Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI

Large-scale patterns of spontaneous whole-brain activity seen in resting-state functional magnetic resonance imaging (rs-fMRI) are in part believed to arise from neural populations interacting through the structural network (Honey, Kötter, Breakspear, & Sporns, 2007 ). Gene...

Full description

Bibliographic Details
Main Authors: Kashyap, Amrit, Keilholz, Shella
Format: Article
Language:English
Published: The MIT Press 2020-01-01
Series:Network Neuroscience
Online Access:https://www.mitpressjournals.org/doi/abs/10.1162/netn_a_00129
id doaj-ff4a97d62ebb4eb1b79350444d28885e
record_format Article
spelling doaj-ff4a97d62ebb4eb1b79350444d28885e2020-11-25T02:00:22ZengThe MIT PressNetwork Neuroscience2472-17512020-01-014244846610.1162/netn_a_00129Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRIKashyap, AmritKeilholz, Shella Large-scale patterns of spontaneous whole-brain activity seen in resting-state functional magnetic resonance imaging (rs-fMRI) are in part believed to arise from neural populations interacting through the structural network (Honey, Kötter, Breakspear, & Sporns, 2007 ). Generative models that simulate this network activity, called brain network models (BNM), are able to reproduce global averaged properties of empirical rs-fMRI activity such as functional connectivity (FC) but perform poorly in reproducing unique trajectories and state transitions that are observed over the span of minutes in whole-brain data (Cabral, Kringelbach, & Deco, 2017 ; Kashyap & Keilholz, 2019 ). The manuscript demonstrates that by using recurrent neural networks, it can fit the BNM in a novel way to the rs-fMRI data and predict large amounts of variance between subsequent measures of rs-fMRI data. Simulated data also contain unique repeating trajectories observed in rs-fMRI, called quasiperiodic patterns (QPP), that span 20 s and complex state transitions observed using k-means analysis on windowed FC matrices (Allen et al., 2012 ; Majeed et al., 2011 ). Our approach is able to estimate the manifold of rs-fMRI dynamics by training on generating subsequent time points, and it can simulate complex resting-state trajectories better than the traditional generative approaches. https://www.mitpressjournals.org/doi/abs/10.1162/netn_a_00129
collection DOAJ
language English
format Article
sources DOAJ
author Kashyap, Amrit
Keilholz, Shella
spellingShingle Kashyap, Amrit
Keilholz, Shella
Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI
Network Neuroscience
author_facet Kashyap, Amrit
Keilholz, Shella
author_sort Kashyap, Amrit
title Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI
title_short Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI
title_full Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI
title_fullStr Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI
title_full_unstemmed Brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fMRI
title_sort brain network constraints and recurrent neural networks reproduce unique trajectories and state transitions seen over the span of minutes in resting-state fmri
publisher The MIT Press
series Network Neuroscience
issn 2472-1751
publishDate 2020-01-01
description Large-scale patterns of spontaneous whole-brain activity seen in resting-state functional magnetic resonance imaging (rs-fMRI) are in part believed to arise from neural populations interacting through the structural network (Honey, Kötter, Breakspear, & Sporns, 2007 ). Generative models that simulate this network activity, called brain network models (BNM), are able to reproduce global averaged properties of empirical rs-fMRI activity such as functional connectivity (FC) but perform poorly in reproducing unique trajectories and state transitions that are observed over the span of minutes in whole-brain data (Cabral, Kringelbach, & Deco, 2017 ; Kashyap & Keilholz, 2019 ). The manuscript demonstrates that by using recurrent neural networks, it can fit the BNM in a novel way to the rs-fMRI data and predict large amounts of variance between subsequent measures of rs-fMRI data. Simulated data also contain unique repeating trajectories observed in rs-fMRI, called quasiperiodic patterns (QPP), that span 20 s and complex state transitions observed using k-means analysis on windowed FC matrices (Allen et al., 2012 ; Majeed et al., 2011 ). Our approach is able to estimate the manifold of rs-fMRI dynamics by training on generating subsequent time points, and it can simulate complex resting-state trajectories better than the traditional generative approaches.
url https://www.mitpressjournals.org/doi/abs/10.1162/netn_a_00129
work_keys_str_mv AT kashyapamrit brainnetworkconstraintsandrecurrentneuralnetworksreproduceuniquetrajectoriesandstatetransitionsseenoverthespanofminutesinrestingstatefmri
AT keilholzshella brainnetworkconstraintsandrecurrentneuralnetworksreproduceuniquetrajectoriesandstatetransitionsseenoverthespanofminutesinrestingstatefmri
_version_ 1724961010387779584