On the use of generative deep neural networks to synthesize artificial multichannel EEG signals

Ozan Özdenizci; Deniz Erdogmus

doi:10.1109/NER49283.2021.9441381

On the use of generative deep neural networks to synthesize artificial multichannel EEG signals

Ozan Özdenizci, Deniz Erdogmus

Institute of Theoretical Computer Science (7080)

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

Recent promises of generative deep learning lately brought interest to its potential uses in neural engineering. In this paper we firstly review recently emerging studies on generating artificial electroencephalography (EEG) signals with deep neural networks. Subsequently, we present our feasibility experiments on generating condition-specific multichannel EEG signals using conditional variational autoencoders. By manipulating real resting-state EEG epochs, we present an approach to synthetically generate time-series multichannel signals that show spectro-temporal EEG patterns which are expected to be observed during distinct motor imagery conditions.

Original language	English
Title of host publication	2021 10th International IEEE/EMBS Conference on Neural Engineering, NER 2021
Pages	427-430
Number of pages	4
ISBN (Electronic)	9781728143378
DOIs	https://doi.org/10.1109/NER49283.2021.9441381
Publication status	Published - 2021
Event	10th International IEEE/EMBS Conference on Neural Engineering - Virtuell Duration: 4 May 2021 → 6 May 2021

Conference

Conference	10th International IEEE/EMBS Conference on Neural Engineering
Abbreviated title	NER '21
City	Virtuell
Period	4/05/21 → 6/05/21

ASJC Scopus subject areas

Artificial Intelligence
Mechanical Engineering

Fields of Expertise

Human- & Biotechnology
Information, Communication & Computing

Access to Document

10.1109/NER49283.2021.9441381

https://arxiv.org/pdf/2102.08061.pdf

Cite this

@inproceedings{62434f0f2695412297aa435d12bd7955,

title = "On the use of generative deep neural networks to synthesize artificial multichannel EEG signals",

abstract = "Recent promises of generative deep learning lately brought interest to its potential uses in neural engineering. In this paper we firstly review recently emerging studies on generating artificial electroencephalography (EEG) signals with deep neural networks. Subsequently, we present our feasibility experiments on generating condition-specific multichannel EEG signals using conditional variational autoencoders. By manipulating real resting-state EEG epochs, we present an approach to synthetically generate time-series multichannel signals that show spectro-temporal EEG patterns which are expected to be observed during distinct motor imagery conditions.",

author = "Ozan {\"O}zdenizci and Deniz Erdogmus",

year = "2021",

doi = "10.1109/NER49283.2021.9441381",

language = "English",

pages = "427--430",

booktitle = "2021 10th International IEEE/EMBS Conference on Neural Engineering, NER 2021",

note = "10th International IEEE/EMBS Conference on Neural Engineering, NER '21 ; Conference date: 04-05-2021 Through 06-05-2021",

}

TY - GEN

T1 - On the use of generative deep neural networks to synthesize artificial multichannel EEG signals

AU - Özdenizci, Ozan

AU - Erdogmus, Deniz

PY - 2021

Y1 - 2021

N2 - Recent promises of generative deep learning lately brought interest to its potential uses in neural engineering. In this paper we firstly review recently emerging studies on generating artificial electroencephalography (EEG) signals with deep neural networks. Subsequently, we present our feasibility experiments on generating condition-specific multichannel EEG signals using conditional variational autoencoders. By manipulating real resting-state EEG epochs, we present an approach to synthetically generate time-series multichannel signals that show spectro-temporal EEG patterns which are expected to be observed during distinct motor imagery conditions.

AB - Recent promises of generative deep learning lately brought interest to its potential uses in neural engineering. In this paper we firstly review recently emerging studies on generating artificial electroencephalography (EEG) signals with deep neural networks. Subsequently, we present our feasibility experiments on generating condition-specific multichannel EEG signals using conditional variational autoencoders. By manipulating real resting-state EEG epochs, we present an approach to synthetically generate time-series multichannel signals that show spectro-temporal EEG patterns which are expected to be observed during distinct motor imagery conditions.

UR - http://www.scopus.com/inward/record.url?scp=85107510817&partnerID=8YFLogxK

U2 - 10.1109/NER49283.2021.9441381

DO - 10.1109/NER49283.2021.9441381

M3 - Conference paper

SP - 427

EP - 430

BT - 2021 10th International IEEE/EMBS Conference on Neural Engineering, NER 2021

T2 - 10th International IEEE/EMBS Conference on Neural Engineering

Y2 - 4 May 2021 through 6 May 2021

ER -

On the use of generative deep neural networks to synthesize artificial multichannel EEG signals

Abstract

Conference

ASJC Scopus subject areas

Fields of Expertise

Access to Document

Other files and links

Fingerprint

Cite this