PMID- 33254159
OWN - NLM
STAT- MEDLINE
DCOM- 20220307
LR  - 20220307
IS  - 1741-2552 (Electronic)
IS  - 1741-2552 (Linking)
VI  - 18
IP  - 3
DP  - 2021 Mar 9
TI  - Adaptive latent state modeling of brain network dynamics with real-time learning 
      rate optimization.
LID - 10.1088/1741-2552/abcefd [doi]
AB  - Objective. Dynamic latent state models are widely used to characterize the 
      dynamics of brain network activity for various neural signal types. To date, 
      dynamic latent state models have largely been developed for stationary brain 
      network dynamics. However, brain network dynamics can be non-stationary for 
      example due to learning, plasticity or recording instability. To enable modeling 
      these non-stationarities, two problems need to be resolved. First, novel methods 
      should be developed that can adaptively update the parameters of latent state 
      models, which is difficult due to the state being latent. Second, new methods are 
      needed to optimize the adaptation learning rate, which specifies how fast new 
      neural observations update the model parameters and can significantly influence 
      adaptation accuracy.Approach. We develop a Rate Optimized-adaptive Linear 
      State-Space Modeling (RO-adaptive LSSM) algorithm that solves these two problems. 
      First, to enable adaptation, we derive a computation- and memory-efficient 
      adaptive LSSM fitting algorithm that updates the LSSM parameters recursively and 
      in real time in the presence of the latent state. Second, we develop a real-time 
      learning rate optimization algorithm. We use comprehensive simulations of a broad 
      range of non-stationary brain network dynamics to validate both algorithms, which 
      together constitute the RO-adaptive LSSM.Main results. We show that the adaptive 
      LSSM fitting algorithm can accurately track the broad simulated non-stationary 
      brain network dynamics. We also find that the learning rate significantly affects 
      the LSSM fitting accuracy. Finally, we show that the real-time learning rate 
      optimization algorithm can run in parallel with the adaptive LSSM fitting 
      algorithm. Doing so, the combined RO-adaptive LSSM algorithm rapidly converges to 
      the optimal learning rate and accurately tracks non-stationarities.Significance. 
      These algorithms can be used to study time-varying neural dynamics underlying 
      various brain functions and enhance future neurotechnologies such as 
      brain-machine interfaces and closed-loop brain stimulation systems.
CI  - (c) 2021 IOP Publishing Ltd.
FAU - Yang, Yuxiao
AU  - Yang Y
AUID- ORCID: 0000-0001-9208-7956
AD  - Ming Hsieh Department of Electrical and Computer Engineering, Viterbi School of 
      Engineering, University of Southern California, Los Angeles, CA, United States of 
      America.
AD  - These authors contributed equally to this work.
FAU - Ahmadipour, Parima
AU  - Ahmadipour P
AUID- ORCID: 0000-0003-2998-0703
AD  - Ming Hsieh Department of Electrical and Computer Engineering, Viterbi School of 
      Engineering, University of Southern California, Los Angeles, CA, United States of 
      America.
AD  - These authors contributed equally to this work.
FAU - Shanechi, Maryam M
AU  - Shanechi MM
AUID- ORCID: 0000-0002-0544-7720
AD  - Ming Hsieh Department of Electrical and Computer Engineering, Viterbi School of 
      Engineering, University of Southern California, Los Angeles, CA, United States of 
      America.
AD  - Neuroscience Graduate Program, University of Southern California, Los Angeles, 
      CA, United States of America.
LA  - eng
PT  - Journal Article
PT  - Research Support, U.S. Gov't, Non-P.H.S.
DEP - 20210309
PL  - England
TA  - J Neural Eng
JT  - Journal of neural engineering
JID - 101217933
SB  - IM
MH  - Algorithms
MH  - *Brain/physiology
MH  - *Brain-Computer Interfaces
MH  - Learning
MH  - Stereotaxic Techniques
OTO - NOTNLM
OT  - adaptation
OT  - brain network dynamics
OT  - dynamic latent state models
OT  - learning rate
OT  - state-space models
EDAT- 2020/12/01 06:00
MHDA- 2022/03/08 06:00
CRDT- 2020/11/30 20:16
PHST- 2020/05/07 00:00 [received]
PHST- 2020/11/30 00:00 [accepted]
PHST- 2020/12/01 06:00 [pubmed]
PHST- 2022/03/08 06:00 [medline]
PHST- 2020/11/30 20:16 [entrez]
AID - 10.1088/1741-2552/abcefd [doi]
PST - epublish
SO  - J Neural Eng. 2021 Mar 9;18(3). doi: 10.1088/1741-2552/abcefd.