Research
Print page Print page
Switch language
The Capital Region of Denmark - a part of Copenhagen University Hospital
Published

Explaining deep neural networks for knowledge discovery in electrocardiogram analysis

Research output: Contribution to journalJournal articleResearchpeer-review

  1. Evaluation of 2D super-resolution ultrasound imaging of the rat renal vasculature using ex vivo micro-computed tomography

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. Automatic airway segmentation from computed tomography using robust and efficient 3-D convolutional neural networks

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Cardiac hypoxic resistance and decreasing lactate during maximum apnea in elite breath hold divers

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. Detection of biological signals from a live mammalian muscle using an early stage diamond quantum sensor

    Research output: Contribution to journalJournal articleResearchpeer-review

  1. Tre ønsker for 2022: Mere og bedre fokus på populationer, samarbejder og brug af data.

    Research output: Contribution to journalContribution to newspaper - Feature articleCommunication

  2. Rare coding variants in 35 genes associate with circulating lipid levels-A multi-ancestry analysis of 170,000 exomes

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. The power of genetic diversity in genome-wide association studies of lipids

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. The age-specific incidence of type 2 diabetes in Danish elderly adults – the role of size at birth

    Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

View graph of relations

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.

Original languageEnglish
Article number10949
JournalScientific Reports
Volume11
Issue number1
Pages (from-to)10949
ISSN2045-2322
DOIs
Publication statusPublished - 26 May 2021
Externally publishedYes

Bibliographical note

Publisher Copyright:
© 2021, The Author(s).

Copyright:
Copyright 2021 Elsevier B.V., All rights reserved.

ID: 65837298