Segmentation of Atrial Electrical Activity in Intracardiac
Electrograms (IECGs) Using Convolutional Neural Network (CNN)
Trained on Small Imbalanced Dataset

D 2021

Segmentation of Atrial Electrical Activity in Intracardiac Electrograms (IECGs) Using Convolutional Neural Network (CNN) Trained on Small Imbalanced Dataset

HEJČ, Jakub, David POSPÍŠIL, Petra NOVOTNÁ, Martin PEŠL, Oto JANOUŠEK et. al.

Basic information

Original name

Segmentation of Atrial Electrical Activity in Intracardiac Electrograms (IECGs) Using Convolutional Neural Network (CNN) Trained on Small Imbalanced Dataset

Authors

HEJČ, Jakub (203 Czech Republic), David POSPÍŠIL (203 Czech Republic, belonging to the institution), Petra NOVOTNÁ (203 Czech Republic), Martin PEŠL (203 Czech Republic, belonging to the institution), Oto JANOUŠEK (203 Czech Republic), Marina RONZHINA and Zdeněk STÁREK (203 Czech Republic, belonging to the institution)

Edition

United States, 2021 Computing in Cardiology (CinC), p. 1-4, 4 pp. 2021

Publisher

IEEE

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

30201 Cardiac and Cardiovascular systems

Country of publisher

United States of America

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

electronic version available online

References:

URL

RIV identification code

RIV/00216224:14110/21:00124596

Organization unit

Faculty of Medicine

ISBN

978-1-6654-7916-5

ISSN

DOI

http://dx.doi.org/10.23919/CinC53138.2021.9662729

UT WoS

000821955000051

Keywords in English

Atrial Electrical Activity; IECGs; CNN)

Abstract

V originále

Timing pattern of intracardiac atrial activity recorded by multipolar catheter in the coronary sinus (CS) provides insightful information about the type and approximate origin of common non-complex arrhythmias. Depending on the anatomy of the CS, the atrial activity can be substantially disturbed by ventricular far field complex preventing accurate segmentation by convential methods. In this paper, we present small clinically validated database of 326 surface 12-lead and intracardiac electrograms (ECG and IEGs) and a simple deep learning framework for semantic beat-to-beat segmentation of atrial activity in CS recordings. The model is based on a residual convolutional neural network (CNN) combined with pyramidal upsampling decoder. It is capable to recognize well between atrial and ventricular signals recorded by decapolar CS catheter in multiple arrhythmic scenarios reaching dice score of 0.875 on evaluation dataset. To address a dataset size and imbalance issues, we have adopted several preprocessing and learning techniques with adequate evaluation of its impact on the model performance.

Links

MUNI/A/1450/2021, interní kód MU

Name: Nevyřešené otázky a nové metody hodnocení elektrokardiografického signálu a struktur myokardu III. (Acronym: ECG2022)

Investor: Masaryk University

Citovat

HEJČ, Jakub, David POSPÍŠIL, Petra NOVOTNÁ, Martin PEŠL, Oto JANOUŠEK, Marina RONZHINA and Zdeněk STÁREK. Segmentation of Atrial Electrical Activity in Intracardiac Electrograms (IECGs) Using Convolutional Neural Network (CNN) Trained on Small Imbalanced Dataset. Online. In 2021 Computing in Cardiology (CinC). United States: IEEE, 2021, p. 1-4. ISBN 978-1-6654-7916-5. Available from: https://dx.doi.org/10.23919/CinC53138.2021.9662729.

@inproceedings{1853517,
   author = {Hejč, Jakub and Pospíšil, David and Novotná, Petra and Pešl, Martin and Janoušek, Oto and Ronzhina, Marina and Stárek, Zdeněk},
   address = {United States},
   booktitle = {2021 Computing in Cardiology (CinC)},
   doi = {http://dx.doi.org/10.23919/CinC53138.2021.9662729},
   keywords = {Atrial Electrical Activity; IECGs; CNN)},
   howpublished = {elektronická verze "online"},
   language = {eng},
   location = {United States},
   isbn = {978-1-6654-7916-5},
   pages = {1-4},
   publisher = {IEEE},
   title = {Segmentation of Atrial Electrical Activity in Intracardiac Electrograms (IECGs) Using Convolutional Neural Network (CNN) Trained on Small Imbalanced Dataset},
   url = {https://ieeexplore.ieee.org/document/9662729},
   year = {2021}
}

TY  - JOUR
ID  - 1853517
AU  - Hejč, Jakub - Pospíšil, David - Novotná, Petra - Pešl, Martin - Janoušek, Oto - Ronzhina, Marina - Stárek, Zdeněk
PY  - 2021
TI  - Segmentation of Atrial Electrical Activity in Intracardiac Electrograms (IECGs) Using Convolutional Neural Network (CNN) Trained on Small Imbalanced Dataset
PB  - IEEE
CY  - United States
SN  - 9781665479165
KW  - Atrial Electrical Activity
KW  - IECGs
KW  - CNN)
UR  - https://ieeexplore.ieee.org/document/9662729
N2  - Timing pattern of intracardiac atrial activity recorded by multipolar catheter in the coronary sinus (CS) provides insightful information about the type and approximate origin of common non-complex arrhythmias. Depending on the anatomy of the CS, the atrial activity can be substantially disturbed by ventricular far field complex preventing accurate segmentation by convential methods. In this paper, we present small clinically validated database of 326 surface 12-lead and intracardiac electrograms (ECG and IEGs) and a simple deep learning framework for semantic beat-to-beat segmentation of atrial activity in CS recordings. The model is based on a residual convolutional neural network (CNN) combined with pyramidal upsampling decoder. It is capable to recognize well between atrial and ventricular signals recorded by decapolar CS catheter in multiple arrhythmic scenarios reaching dice score of 0.875 on evaluation dataset. To address a dataset size and imbalance issues, we have adopted several preprocessing and learning techniques with adequate evaluation of its impact on the model performance.
ER  -

HEJČ, Jakub, David POSPÍŠIL, Petra NOVOTNÁ, Martin PEŠL, Oto JANOUŠEK, Marina RONZHINA and Zdeněk STÁREK. Segmentation of Atrial Electrical Activity in Intracardiac Electrograms (IECGs) Using Convolutional Neural Network (CNN) Trained on Small Imbalanced Dataset. Online. In \textit{2021 Computing in Cardiology (CinC)}. United States: IEEE, 2021, p.~1-4. ISBN~978-1-6654-7916-5. Available from: https://dx.doi.org/10.23919/CinC53138.2021.9662729.

Detailed Information on Publication Record