Explaining deep neural networks for knowledge discovery in electrocardiogram analysis

Abstract Deep learning-based tools may annotate and interpret medical data more quickly, consistently, and accurately than medical doctors. However, as medical doctors are ultimately responsible for clinical decision-making, any deep learning-based prediction should be accompanied by an explanation...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Steven A. Hicks, Jonas L. Isaksen, Vajira Thambawita, Jonas Ghouse, Gustav Ahlberg, Allan Linneberg, Niels Grarup, Inga Strümke, Christina Ellervik, Morten Salling Olesen, Torben Hansen, Claus Graff, Niels-Henrik Holstein-Rathlou, Pål Halvorsen, Mary M. Maleckar, Michael A. Riegler, Jørgen K. Kanters
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/6f83e0ec46bc41759fd87f303ddffe2e
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:6f83e0ec46bc41759fd87f303ddffe2e
record_format	dspace
spelling	oai:doaj.org-article:6f83e0ec46bc41759fd87f303ddffe2e2021-12-02T16:53:20ZExplaining deep neural networks for knowledge discovery in electrocardiogram analysis10.1038/s41598-021-90285-52045-2322https://doaj.org/article/6f83e0ec46bc41759fd87f303ddffe2e2021-05-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-90285-5https://doaj.org/toc/2045-2322Abstract Deep learning-based tools may annotate and interpret medical data more quickly, consistently, and accurately than medical doctors. However, as medical doctors are ultimately responsible for clinical decision-making, any deep learning-based prediction should be accompanied by an explanation that a human can understand. We present an approach called electrocardiogram gradient class activation map (ECGradCAM), which is used to generate attention maps and explain the reasoning behind deep learning-based decision-making in ECG analysis. Attention maps may be used in the clinic to aid diagnosis, discover new medical knowledge, and identify novel features and characteristics of medical tests. In this paper, we showcase how ECGradCAM attention maps can unmask how a novel deep learning model measures both amplitudes and intervals in 12-lead electrocardiograms, and we show an example of how attention maps may be used to develop novel ECG features.Steven A. HicksJonas L. IsaksenVajira ThambawitaJonas GhouseGustav AhlbergAllan LinnebergNiels GrarupInga StrümkeChristina EllervikMorten Salling OlesenTorben HansenClaus GraffNiels-Henrik Holstein-RathlouPål HalvorsenMary M. MaleckarMichael A. RieglerJørgen K. KantersNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Steven A. Hicks Jonas L. Isaksen Vajira Thambawita Jonas Ghouse Gustav Ahlberg Allan Linneberg Niels Grarup Inga Strümke Christina Ellervik Morten Salling Olesen Torben Hansen Claus Graff Niels-Henrik Holstein-Rathlou Pål Halvorsen Mary M. Maleckar Michael A. Riegler Jørgen K. Kanters Explaining deep neural networks for knowledge discovery in electrocardiogram analysis
description	Abstract Deep learning-based tools may annotate and interpret medical data more quickly, consistently, and accurately than medical doctors. However, as medical doctors are ultimately responsible for clinical decision-making, any deep learning-based prediction should be accompanied by an explanation that a human can understand. We present an approach called electrocardiogram gradient class activation map (ECGradCAM), which is used to generate attention maps and explain the reasoning behind deep learning-based decision-making in ECG analysis. Attention maps may be used in the clinic to aid diagnosis, discover new medical knowledge, and identify novel features and characteristics of medical tests. In this paper, we showcase how ECGradCAM attention maps can unmask how a novel deep learning model measures both amplitudes and intervals in 12-lead electrocardiograms, and we show an example of how attention maps may be used to develop novel ECG features.
format	article
author	Steven A. Hicks Jonas L. Isaksen Vajira Thambawita Jonas Ghouse Gustav Ahlberg Allan Linneberg Niels Grarup Inga Strümke Christina Ellervik Morten Salling Olesen Torben Hansen Claus Graff Niels-Henrik Holstein-Rathlou Pål Halvorsen Mary M. Maleckar Michael A. Riegler Jørgen K. Kanters
author_facet	Steven A. Hicks Jonas L. Isaksen Vajira Thambawita Jonas Ghouse Gustav Ahlberg Allan Linneberg Niels Grarup Inga Strümke Christina Ellervik Morten Salling Olesen Torben Hansen Claus Graff Niels-Henrik Holstein-Rathlou Pål Halvorsen Mary M. Maleckar Michael A. Riegler Jørgen K. Kanters
author_sort	Steven A. Hicks
title	Explaining deep neural networks for knowledge discovery in electrocardiogram analysis
title_short	Explaining deep neural networks for knowledge discovery in electrocardiogram analysis
title_full	Explaining deep neural networks for knowledge discovery in electrocardiogram analysis
title_fullStr	Explaining deep neural networks for knowledge discovery in electrocardiogram analysis
title_full_unstemmed	Explaining deep neural networks for knowledge discovery in electrocardiogram analysis
title_sort	explaining deep neural networks for knowledge discovery in electrocardiogram analysis
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/6f83e0ec46bc41759fd87f303ddffe2e
work_keys_str_mv	AT stevenahicks explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT jonaslisaksen explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT vajirathambawita explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT jonasghouse explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT gustavahlberg explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT allanlinneberg explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT nielsgrarup explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT ingastrumke explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT christinaellervik explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT mortensallingolesen explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT torbenhansen explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT clausgraff explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT nielshenrikholsteinrathlou explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT palhalvorsen explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT marymmaleckar explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT michaelariegler explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis AT jørgenkkanters explainingdeepneuralnetworksforknowledgediscoveryinelectrocardiogramanalysis
_version_	1718382865745444864

Explaining deep neural networks for knowledge discovery in electrocardiogram analysis

Ejemplares similares