Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network

This study presents a general trainable network of Hopf oscillators to model high-dimensional electroencephalogram (EEG) signals across different sleep stages. The proposed architecture consists of two main components: a layer of interconnected oscillators and a complex-valued feed-forward network d...

Full description

Saved in:
Bibliographic Details
Main Authors: Sayan Ghosh, Dipayan Biswas, N. R. Rohan, Sujith Vijayan, V. Srinivasa Chakravarthy
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-05-01
Series:Frontiers in Neuroinformatics
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fninf.2025.1513374/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849727602767429632
author Sayan Ghosh
Dipayan Biswas
N. R. Rohan
Sujith Vijayan
V. Srinivasa Chakravarthy
author_facet Sayan Ghosh
Dipayan Biswas
N. R. Rohan
Sujith Vijayan
V. Srinivasa Chakravarthy
author_sort Sayan Ghosh
collection DOAJ
description This study presents a general trainable network of Hopf oscillators to model high-dimensional electroencephalogram (EEG) signals across different sleep stages. The proposed architecture consists of two main components: a layer of interconnected oscillators and a complex-valued feed-forward network designed with and without a hidden layer. Incorporating a hidden layer in the feed-forward network leads to lower reconstruction errors than the simpler version without it. Our model reconstructs EEG signals across all five sleep stages and predicts the subsequent 5 s of EEG activity. The predicted data closely aligns with the empirical EEG regarding mean absolute error, power spectral similarity, and complexity measures. We propose three models, each representing a stage of increasing complexity from initial training to architectures with and without hidden layers. In these models, the oscillators initially lack spatial localization. However, we introduce spatial constraints in the final two models by superimposing spherical shells and rectangular geometries onto the oscillator network. Overall, the proposed model represents a step toward constructing a large-scale, biologically inspired model of brain dynamics.
format Article
id doaj-art-3764260c5a0f485ea64effce5f07e395
institution DOAJ
issn 1662-5196
language English
publishDate 2025-05-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Neuroinformatics
spelling doaj-art-3764260c5a0f485ea64effce5f07e3952025-08-20T03:09:48ZengFrontiers Media S.A.Frontiers in Neuroinformatics1662-51962025-05-011910.3389/fninf.2025.15133741513374Modeling of whole brain sleep electroencephalogram using deep oscillatory neural networkSayan Ghosh0Dipayan Biswas1N. R. Rohan2Sujith Vijayan3V. Srinivasa Chakravarthy4Indian Institute of Technology Madras, Chennai, IndiaIndian Institute of Technology Madras, Chennai, IndiaIndian Institute of Technology Madras, Chennai, IndiaVirginia Tech, Blacksburg, VA, United StatesIndian Institute of Technology Madras, Chennai, IndiaThis study presents a general trainable network of Hopf oscillators to model high-dimensional electroencephalogram (EEG) signals across different sleep stages. The proposed architecture consists of two main components: a layer of interconnected oscillators and a complex-valued feed-forward network designed with and without a hidden layer. Incorporating a hidden layer in the feed-forward network leads to lower reconstruction errors than the simpler version without it. Our model reconstructs EEG signals across all five sleep stages and predicts the subsequent 5 s of EEG activity. The predicted data closely aligns with the empirical EEG regarding mean absolute error, power spectral similarity, and complexity measures. We propose three models, each representing a stage of increasing complexity from initial training to architectures with and without hidden layers. In these models, the oscillators initially lack spatial localization. However, we introduce spatial constraints in the final two models by superimposing spherical shells and rectangular geometries onto the oscillator network. Overall, the proposed model represents a step toward constructing a large-scale, biologically inspired model of brain dynamics.https://www.frontiersin.org/articles/10.3389/fninf.2025.1513374/fullEEGHopf oscillatorsleep stages modelinglarge scale brain dynamicsbiomedical signal analysisHopf oscillator model
spellingShingle Sayan Ghosh
Dipayan Biswas
N. R. Rohan
Sujith Vijayan
V. Srinivasa Chakravarthy
Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
Frontiers in Neuroinformatics
EEG
Hopf oscillator
sleep stages modeling
large scale brain dynamics
biomedical signal analysis
Hopf oscillator model
title Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
title_full Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
title_fullStr Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
title_full_unstemmed Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
title_short Modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
title_sort modeling of whole brain sleep electroencephalogram using deep oscillatory neural network
topic EEG
Hopf oscillator
sleep stages modeling
large scale brain dynamics
biomedical signal analysis
Hopf oscillator model
url https://www.frontiersin.org/articles/10.3389/fninf.2025.1513374/full
work_keys_str_mv AT sayanghosh modelingofwholebrainsleepelectroencephalogramusingdeeposcillatoryneuralnetwork
AT dipayanbiswas modelingofwholebrainsleepelectroencephalogramusingdeeposcillatoryneuralnetwork
AT nrrohan modelingofwholebrainsleepelectroencephalogramusingdeeposcillatoryneuralnetwork
AT sujithvijayan modelingofwholebrainsleepelectroencephalogramusingdeeposcillatoryneuralnetwork
AT vsrinivasachakravarthy modelingofwholebrainsleepelectroencephalogramusingdeeposcillatoryneuralnetwork