Evaluating the feasibility of 12-lead electrocardiogram reconstruction from limited leads using deep learning
AI-based 12-lead ECG reconstruction results in a regression-to-the-mean effect rather than personalized output, rendering it unsuitable for clinical use.
Authors
- Alexandru Dorobanțiu (ORCID: https://orcid.org/0000-0003-4982-6930)
- Arun Raghav Mahankali Sridhar
- Vajira Thambawita (ORCID: https://orcid.org/0000-0001-6026-0929)
- Jonas L. Isaksen (ORCID: https://orcid.org/0000-0003-3227-1131)
- Jørgen K. Kanters (ORCID: https://orcid.org/0000-0002-3267-4910)
- Claus Graff
- Michael A. Riegler (ORCID: https://orcid.org/0000-0002-3153-2064)
- Oriana Presacan (ORCID: https://orcid.org/0009-0005-6909-4476)
- Tobias Willi
Institutions
- University of Copenhagen (DK)
- Lucian Blaga University of Sibiu (RO)
- MultiCare Health System (US)
- University of California, San Francisco (US)
- OsloMet – Oslo Metropolitan University (NO)
- Simula Research Laboratory (NO)
- Simula Metropolitan Center for Digital Engineering (NO)
- KTH Royal Institute of Technology (SE)
- Aalborg University (DK)
Publication Details
- Journal
- Communications Medicine
- Published
- 2025-04-25
- DOI
- https://doi.org/10.1038/s43856-025-00814-w
- Citations
- 8
- Primary Topic
- ECG Monitoring and Analysis
- Type
- article
- Field-Weighted Citation Impact
- 11.11
Funders
- Norges Forskningsråd