Deep learning methods for screening patients' S-ICD implantation eligibility.

Anthony Dunn; Mohamed Elrefai; Paul Roberts; Stefano Coniglio; Benedict Wiles; Alain Zemkoho

doi:10.1016/j.artmed.2021.102139

Deep learning methods for screening patients' S-ICD implantation eligibility.

Anthony Dunn
, Mohamed Elrefai
, Paul Roberts
, Stefano Coniglio
, Benedict Wiles
, Alain Zemkoho^* (Corresponding Author)

^*Corresponding author for this work

University of Southampton
St George's University Hospitals NHS Foundation Trust
University Hospital Southampton NHS Foundation Trust

Research output: Contribution to journal › Article › peer-review

17 Citations (Scopus)

3 Downloads (Pure)

Abstract

Subcutaneous Implantable Cardioverter-Defibrillators (S-ICDs) are used for prevention of sudden cardiac death triggered by ventricular arrhythmias. T Wave Over Sensing (TWOS) is an inherent risk with S-ICDs which can lead to inappropriate shocks. A major predictor of TWOS is a high T:R ratio (the ratio between the amplitudes of the T and R waves). Currently, patients' Electrocardiograms (ECGs) are screened over 10 s to measure the T:R ratio to determine the patients' eligibility for S-ICD implantation. Due to temporal variations in the T:R ratio, 10 s is not a long enough window to reliably determine the normal values of a patient's T:R ratio. In this paper, we develop a convolutional neural network (CNN) based model utilising phase space reconstruction matrices to predict T:R ratios from 10-second ECG segments without explicitly locating the R or T waves, thus avoiding the issue of TWOS. This tool can be used to automatically screen patients over a much longer period and provide an in-depth description of the behavior of the T:R ratio over that period. The tool can also enable much more reliable and descriptive screenings to better assess patients' eligibility for S-ICD implantation.

Original language	English
Article number	102139
Number of pages	12
Journal	Artificial Intelligence in Medicine
Volume	119
Early online date	9 Aug 2021
DOIs	https://doi.org/10.1016/j.artmed.2021.102139
Publication status	Published - Sept 2021
Externally published	Yes

Bibliographical note

Acknowledgments
The work of Anthony J. Dunn is jointly funded by Decision Analysis Services Ltd. and EPSRC through the Studentship with Reference EP/R513325/1. The work of Alain B. Zemkoho is supported by the EPSRC grant EP/V049038/1 and the Alan Turing Institute under the EPSRC grant EP/N510129/1.

The feedback provided by Sion Cave (DAS Ltd) on the initial draft of the paper is gratefully acknowledged.

Funding

The work of Anthony J. Dunn is jointly funded by Decision Analysis Services Ltd. and EPSRC through the Studentship with Reference EP/R513325/1. The work of Alain B. Zemkoho is supported by the EPSRC grant EP/V049038/1 and the Alan Turing Institute under the EPSRC grant EP/N510129/1.

Funders	Funder number
Decision Analysis Services Ltd.
Engineering and Physical Sciences Research Council	EP/R513325/1, EP/V049038/1 , EP/N510129/1.

Keywords

Subcutaneous implantable cardioverter-defibrillators
Sudden cardiac death
Ventricular arrhythmia
Electrocardiogram
Deep learning
Convolutional neural networks
Phase space reconstruction
Patient screening

Access to Document

10.1016/j.artmed.2021.102139Licence: CC BY

Dunn_etal_AIIM_Deep_Learning_Methods_VOR
0933-3657/© 2021 University of Southampton. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license https://creativecommons.org/licenses/by-nc-nd/4.0/
Final published version, 2.76 MBLicence: CC BY-NC-ND

Cite this

@article{73f7b8196eac48f9881b655c650ed528,

title = "Deep learning methods for screening patients' S-ICD implantation eligibility.",

abstract = "Subcutaneous Implantable Cardioverter-Defibrillators (S-ICDs) are used for prevention of sudden cardiac death triggered by ventricular arrhythmias. T Wave Over Sensing (TWOS) is an inherent risk with S-ICDs which can lead to inappropriate shocks. A major predictor of TWOS is a high T:R ratio (the ratio between the amplitudes of the T and R waves). Currently, patients' Electrocardiograms (ECGs) are screened over 10 s to measure the T:R ratio to determine the patients' eligibility for S-ICD implantation. Due to temporal variations in the T:R ratio, 10 s is not a long enough window to reliably determine the normal values of a patient's T:R ratio. In this paper, we develop a convolutional neural network (CNN) based model utilising phase space reconstruction matrices to predict T:R ratios from 10-second ECG segments without explicitly locating the R or T waves, thus avoiding the issue of TWOS. This tool can be used to automatically screen patients over a much longer period and provide an in-depth description of the behavior of the T:R ratio over that period. The tool can also enable much more reliable and descriptive screenings to better assess patients' eligibility for S-ICD implantation.",

keywords = "Subcutaneous implantable cardioverter-defibrillators, Sudden cardiac death, Ventricular arrhythmia, Electrocardiogram, Deep learning, Convolutional neural networks, Phase space reconstruction, Patient screening",

author = "Anthony Dunn and Mohamed Elrefai and Paul Roberts and Stefano Coniglio and Benedict Wiles and Alain Zemkoho",

note = "Acknowledgments The work of Anthony J. Dunn is jointly funded by Decision Analysis Services Ltd. and EPSRC through the Studentship with Reference EP/R513325/1. The work of Alain B. Zemkoho is supported by the EPSRC grant EP/V049038/1 and the Alan Turing Institute under the EPSRC grant EP/N510129/1. The feedback provided by Sion Cave (DAS Ltd) on the initial draft of the paper is gratefully acknowledged.",

year = "2021",

month = sep,

doi = "10.1016/j.artmed.2021.102139",

language = "English",

volume = "119",

journal = "Artificial Intelligence in Medicine",

issn = "0933-3657",

publisher = "Elsevier BV",

}

TY - JOUR

T1 - Deep learning methods for screening patients' S-ICD implantation eligibility.

AU - Dunn, Anthony

AU - Elrefai, Mohamed

AU - Roberts, Paul

AU - Coniglio, Stefano

AU - Wiles, Benedict

AU - Zemkoho, Alain

N1 - Acknowledgments The work of Anthony J. Dunn is jointly funded by Decision Analysis Services Ltd. and EPSRC through the Studentship with Reference EP/R513325/1. The work of Alain B. Zemkoho is supported by the EPSRC grant EP/V049038/1 and the Alan Turing Institute under the EPSRC grant EP/N510129/1. The feedback provided by Sion Cave (DAS Ltd) on the initial draft of the paper is gratefully acknowledged.

PY - 2021/9

Y1 - 2021/9

N2 - Subcutaneous Implantable Cardioverter-Defibrillators (S-ICDs) are used for prevention of sudden cardiac death triggered by ventricular arrhythmias. T Wave Over Sensing (TWOS) is an inherent risk with S-ICDs which can lead to inappropriate shocks. A major predictor of TWOS is a high T:R ratio (the ratio between the amplitudes of the T and R waves). Currently, patients' Electrocardiograms (ECGs) are screened over 10 s to measure the T:R ratio to determine the patients' eligibility for S-ICD implantation. Due to temporal variations in the T:R ratio, 10 s is not a long enough window to reliably determine the normal values of a patient's T:R ratio. In this paper, we develop a convolutional neural network (CNN) based model utilising phase space reconstruction matrices to predict T:R ratios from 10-second ECG segments without explicitly locating the R or T waves, thus avoiding the issue of TWOS. This tool can be used to automatically screen patients over a much longer period and provide an in-depth description of the behavior of the T:R ratio over that period. The tool can also enable much more reliable and descriptive screenings to better assess patients' eligibility for S-ICD implantation.

AB - Subcutaneous Implantable Cardioverter-Defibrillators (S-ICDs) are used for prevention of sudden cardiac death triggered by ventricular arrhythmias. T Wave Over Sensing (TWOS) is an inherent risk with S-ICDs which can lead to inappropriate shocks. A major predictor of TWOS is a high T:R ratio (the ratio between the amplitudes of the T and R waves). Currently, patients' Electrocardiograms (ECGs) are screened over 10 s to measure the T:R ratio to determine the patients' eligibility for S-ICD implantation. Due to temporal variations in the T:R ratio, 10 s is not a long enough window to reliably determine the normal values of a patient's T:R ratio. In this paper, we develop a convolutional neural network (CNN) based model utilising phase space reconstruction matrices to predict T:R ratios from 10-second ECG segments without explicitly locating the R or T waves, thus avoiding the issue of TWOS. This tool can be used to automatically screen patients over a much longer period and provide an in-depth description of the behavior of the T:R ratio over that period. The tool can also enable much more reliable and descriptive screenings to better assess patients' eligibility for S-ICD implantation.

KW - Subcutaneous implantable cardioverter-defibrillators

KW - Sudden cardiac death

KW - Ventricular arrhythmia

KW - Electrocardiogram

KW - Deep learning

KW - Convolutional neural networks

KW - Phase space reconstruction

KW - Patient screening

UR - https://doi.org/10.1016/j.artmed.2021.102139

U2 - 10.1016/j.artmed.2021.102139

DO - 10.1016/j.artmed.2021.102139

M3 - Article

C2 - 34531008

SN - 0933-3657

VL - 119

JO - Artificial Intelligence in Medicine

JF - Artificial Intelligence in Medicine

M1 - 102139

ER -

Deep learning methods for screening patients' S-ICD implantation eligibility.

Abstract

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this