Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks
Generative Adversarial Networks (GANs) are promising analytical tools in machine learning applications. Characterizing atypical neurodevelopmental processes might be useful in establishing diagnostic and prognostic biomarkers of psychiatric disorders. In this article, we investigate the potential of...
Main Authors: | , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2023-01-01
|
Series: | Frontiers in Neuroscience |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2022.1025492/full |
_version_ | 1797957395217907712 |
---|---|
author | Pedro Machado Nery Dos Santos Sérgio Leonardo Mendes Claudinei Biazoli Ary Gadelha Ary Gadelha Giovanni Abrahão Salum Giovanni Abrahão Salum Euripedes Constantino Miguel Euripedes Constantino Miguel Luis Augusto Rohde Luis Augusto Rohde Luis Augusto Rohde João Ricardo Sato João Ricardo Sato João Ricardo Sato João Ricardo Sato |
author_facet | Pedro Machado Nery Dos Santos Sérgio Leonardo Mendes Claudinei Biazoli Ary Gadelha Ary Gadelha Giovanni Abrahão Salum Giovanni Abrahão Salum Euripedes Constantino Miguel Euripedes Constantino Miguel Luis Augusto Rohde Luis Augusto Rohde Luis Augusto Rohde João Ricardo Sato João Ricardo Sato João Ricardo Sato João Ricardo Sato |
author_sort | Pedro Machado Nery Dos Santos |
collection | DOAJ |
description | Generative Adversarial Networks (GANs) are promising analytical tools in machine learning applications. Characterizing atypical neurodevelopmental processes might be useful in establishing diagnostic and prognostic biomarkers of psychiatric disorders. In this article, we investigate the potential of GANs models combined with functional connectivity (FC) measures to build a predictive neurotypicality score 3-years after scanning. We used a ROI-to-ROI analysis of resting-state functional magnetic resonance imaging (fMRI) data from a community-based cohort of children and adolescents (377 neurotypical and 126 atypical participants). Models were trained on data from neurotypical participants, capturing their sample variability of FC. The discriminator subnetwork of each GAN model discriminated between the learned neurotypical functional connectivity pattern and atypical or unrelated patterns. Discriminator models were combined in ensembles, improving discrimination performance. Explanations for the model’s predictions are provided using the LIME (Local Interpretable Model-Agnostic) algorithm and local hubs are identified in light of these explanations. Our findings suggest this approach is a promising strategy to build potential biomarkers based on functional connectivity. |
first_indexed | 2024-04-11T00:03:38Z |
format | Article |
id | doaj.art-b371b84e28ae4c9882ac85918af41b68 |
institution | Directory Open Access Journal |
issn | 1662-453X |
language | English |
last_indexed | 2024-04-11T00:03:38Z |
publishDate | 2023-01-01 |
publisher | Frontiers Media S.A. |
record_format | Article |
series | Frontiers in Neuroscience |
spelling | doaj.art-b371b84e28ae4c9882ac85918af41b682023-01-09T14:21:59ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2023-01-011610.3389/fnins.2022.10254921025492Assessing atypical brain functional connectivity development: An approach based on generative adversarial networksPedro Machado Nery Dos Santos0Sérgio Leonardo Mendes1Claudinei Biazoli2Ary Gadelha3Ary Gadelha4Giovanni Abrahão Salum5Giovanni Abrahão Salum6Euripedes Constantino Miguel7Euripedes Constantino Miguel8Luis Augusto Rohde9Luis Augusto Rohde10Luis Augusto Rohde11João Ricardo Sato12João Ricardo Sato13João Ricardo Sato14João Ricardo Sato15Center of Mathematics, Computing, and Cognition, Universidade Federal do ABC, São Bernardo do Campo, Santo André, BrazilCenter of Mathematics, Computing, and Cognition, Universidade Federal do ABC, São Bernardo do Campo, Santo André, BrazilCenter of Mathematics, Computing, and Cognition, Universidade Federal do ABC, São Bernardo do Campo, Santo André, BrazilLaboratory of Integrative Neuroscience, Universidade Federal de São Paulo, São Paulo, BrazilNational Institute of Developmental Psychiatry for Children and Adolescents (CNPq), São Paulo, BrazilNational Institute of Developmental Psychiatry for Children and Adolescents (CNPq), São Paulo, BrazilDepartment of Psychiatry, Hospital de Clínicas de Porto Alegre, Federal University of Rio Grande do Sul, Porto Alegre, BrazilNational Institute of Developmental Psychiatry for Children and Adolescents (CNPq), São Paulo, BrazilDepartment of Psychiatry, School of Medicine, University of São Paulo, São Paulo, BrazilNational Institute of Developmental Psychiatry for Children and Adolescents (CNPq), São Paulo, BrazilDepartment of Psychiatry, Hospital de Clínicas de Porto Alegre, Federal University of Rio Grande do Sul, Porto Alegre, BrazilUniEduK, Jaguariúna, BrazilCenter of Mathematics, Computing, and Cognition, Universidade Federal do ABC, São Bernardo do Campo, Santo André, BrazilLaboratory of Integrative Neuroscience, Universidade Federal de São Paulo, São Paulo, BrazilNational Institute of Developmental Psychiatry for Children and Adolescents (CNPq), São Paulo, BrazilBig Data, Hospital Israelita Albert Einstein, São Paulo, BrazilGenerative Adversarial Networks (GANs) are promising analytical tools in machine learning applications. Characterizing atypical neurodevelopmental processes might be useful in establishing diagnostic and prognostic biomarkers of psychiatric disorders. In this article, we investigate the potential of GANs models combined with functional connectivity (FC) measures to build a predictive neurotypicality score 3-years after scanning. We used a ROI-to-ROI analysis of resting-state functional magnetic resonance imaging (fMRI) data from a community-based cohort of children and adolescents (377 neurotypical and 126 atypical participants). Models were trained on data from neurotypical participants, capturing their sample variability of FC. The discriminator subnetwork of each GAN model discriminated between the learned neurotypical functional connectivity pattern and atypical or unrelated patterns. Discriminator models were combined in ensembles, improving discrimination performance. Explanations for the model’s predictions are provided using the LIME (Local Interpretable Model-Agnostic) algorithm and local hubs are identified in light of these explanations. Our findings suggest this approach is a promising strategy to build potential biomarkers based on functional connectivity.https://www.frontiersin.org/articles/10.3389/fnins.2022.1025492/fullmachine learning (ML)biomarkerneural networkschildrenfunctional connectivityGANs |
spellingShingle | Pedro Machado Nery Dos Santos Sérgio Leonardo Mendes Claudinei Biazoli Ary Gadelha Ary Gadelha Giovanni Abrahão Salum Giovanni Abrahão Salum Euripedes Constantino Miguel Euripedes Constantino Miguel Luis Augusto Rohde Luis Augusto Rohde Luis Augusto Rohde João Ricardo Sato João Ricardo Sato João Ricardo Sato João Ricardo Sato Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks Frontiers in Neuroscience machine learning (ML) biomarker neural networks children functional connectivity GANs |
title | Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks |
title_full | Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks |
title_fullStr | Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks |
title_full_unstemmed | Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks |
title_short | Assessing atypical brain functional connectivity development: An approach based on generative adversarial networks |
title_sort | assessing atypical brain functional connectivity development an approach based on generative adversarial networks |
topic | machine learning (ML) biomarker neural networks children functional connectivity GANs |
url | https://www.frontiersin.org/articles/10.3389/fnins.2022.1025492/full |
work_keys_str_mv | AT pedromachadonerydossantos assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT sergioleonardomendes assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT claudineibiazoli assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT arygadelha assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT arygadelha assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT giovanniabrahaosalum assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT giovanniabrahaosalum assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT euripedesconstantinomiguel assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT euripedesconstantinomiguel assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT luisaugustorohde assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT luisaugustorohde assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT luisaugustorohde assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT joaoricardosato assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT joaoricardosato assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT joaoricardosato assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks AT joaoricardosato assessingatypicalbrainfunctionalconnectivitydevelopmentanapproachbasedongenerativeadversarialnetworks |