Cargando…

Bayesian multi-task learning for decoding multi-subject neuroimaging data

Decoding models based on pattern recognition (PR) are becoming increasingly important tools for neuroimaging data analysis. In contrast to alternative (mass-univariate) encoding approaches that use hierarchical models to capture inter-subject variability, inter-subject differences are not typically...

Descripción completa

Detalles Bibliográficos
Autores principales: Marquand, Andre F., Brammer, Michael, Williams, Steven C.R., Doyle, Orla M.
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Academic Press 2014
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4010954/
https://www.ncbi.nlm.nih.gov/pubmed/24531053
http://dx.doi.org/10.1016/j.neuroimage.2014.02.008
_version_ 1782479931227242496
author Marquand, Andre F.
Brammer, Michael
Williams, Steven C.R.
Doyle, Orla M.
author_facet Marquand, Andre F.
Brammer, Michael
Williams, Steven C.R.
Doyle, Orla M.
author_sort Marquand, Andre F.
collection PubMed
description Decoding models based on pattern recognition (PR) are becoming increasingly important tools for neuroimaging data analysis. In contrast to alternative (mass-univariate) encoding approaches that use hierarchical models to capture inter-subject variability, inter-subject differences are not typically handled efficiently in PR. In this work, we propose to overcome this problem by recasting the decoding problem in a multi-task learning (MTL) framework. In MTL, a single PR model is used to learn different but related “tasks” simultaneously. The primary advantage of MTL is that it makes more efficient use of the data available and leads to more accurate models by making use of the relationships between tasks. In this work, we construct MTL models where each subject is modelled by a separate task. We use a flexible covariance structure to model the relationships between tasks and induce coupling between them using Gaussian process priors. We present an MTL method for classification problems and demonstrate a novel mapping method suitable for PR models. We apply these MTL approaches to classifying many different contrasts in a publicly available fMRI dataset and show that the proposed MTL methods produce higher decoding accuracy and more consistent discriminative activity patterns than currently used techniques. Our results demonstrate that MTL provides a promising method for multi-subject decoding studies by focusing on the commonalities between a group of subjects rather than the idiosyncratic properties of different subjects.
format Online
Article
Text
id pubmed-4010954
institution National Center for Biotechnology Information
language English
publishDate 2014
publisher Academic Press
record_format MEDLINE/PubMed
spelling pubmed-40109542014-05-15 Bayesian multi-task learning for decoding multi-subject neuroimaging data Marquand, Andre F. Brammer, Michael Williams, Steven C.R. Doyle, Orla M. Neuroimage Article Decoding models based on pattern recognition (PR) are becoming increasingly important tools for neuroimaging data analysis. In contrast to alternative (mass-univariate) encoding approaches that use hierarchical models to capture inter-subject variability, inter-subject differences are not typically handled efficiently in PR. In this work, we propose to overcome this problem by recasting the decoding problem in a multi-task learning (MTL) framework. In MTL, a single PR model is used to learn different but related “tasks” simultaneously. The primary advantage of MTL is that it makes more efficient use of the data available and leads to more accurate models by making use of the relationships between tasks. In this work, we construct MTL models where each subject is modelled by a separate task. We use a flexible covariance structure to model the relationships between tasks and induce coupling between them using Gaussian process priors. We present an MTL method for classification problems and demonstrate a novel mapping method suitable for PR models. We apply these MTL approaches to classifying many different contrasts in a publicly available fMRI dataset and show that the proposed MTL methods produce higher decoding accuracy and more consistent discriminative activity patterns than currently used techniques. Our results demonstrate that MTL provides a promising method for multi-subject decoding studies by focusing on the commonalities between a group of subjects rather than the idiosyncratic properties of different subjects. Academic Press 2014-05-15 /pmc/articles/PMC4010954/ /pubmed/24531053 http://dx.doi.org/10.1016/j.neuroimage.2014.02.008 Text en © 2014 The Authors http://creativecommons.org/licenses/by/3.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).
spellingShingle Article
Marquand, Andre F.
Brammer, Michael
Williams, Steven C.R.
Doyle, Orla M.
Bayesian multi-task learning for decoding multi-subject neuroimaging data
title Bayesian multi-task learning for decoding multi-subject neuroimaging data
title_full Bayesian multi-task learning for decoding multi-subject neuroimaging data
title_fullStr Bayesian multi-task learning for decoding multi-subject neuroimaging data
title_full_unstemmed Bayesian multi-task learning for decoding multi-subject neuroimaging data
title_short Bayesian multi-task learning for decoding multi-subject neuroimaging data
title_sort bayesian multi-task learning for decoding multi-subject neuroimaging data
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4010954/
https://www.ncbi.nlm.nih.gov/pubmed/24531053
http://dx.doi.org/10.1016/j.neuroimage.2014.02.008
work_keys_str_mv AT marquandandref bayesianmultitasklearningfordecodingmultisubjectneuroimagingdata
AT brammermichael bayesianmultitasklearningfordecodingmultisubjectneuroimagingdata
AT williamsstevencr bayesianmultitasklearningfordecodingmultisubjectneuroimagingdata
AT doyleorlam bayesianmultitasklearningfordecodingmultisubjectneuroimagingdata