Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI
A major tenet in theoretical neuroscience is that cognitive and behavioral processes are ultimately implemented in terms of the neural system dynamics. Accordingly, a major aim for the analysis of neurophysiological measurements should lie in the identification of the computational dynamics underlyi...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
August21, 2019
|
| In: |
PLoS Computational Biology
Year: 2019, Volume: 15, Issue: 8 |
| ISSN: | 1553-7358 |
| DOI: | 10.1371/journal.pcbi.1007263 |
| Online Access: | Verlag, Volltext: https://doi.org/10.1371/journal.pcbi.1007263 Verlag: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1007263 
|
| Author Notes: | Georgia Koppe, Hazem Toutounji, Peter Kirsch, Stefanie Lis, Daniel Durstewitz |
MARC
| LEADER | 00000caa a2200000 c 4500 | ||
|---|---|---|---|
| 001 | 168499120X | ||
| 003 | DE-627 | ||
| 005 | 20220817181015.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 191210s2019 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1371/journal.pcbi.1007263 |2 doi | |
| 035 | |a (DE-627)168499120X | ||
| 035 | |a (DE-599)KXP168499120X | ||
| 035 | |a (OCoLC)1341280830 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 084 | |a 33 |2 sdnb | ||
| 100 | 1 | |a Koppe, Georgia |d 1984- |e VerfasserIn |0 (DE-588)1095801198 |0 (DE-627)856418498 |0 (DE-576)467814724 |4 aut | |
| 245 | 1 | 0 | |a Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI |c Georgia Koppe, Hazem Toutounji, Peter Kirsch, Stefanie Lis, Daniel Durstewitz |
| 264 | 1 | |c August21, 2019 | |
| 300 | |a 35 | ||
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a A major tenet in theoretical neuroscience is that cognitive and behavioral processes are ultimately implemented in terms of the neural system dynamics. Accordingly, a major aim for the analysis of neurophysiological measurements should lie in the identification of the computational dynamics underlying task processing. Here we advance a state space model (SSM) based on generative piecewise-linear recurrent neural networks (PLRNN) to assess dynamics from neuroimaging data. In contrast to many other nonlinear time series models which have been proposed for reconstructing latent dynamics, our model is easily interpretable in neural terms, amenable to systematic dynamical systems analysis of the resulting set of equations, and can straightforwardly be transformed into an equivalent continuous-time dynamical system. The major contributions of this paper are the introduction of a new observation model suitable for functional magnetic resonance imaging (fMRI) coupled to the latent PLRNN, an efficient stepwise training procedure that forces the latent model to capture the ‘true’ underlying dynamics rather than just fitting (or predicting) the observations, and of an empirical measure based on the Kullback-Leibler divergence to evaluate from empirical time series how well this goal of approximating the underlying dynamics has been achieved. We validate and illustrate the power of our approach on simulated ‘ground-truth’ dynamical systems as well as on experimental fMRI time series, and demonstrate that the learnt dynamics harbors task-related nonlinear structure that a linear dynamical model fails to capture. Given that fMRI is one of the most common techniques for measuring brain activity non-invasively in human subjects, this approach may provide a novel step toward analyzing aberrant (nonlinear) dynamics for clinical assessment or neuroscientific research. | ||
| 650 | 4 | |a Algorithms | |
| 650 | 4 | |a Dynamical systems | |
| 650 | 4 | |a Cognition | |
| 650 | 4 | |a Covariance | |
| 650 | 4 | |a Functional magnetic resonance imaging | |
| 650 | 4 | |a Nonlinear dynamics | |
| 650 | 4 | |a Nonlinear systems | |
| 650 | 4 | |a System instability | |
| 700 | 1 | |a Toutounji, Hazem |e VerfasserIn |0 (DE-588)1185197060 |0 (DE-627)1664550526 |4 aut | |
| 700 | 1 | |a Kirsch, Peter |d 1966- |e VerfasserIn |0 (DE-588)142172197 |0 (DE-627)634419218 |0 (DE-576)328258423 |4 aut | |
| 700 | 1 | |a Lis, Stefanie |d 1963- |e VerfasserIn |0 (DE-588)123104297 |0 (DE-627)08234969X |0 (DE-576)293557675 |4 aut | |
| 700 | 1 | |a Durstewitz, Daniel |d 1967- |e VerfasserIn |0 (DE-588)12042021X |0 (DE-627)080664008 |0 (DE-576)174757050 |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |a Public Library of Science |t PLoS Computational Biology |d San Francisco, Calif. : Public Library of Science, 2005 |g 15(2019,8) Artikel-Nummer e1007263, 35 Seiten |h Online-Ressource |w (DE-627)491436017 |w (DE-600)2193340-6 |w (DE-576)273890492 |x 1553-7358 |7 nnas |
| 773 | 1 | 8 | |g volume:15 |g year:2019 |g number:8 |g extent:35 |a Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI |
| 856 | 4 | 0 | |u https://doi.org/10.1371/journal.pcbi.1007263 |x Verlag |x Resolving-System |3 Volltext |
| 856 | 4 | 0 | |u https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1007263 |x Verlag |
| 951 | |a AR | ||
| 992 | |a 20191210 | ||
| 993 | |a Article | ||
| 994 | |a 2019 | ||
| 998 | |g 12042021X |a Durstewitz, Daniel |m 12042021X:Durstewitz, Daniel |d 130000 |e 130000PD12042021X |k 0/130000/ |p 5 |y j | ||
| 998 | |g 123104297 |a Lis, Stefanie |m 123104297:Lis, Stefanie |d 60000 |e 60000PL123104297 |k 0/60000/ |p 4 | ||
| 998 | |g 142172197 |a Kirsch, Peter |m 142172197:Kirsch, Peter |d 60000 |e 60000PK142172197 |k 0/60000/ |p 3 | ||
| 998 | |g 1095801198 |a Koppe, Georgia |m 1095801198:Koppe, Georgia |d 60000 |e 60000PK1095801198 |k 0/60000/ |p 1 |x j | ||
| 999 | |a KXP-PPN168499120X |e 3563213143 | ||
| BIB | |a Y | ||
| SER | |a journal | ||
| JSO | |a {"relHost":[{"pubHistory":["1.2005 -"],"title":[{"title":"PLoS Computational Biology","subtitle":"a new community journal","title_sort":"PLoS Computational Biology"}],"part":{"extent":"35","volume":"15","issue":"8","year":"2019","text":"15(2019,8) Artikel-Nummer e1007263, 35 Seiten"},"language":["eng"],"id":{"eki":["491436017"],"zdb":["2193340-6"],"issn":["1553-7358"]},"disp":"Public Library of SciencePLoS Computational Biology","physDesc":[{"extent":"Online-Ressource"}],"corporate":[{"role":"aut","display":"Public Library of Science"}],"note":["Gesehen am 23. November 2020"],"recId":"491436017","origin":[{"dateIssuedDisp":"2005-","publisher":"Public Library of Science","dateIssuedKey":"2005","publisherPlace":"San Francisco, Calif."}],"type":{"bibl":"periodical","media":"Online-Ressource"},"name":{"displayForm":["publ. by the Public Library of Science (PLoS) in association with the International Society for Computational Biology (ISCB)"]}}],"physDesc":[{"extent":"35 S."}],"id":{"doi":["10.1371/journal.pcbi.1007263"],"eki":["168499120X"]},"type":{"bibl":"article-journal","media":"Online-Ressource"},"recId":"168499120X","origin":[{"dateIssuedDisp":"August21, 2019","dateIssuedKey":"2019"}],"person":[{"given":"Georgia","role":"aut","display":"Koppe, Georgia","family":"Koppe"},{"display":"Toutounji, Hazem","role":"aut","given":"Hazem","family":"Toutounji"},{"given":"Peter","role":"aut","display":"Kirsch, Peter","family":"Kirsch"},{"family":"Lis","display":"Lis, Stefanie","role":"aut","given":"Stefanie"},{"family":"Durstewitz","role":"aut","given":"Daniel","display":"Durstewitz, Daniel"}],"language":["eng"],"name":{"displayForm":["Georgia Koppe, Hazem Toutounji, Peter Kirsch, Stefanie Lis, Daniel Durstewitz"]},"title":[{"title_sort":"Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI","title":"Identifying nonlinear dynamical systems via generative recurrent neural networks with applications to fMRI"}]} | ||
| SRT | |a KOPPEGEORGIDENTIFYIN2120 | ||