Sparse Bayesian learning for identifying imaging biomarkers in AD prediction

Li Shen, Yuan Qi, Sungeun Kim, Kwangsik Nho, Jing Wan, Shannon L. Risacher, Andrew J. Saykin

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

Abstract

We apply sparse Bayesian learning methods, automatic relevance determination (ARD) and predictive ARD (PARD), to Alzheimer's disease (AD) classification to make accurate prediction and identify critical imaging markers relevant to AD at the same time. ARD is one of the most successful Bayesian feature selection methods. PARD is a powerful Bayesian feature selection method, and provides sparse models that is easy to interpret. PARD selects the model with the best estimate of the predictive performance instead of choosing the one with the largest marginal model likelihood. Comparative study with support vector machine (SVM) shows that ARD/PARD in general outperform SVM in terms of prediction accuracy. Additional comparison with surface-based general linear model (GLM) analysis shows that regions with strongest signals are identified by both GLM and ARD/PARD. While GLM P-map returns significant regions all over the cortex, ARD/PARD provide a small number of relevant and meaningful imaging markers with predictive power, including both cortical and subcortical measures.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages611-618
Number of pages8
Volume6363 LNCS
EditionPART 3
DOIs
StatePublished - 2010
Event13th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2010 - Beijing, China
Duration: Sep 20 2010Sep 24 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume6363 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other13th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2010
CountryChina
CityBeijing
Period9/20/109/24/10

Fingerprint

Bayesian Learning
Alzheimer's Disease
Biomarkers
Imaging
Imaging techniques
Prediction
Linear Model
Feature Selection
Support vector machines
Feature extraction
Support Vector Machine
Marginal Model
Return Map
Cortex
Model Analysis
Comparative Study
Relevance
Likelihood
Model
Estimate

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Shen, L., Qi, Y., Kim, S., Nho, K., Wan, J., Risacher, S. L., & Saykin, A. J. (2010). Sparse Bayesian learning for identifying imaging biomarkers in AD prediction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (PART 3 ed., Vol. 6363 LNCS, pp. 611-618). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6363 LNCS, No. PART 3). https://doi.org/10.1007/978-3-642-15711-0_76

Sparse Bayesian learning for identifying imaging biomarkers in AD prediction. / Shen, Li; Qi, Yuan; Kim, Sungeun; Nho, Kwangsik; Wan, Jing; Risacher, Shannon L.; Saykin, Andrew J.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 6363 LNCS PART 3. ed. 2010. p. 611-618 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6363 LNCS, No. PART 3).

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

Shen, L, Qi, Y, Kim, S, Nho, K, Wan, J, Risacher, SL & Saykin, AJ 2010, Sparse Bayesian learning for identifying imaging biomarkers in AD prediction. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 edn, vol. 6363 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 6363 LNCS, pp. 611-618, 13th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2010, Beijing, China, 9/20/10. https://doi.org/10.1007/978-3-642-15711-0_76
Shen L, Qi Y, Kim S, Nho K, Wan J, Risacher SL et al. Sparse Bayesian learning for identifying imaging biomarkers in AD prediction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 ed. Vol. 6363 LNCS. 2010. p. 611-618. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3). https://doi.org/10.1007/978-3-642-15711-0_76
Shen, Li ; Qi, Yuan ; Kim, Sungeun ; Nho, Kwangsik ; Wan, Jing ; Risacher, Shannon L. ; Saykin, Andrew J. / Sparse Bayesian learning for identifying imaging biomarkers in AD prediction. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 6363 LNCS PART 3. ed. 2010. pp. 611-618 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
@inproceedings{ce1d171cb956418498fdb44b26707ebc,
title = "Sparse Bayesian learning for identifying imaging biomarkers in AD prediction",
abstract = "We apply sparse Bayesian learning methods, automatic relevance determination (ARD) and predictive ARD (PARD), to Alzheimer's disease (AD) classification to make accurate prediction and identify critical imaging markers relevant to AD at the same time. ARD is one of the most successful Bayesian feature selection methods. PARD is a powerful Bayesian feature selection method, and provides sparse models that is easy to interpret. PARD selects the model with the best estimate of the predictive performance instead of choosing the one with the largest marginal model likelihood. Comparative study with support vector machine (SVM) shows that ARD/PARD in general outperform SVM in terms of prediction accuracy. Additional comparison with surface-based general linear model (GLM) analysis shows that regions with strongest signals are identified by both GLM and ARD/PARD. While GLM P-map returns significant regions all over the cortex, ARD/PARD provide a small number of relevant and meaningful imaging markers with predictive power, including both cortical and subcortical measures.",
author = "Li Shen and Yuan Qi and Sungeun Kim and Kwangsik Nho and Jing Wan and Risacher, {Shannon L.} and Saykin, {Andrew J.}",
year = "2010",
doi = "10.1007/978-3-642-15711-0_76",
language = "English",
isbn = "3642157106",
volume = "6363 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 3",
pages = "611--618",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
edition = "PART 3",

}

TY - GEN

T1 - Sparse Bayesian learning for identifying imaging biomarkers in AD prediction

AU - Shen, Li

AU - Qi, Yuan

AU - Kim, Sungeun

AU - Nho, Kwangsik

AU - Wan, Jing

AU - Risacher, Shannon L.

AU - Saykin, Andrew J.

PY - 2010

Y1 - 2010

N2 - We apply sparse Bayesian learning methods, automatic relevance determination (ARD) and predictive ARD (PARD), to Alzheimer's disease (AD) classification to make accurate prediction and identify critical imaging markers relevant to AD at the same time. ARD is one of the most successful Bayesian feature selection methods. PARD is a powerful Bayesian feature selection method, and provides sparse models that is easy to interpret. PARD selects the model with the best estimate of the predictive performance instead of choosing the one with the largest marginal model likelihood. Comparative study with support vector machine (SVM) shows that ARD/PARD in general outperform SVM in terms of prediction accuracy. Additional comparison with surface-based general linear model (GLM) analysis shows that regions with strongest signals are identified by both GLM and ARD/PARD. While GLM P-map returns significant regions all over the cortex, ARD/PARD provide a small number of relevant and meaningful imaging markers with predictive power, including both cortical and subcortical measures.

AB - We apply sparse Bayesian learning methods, automatic relevance determination (ARD) and predictive ARD (PARD), to Alzheimer's disease (AD) classification to make accurate prediction and identify critical imaging markers relevant to AD at the same time. ARD is one of the most successful Bayesian feature selection methods. PARD is a powerful Bayesian feature selection method, and provides sparse models that is easy to interpret. PARD selects the model with the best estimate of the predictive performance instead of choosing the one with the largest marginal model likelihood. Comparative study with support vector machine (SVM) shows that ARD/PARD in general outperform SVM in terms of prediction accuracy. Additional comparison with surface-based general linear model (GLM) analysis shows that regions with strongest signals are identified by both GLM and ARD/PARD. While GLM P-map returns significant regions all over the cortex, ARD/PARD provide a small number of relevant and meaningful imaging markers with predictive power, including both cortical and subcortical measures.

UR - http://www.scopus.com/inward/record.url?scp=78349252168&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78349252168&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-15711-0_76

DO - 10.1007/978-3-642-15711-0_76

M3 - Conference contribution

SN - 3642157106

SN - 9783642157103

VL - 6363 LNCS

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

SP - 611

EP - 618

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

ER -