target response variables. In this regard, we use the optimal regression coefficients learned in one hierarchy as feature weighting factors in the following hierarchy, and formulate a weighted sparse multi-task learning method. Lastly, we also take into account the distributional characteristics of samples per class and use clustering-induced subclass label vectors as target response values in our sparse regression model. In our experiments on the ADNI cohort, we performed both binary and multi-class classification tasks in AD/MCI diagnosis and showed the superiority of the proposed method by comparing with the state-of-the-art methods.",
keywords = "Alzheimer’s disease (AD), Deep architecture, Feature selection, Magnetic resonance imaging (MRI), Mild cognitive impairment (MCI), Multi-task learning, Positron emission topography (PET), Sparse least squared regression",
author = "Heung-Il Suk and Lee, {Seong Whan} and Dinggang Shen and {The Alzheimer’S Disease Neuroimaging Initiative}, {Alzheimer’S Disease Neuroimaging Initiative}",
year = "2015",
month = "5",
day = "21",
doi = "10.1007/s00429-015-1059-y",
language = "English",
journal = "Anatomische Hefte",
issn = "0177-5154",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Deep sparse multi-task learning for feature selection in Alzheimer’s disease diagnosis

AU - Suk, Heung-Il

AU - Lee, Seong Whan

AU - Shen, Dinggang

AU - The Alzheimer’S Disease Neuroimaging Initiative, Alzheimer’S Disease Neuroimaging Initiative

PY - 2015/5/21

Y1 - 2015/5/21

N2 - Recently, neuroimaging-based Alzheimer’s disease (AD) or mild cognitive impairment (MCI) diagnosis has attracted researchers in the field, due to the increasing prevalence of the diseases. Unfortunately, the unfavorable high-dimensional nature of neuroimaging data, but a limited small number of samples available, makes it challenging to build a robust computer-aided diagnosis system. Machine learning techniques have been considered as a useful tool in this respect and, among various methods, sparse regression has shown its validity in the literature. However, to our best knowledge, the existing sparse regression methods mostly try to select features based on the optimal regression coefficients in one step. We argue that since the training feature vectors are composed of both informative and uninformative or less informative features, the resulting optimal regression coefficients are inevidently affected by the uninformative or less informative features. To this end, we first propose a novel deep architecture to recursively discard uninformative features by performing sparse multi-task learning in a hierarchical fashion. We further hypothesize that the optimal regression coefficients reflect the relative importance of features in representing the target response variables. In this regard, we use the optimal regression coefficients learned in one hierarchy as feature weighting factors in the following hierarchy, and formulate a weighted sparse multi-task learning method. Lastly, we also take into account the distributional characteristics of samples per class and use clustering-induced subclass label vectors as target response values in our sparse regression model. In our experiments on the ADNI cohort, we performed both binary and multi-class classification tasks in AD/MCI diagnosis and showed the superiority of the proposed method by comparing with the state-of-the-art methods.

AB - Recently, neuroimaging-based Alzheimer’s disease (AD) or mild cognitive impairment (MCI) diagnosis has attracted researchers in the field, due to the increasing prevalence of the diseases. Unfortunately, the unfavorable high-dimensional nature of neuroimaging data, but a limited small number of samples available, makes it challenging to build a robust computer-aided diagnosis system. Machine learning techniques have been considered as a useful tool in this respect and, among various methods, sparse regression has shown its validity in the literature. However, to our best knowledge, the existing sparse regression methods mostly try to select features based on the optimal regression coefficients in one step. We argue that since the training feature vectors are composed of both informative and uninformative or less informative features, the resulting optimal regression coefficients are inevidently affected by the uninformative or less informative features. To this end, we first propose a novel deep architecture to recursively discard uninformative features by performing sparse multi-task learning in a hierarchical fashion. We further hypothesize that the optimal regression coefficients reflect the relative importance of features in representing the target response variables. In this regard, we use the optimal regression coefficients learned in one hierarchy as feature weighting factors in the following hierarchy, and formulate a weighted sparse multi-task learning method. Lastly, we also take into account the distributional characteristics of samples per class and use clustering-induced subclass label vectors as target response values in our sparse regression model. In our experiments on the ADNI cohort, we performed both binary and multi-class classification tasks in AD/MCI diagnosis and showed the superiority of the proposed method by comparing with the state-of-the-art methods.

KW - Alzheimer’s disease (AD)

KW - Deep architecture

KW - Feature selection

KW - Magnetic resonance imaging (MRI)

KW - Mild cognitive impairment (MCI)

KW - Multi-task learning

KW - Positron emission topography (PET)

KW - Sparse least squared regression

UR - http://www.scopus.com/inward/record.url?scp=84929692240&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929692240&partnerID=8YFLogxK

U2 - 10.1007/s00429-015-1059-y

DO - 10.1007/s00429-015-1059-y

M3 - Article

JO - Anatomische Hefte

JF - Anatomische Hefte

SN - 0177-5154

ER -