Variational AutoEncoder For Regression: Application to Brain Aging Analysis

While unsupervised variational autoencoders (VAE) have become a powerful tool in neuroimage analysis, their application to supervised learning is under-explored. We aim to close this gap by proposing a unified probabilistic model for learning the latent space of imaging data and performing supervise...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention Jg. 11765; S. 823
Hauptverfasser: Zhao, Qingyu, Adeli, Ehsan, Honnorat, Nicolas, Leng, Tuo, Pohl, Kilian M
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Germany 01.01.2019
Online-Zugang:Weitere Angaben
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract While unsupervised variational autoencoders (VAE) have become a powerful tool in neuroimage analysis, their application to supervised learning is under-explored. We aim to close this gap by proposing a unified probabilistic model for learning the latent space of imaging data and performing supervised regression. Based on recent advances in learning disentangled representations, the novel generative process explicitly models the conditional distribution of latent representations with respect to the regression target variable. Performing a variational inference procedure on this model leads to joint regularization between the VAE and a neural-network regressor. In predicting the age of 245 subjects from their structural Magnetic Resonance (MR) images, our model is more accurate than state-of-the-art methods when applied to either region-of-interest (ROI) measurements or raw 3D volume images. More importantly, unlike simple feed-forward neural-networks, disentanglement of age in latent representations allows for intuitive interpretation of the structural developmental patterns of the human brain.
AbstractList While unsupervised variational autoencoders (VAE) have become a powerful tool in neuroimage analysis, their application to supervised learning is under-explored. We aim to close this gap by proposing a unified probabilistic model for learning the latent space of imaging data and performing supervised regression. Based on recent advances in learning disentangled representations, the novel generative process explicitly models the conditional distribution of latent representations with respect to the regression target variable. Performing a variational inference procedure on this model leads to joint regularization between the VAE and a neural-network regressor. In predicting the age of 245 subjects from their structural Magnetic Resonance (MR) images, our model is more accurate than state-of-the-art methods when applied to either region-of-interest (ROI) measurements or raw 3D volume images. More importantly, unlike simple feed-forward neural-networks, disentanglement of age in latent representations allows for intuitive interpretation of the structural developmental patterns of the human brain.
While unsupervised variational autoencoders (VAE) have become a powerful tool in neuroimage analysis, their application to supervised learning is under-explored. We aim to close this gap by proposing a unified probabilistic model for learning the latent space of imaging data and performing supervised regression. Based on recent advances in learning disentangled representations, the novel generative process explicitly models the conditional distribution of latent representations with respect to the regression target variable. Performing a variational inference procedure on this model leads to joint regularization between the VAE and a neural-network regressor. In predicting the age of 245 subjects from their structural Magnetic Resonance (MR) images, our model is more accurate than state-of-the-art methods when applied to either region-of-interest (ROI) measurements or raw 3D volume images. More importantly, unlike simple feed-forward neural-networks, disentanglement of age in latent representations allows for intuitive interpretation of the structural developmental patterns of the human brain.While unsupervised variational autoencoders (VAE) have become a powerful tool in neuroimage analysis, their application to supervised learning is under-explored. We aim to close this gap by proposing a unified probabilistic model for learning the latent space of imaging data and performing supervised regression. Based on recent advances in learning disentangled representations, the novel generative process explicitly models the conditional distribution of latent representations with respect to the regression target variable. Performing a variational inference procedure on this model leads to joint regularization between the VAE and a neural-network regressor. In predicting the age of 245 subjects from their structural Magnetic Resonance (MR) images, our model is more accurate than state-of-the-art methods when applied to either region-of-interest (ROI) measurements or raw 3D volume images. More importantly, unlike simple feed-forward neural-networks, disentanglement of age in latent representations allows for intuitive interpretation of the structural developmental patterns of the human brain.
Author Honnorat, Nicolas
Leng, Tuo
Adeli, Ehsan
Pohl, Kilian M
Zhao, Qingyu
Author_xml – sequence: 1
  givenname: Qingyu
  surname: Zhao
  fullname: Zhao, Qingyu
  organization: Stanford University, Stanford, CA, USA
– sequence: 2
  givenname: Ehsan
  surname: Adeli
  fullname: Adeli, Ehsan
  organization: Stanford University, Stanford, CA, USA
– sequence: 3
  givenname: Nicolas
  surname: Honnorat
  fullname: Honnorat, Nicolas
  organization: SRI International, Menlo Park, CA, USA
– sequence: 4
  givenname: Tuo
  surname: Leng
  fullname: Leng, Tuo
  organization: Stanford University, Stanford, CA, USA
– sequence: 5
  givenname: Kilian M
  surname: Pohl
  fullname: Pohl, Kilian M
  organization: SRI International, Menlo Park, CA, USA
BackLink https://www.ncbi.nlm.nih.gov/pubmed/32705091$$D View this record in MEDLINE/PubMed
BookMark eNo1kEtPxCAYRVlofIz-A2NYukE_KBRwVyfjI05iYtRtQylMSDpQoV3Mv3ei4-ou7jl3cc_RUUzRIXRF4ZYCyDstFakIVEAqxrggqtX0BJ1WTIIATc_Q65fJwUwhRTPgZp7SKtrUu4wfU8bvbpNdKfvyHjfjOAT7S-Ip4YdsQsTNJsQNbvburoRygY69GYq7POQCfT6uPpbPZP329LJs1mTkABPx1FqtO8ld52vqhQRtKafK1sLL3iojVK9434HpakGN9sCU1wZqqblSwrEFuvnbHXP6nl2Z2m0o1g2DiS7NpWWcSaYFsGqPXh_Qudu6vh1z2Jq8a_8PYD-vklib
ContentType Journal Article
DBID NPM
7X8
DOI 10.1007/978-3-030-32245-8_91
DatabaseName PubMed
MEDLINE - Academic
DatabaseTitle PubMed
MEDLINE - Academic
DatabaseTitleList PubMed
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
ExternalDocumentID 32705091
Genre Journal Article
GrantInformation_xml – fundername: NIAAA NIH HHS
  grantid: U24 AA021697
– fundername: NIAAA NIH HHS
  grantid: R37 AA005965
– fundername: NIAAA NIH HHS
  grantid: R37 AA010723
– fundername: NIAAA NIH HHS
  grantid: K05 AA017168
– fundername: NIAAA NIH HHS
  grantid: F32 AA026762
– fundername: NIAAA NIH HHS
  grantid: U01 AA013521
– fundername: NIAAA NIH HHS
  grantid: R01 AA005965
GroupedDBID NPM
7X8
ID FETCH-LOGICAL-p400t-f1cc99b74ebf61f5709c1418c65f7dc8a58d84db0ab651a9f028f9a06794885e2
IEDL.DBID 7X8
ISICitedReferencesCount 67
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000548438900091&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Fri Jul 11 02:42:43 EDT 2025
Wed Feb 19 02:04:23 EST 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-p400t-f1cc99b74ebf61f5709c1418c65f7dc8a58d84db0ab651a9f028f9a06794885e2
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
OpenAccessLink https://www.ncbi.nlm.nih.gov/pmc/articles/7377006
PMID 32705091
PQID 2427295023
PQPubID 23479
ParticipantIDs proquest_miscellaneous_2427295023
pubmed_primary_32705091
PublicationCentury 2000
PublicationDate 2019-01-01
PublicationDateYYYYMMDD 2019-01-01
PublicationDate_xml – month: 01
  year: 2019
  text: 2019-01-01
  day: 01
PublicationDecade 2010
PublicationPlace Germany
PublicationPlace_xml – name: Germany
PublicationTitle Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
PublicationTitleAlternate Med Image Comput Comput Assist Interv
PublicationYear 2019
Score 2.284142
Snippet While unsupervised variational autoencoders (VAE) have become a powerful tool in neuroimage analysis, their application to supervised learning is...
SourceID proquest
pubmed
SourceType Aggregation Database
Index Database
StartPage 823
Title Variational AutoEncoder For Regression: Application to Brain Aging Analysis
URI https://www.ncbi.nlm.nih.gov/pubmed/32705091
https://www.proquest.com/docview/2427295023
Volume 11765
WOSCitedRecordID wos000548438900091&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1LS8QwEA7qevDiA1_riwheg03bNIkXqbKLoC4iKnsraZqIl3bd7fr7nWm7rBdB8NJLGyiTmcw3j8xHyIVHC5M-x6puwJDqnmkjQ1ZIkfjceu5ds9MPcjRS47F-6hJus66tcnEmNgd1UVnMkV-CKwEcKMDFXE8-GbJGYXW1o9BYJb0IoAy2dMmx-nFDri3-gyIzUNxYMJVp_juWbHzKcOu_f7NNNjs0SdN2-3fIiit3yf0bxL9djo-m87oalHhxfUqH1ZQ-u_e287W8oumyeE3rit4gWwRNkbWILmaV7JHX4eDl9o51nAlsAtZYM8-t1TqXsct9wr2QgbY85somwsvCKiNUoeIiD0yeCG60B3zhtcF0EpiycOE-WSur0h0SmgeRtRa-jPAdxJERYB8bmzBUEJWYuE_OF7LJQCex0GBKV81n2VI6fXLQCjibtMMzsiiUOHKGH_1h9THZAHyi24zHCel5sEh3StbtV_0xm541mw3P0dPjN3DDtYc
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Variational+AutoEncoder+For+Regression%3A+Application+to+Brain+Aging+Analysis&rft.jtitle=Medical+image+computing+and+computer-assisted+intervention+%3A+MICCAI+...+International+Conference+on+Medical+Image+Computing+and+Computer-Assisted+Intervention&rft.au=Zhao%2C+Qingyu&rft.au=Adeli%2C+Ehsan&rft.au=Honnorat%2C+Nicolas&rft.au=Leng%2C+Tuo&rft.date=2019-01-01&rft.volume=11765&rft.spage=823&rft_id=info:doi/10.1007%2F978-3-030-32245-8_91&rft.externalDBID=NO_FULL_TEXT