Cortical surface biomarkers for predicting cognitive outcomes using group l2,1 norm

Alzheimer's Disease Neuroimaging Initiative

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

Regression models have been widely studied to investigate the prediction power of neuroimaging measures as biomarkers for inferring cognitive outcomes in the Alzheimer's disease study. Most of these models ignore the interrelated structures either within neuroimaging measures or between cognitive outcomes, and thus may have limited power to yield optimal solutions. To address this issue, we propose to use a new sparse multitask learning model called Group-Sparse Multi-task Regression and Feature Selection (G-SMuRFS) and demonstrate its effectiveness by examining the predictive power of detailed cortical thickness measures toward 3 types of cognitive scores in a large cohort. G-SMuRFS proposes a group-level l2,1-norm strategy to group relevant features together in an anatomically meaningful manner and use this prior knowledge to guide the learning process. This approach also takes into account the correlation among cognitive outcomes for building a more appropriate predictive model. Compared with traditional methods, G-SMuRFS not only demonstrates a superior performance but also identifies a small set of surface markers that are biologically meaningful.

Original languageEnglish
Pages (from-to)S185-S193
JournalNeurobiology of Aging
Volume36
Issue numberS1
DOIs
StatePublished - Jan 1 2015

Fingerprint

Neuroimaging
Biomarkers
Learning
Alzheimer Disease
Power (Psychology)

Keywords

  • Alzheimer's disease neuroimaging initiative (ADNI)
  • Cognitive function prediction
  • Cortical thickness
  • Magnetic resonance imaging (MRI)
  • Sparse learning

ASJC Scopus subject areas

  • Clinical Neurology
  • Neuroscience(all)
  • Aging
  • Developmental Biology
  • Geriatrics and Gerontology

Cite this

Cortical surface biomarkers for predicting cognitive outcomes using group l2,1 norm. / Alzheimer's Disease Neuroimaging Initiative.

In: Neurobiology of Aging, Vol. 36, No. S1, 01.01.2015, p. S185-S193.

Research output: Contribution to journalArticle

Alzheimer's Disease Neuroimaging Initiative. / Cortical surface biomarkers for predicting cognitive outcomes using group l2,1 norm. In: Neurobiology of Aging. 2015 ; Vol. 36, No. S1. pp. S185-S193.
@article{de4b1dbdb24c4fea9fc2ae1af6dc0db2,
title = "Cortical surface biomarkers for predicting cognitive outcomes using group l2,1 norm",
abstract = "Regression models have been widely studied to investigate the prediction power of neuroimaging measures as biomarkers for inferring cognitive outcomes in the Alzheimer's disease study. Most of these models ignore the interrelated structures either within neuroimaging measures or between cognitive outcomes, and thus may have limited power to yield optimal solutions. To address this issue, we propose to use a new sparse multitask learning model called Group-Sparse Multi-task Regression and Feature Selection (G-SMuRFS) and demonstrate its effectiveness by examining the predictive power of detailed cortical thickness measures toward 3 types of cognitive scores in a large cohort. G-SMuRFS proposes a group-level l2,1-norm strategy to group relevant features together in an anatomically meaningful manner and use this prior knowledge to guide the learning process. This approach also takes into account the correlation among cognitive outcomes for building a more appropriate predictive model. Compared with traditional methods, G-SMuRFS not only demonstrates a superior performance but also identifies a small set of surface markers that are biologically meaningful.",
keywords = "Alzheimer's disease neuroimaging initiative (ADNI), Cognitive function prediction, Cortical thickness, Magnetic resonance imaging (MRI), Sparse learning",
author = "{Alzheimer's Disease Neuroimaging Initiative} and Jingwen Yan and Taiyong Li and Hua Wang and Heng Huang and Jing Wan and Kwangsik Nho and Sungeun Kim and Risacher, {Shannon L.} and Andrew Saykin and Li Shen",
year = "2015",
month = "1",
day = "1",
doi = "10.1016/j.neurobiolaging.2014.07.045",
language = "English",
volume = "36",
pages = "S185--S193",
journal = "Neurobiology of Aging",
issn = "0197-4580",
publisher = "Elsevier Inc.",
number = "S1",

}

TY - JOUR

T1 - Cortical surface biomarkers for predicting cognitive outcomes using group l2,1 norm

AU - Alzheimer's Disease Neuroimaging Initiative

AU - Yan, Jingwen

AU - Li, Taiyong

AU - Wang, Hua

AU - Huang, Heng

AU - Wan, Jing

AU - Nho, Kwangsik

AU - Kim, Sungeun

AU - Risacher, Shannon L.

AU - Saykin, Andrew

AU - Shen, Li

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Regression models have been widely studied to investigate the prediction power of neuroimaging measures as biomarkers for inferring cognitive outcomes in the Alzheimer's disease study. Most of these models ignore the interrelated structures either within neuroimaging measures or between cognitive outcomes, and thus may have limited power to yield optimal solutions. To address this issue, we propose to use a new sparse multitask learning model called Group-Sparse Multi-task Regression and Feature Selection (G-SMuRFS) and demonstrate its effectiveness by examining the predictive power of detailed cortical thickness measures toward 3 types of cognitive scores in a large cohort. G-SMuRFS proposes a group-level l2,1-norm strategy to group relevant features together in an anatomically meaningful manner and use this prior knowledge to guide the learning process. This approach also takes into account the correlation among cognitive outcomes for building a more appropriate predictive model. Compared with traditional methods, G-SMuRFS not only demonstrates a superior performance but also identifies a small set of surface markers that are biologically meaningful.

AB - Regression models have been widely studied to investigate the prediction power of neuroimaging measures as biomarkers for inferring cognitive outcomes in the Alzheimer's disease study. Most of these models ignore the interrelated structures either within neuroimaging measures or between cognitive outcomes, and thus may have limited power to yield optimal solutions. To address this issue, we propose to use a new sparse multitask learning model called Group-Sparse Multi-task Regression and Feature Selection (G-SMuRFS) and demonstrate its effectiveness by examining the predictive power of detailed cortical thickness measures toward 3 types of cognitive scores in a large cohort. G-SMuRFS proposes a group-level l2,1-norm strategy to group relevant features together in an anatomically meaningful manner and use this prior knowledge to guide the learning process. This approach also takes into account the correlation among cognitive outcomes for building a more appropriate predictive model. Compared with traditional methods, G-SMuRFS not only demonstrates a superior performance but also identifies a small set of surface markers that are biologically meaningful.

KW - Alzheimer's disease neuroimaging initiative (ADNI)

KW - Cognitive function prediction

KW - Cortical thickness

KW - Magnetic resonance imaging (MRI)

KW - Sparse learning

UR - http://www.scopus.com/inward/record.url?scp=84916613101&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84916613101&partnerID=8YFLogxK

U2 - 10.1016/j.neurobiolaging.2014.07.045

DO - 10.1016/j.neurobiolaging.2014.07.045

M3 - Article

VL - 36

SP - S185-S193

JO - Neurobiology of Aging

JF - Neurobiology of Aging

SN - 0197-4580

IS - S1

ER -