Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Artificial intelligence for direct-to-physician reporting of ambulatory electrocardiography

Johnson, L S LU ; Zadrozniak, P ; Jasina, G ; Grotek-Cuprjak, A ; Andrade, J G ; Svennberg, E ; Diederichsen, S Z ; McIntyre, W F ; Stavrakis, S and Benezet-Mazuecos, J , et al. (2025) In Nature Medicine
Abstract

Developments in ambulatory electrocardiogram (ECG) technology have led to vast amounts of ECG data that currently need to be interpreted by human technicians. Here we tested an artificial intelligence (AI) algorithm for direct-to-physician reporting of ambulatory ECGs. Beat-by-beat annotation of 14,606 individual ambulatory ECG recordings (mean duration = 14 ± 10 days) was performed by certified ECG technicians (n = 167) and an ensemble AI model, called DeepRhythmAI. To compare the performance of the AI model and the technicians, a random sample of 5,235 rhythm events identified by the AI model or by technicians, of which 2,236 events were identified as critical arrhythmias, was selected for annotation by one of 17 cardiologist... (More)

Developments in ambulatory electrocardiogram (ECG) technology have led to vast amounts of ECG data that currently need to be interpreted by human technicians. Here we tested an artificial intelligence (AI) algorithm for direct-to-physician reporting of ambulatory ECGs. Beat-by-beat annotation of 14,606 individual ambulatory ECG recordings (mean duration = 14 ± 10 days) was performed by certified ECG technicians (n = 167) and an ensemble AI model, called DeepRhythmAI. To compare the performance of the AI model and the technicians, a random sample of 5,235 rhythm events identified by the AI model or by technicians, of which 2,236 events were identified as critical arrhythmias, was selected for annotation by one of 17 cardiologist consensus panels. The mean sensitivity of the AI model for the identification of critical arrhythmias was 98.6% (95% confidence interval (CI) = 97.7-99.4), as compared to 80.3% (95% CI = 77.3-83.3%) for the technicians. False-negative findings were observed in 3.2/1,000 patients for the AI model versus 44.3/1,000 patients for the technicians. Accordingly, the relative risk of a missed diagnosis was 14.1 (95% CI = 10.4-19.0) times higher for the technicians. However, a higher false-positive event rate was observed for the AI model (12 (interquartile range (IQR) = 6-74)/1,000 patient days) as compared to the technicians (5 (IQR = 2-153)/1,000 patient days). We conclude that the DeepRhythmAI model has excellent negative predictive value for critical arrhythmias, substantially reducing false-negative findings, but at a modest cost of increased false-positive findings. AI-only analysis to facilitate direct-to-physician reporting could potentially reduce costs and improve access to care and outcomes in patients who need ambulatory ECG monitoring.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and , et al. (More)
; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; and (Less)
organization
publishing date
type
Contribution to journal
publication status
epub
subject
in
Nature Medicine
publisher
Nature Publishing Group
external identifiers
  • scopus:85217554824
  • pmid:39930139
ISSN
1546-170X
DOI
10.1038/s41591-025-03516-x
language
English
LU publication?
yes
additional info
© 2025. The Author(s).
id
c4323388-7163-439f-b1e9-a2c7f8630796
date added to LUP
2025-02-17 19:28:10
date last changed
2025-07-08 15:58:11
@article{c4323388-7163-439f-b1e9-a2c7f8630796,
  abstract     = {{<p>Developments in ambulatory electrocardiogram (ECG) technology have led to vast amounts of ECG data that currently need to be interpreted by human technicians. Here we tested an artificial intelligence (AI) algorithm for direct-to-physician reporting of ambulatory ECGs. Beat-by-beat annotation of 14,606 individual ambulatory ECG recordings (mean duration = 14 ± 10 days) was performed by certified ECG technicians (n = 167) and an ensemble AI model, called DeepRhythmAI. To compare the performance of the AI model and the technicians, a random sample of 5,235 rhythm events identified by the AI model or by technicians, of which 2,236 events were identified as critical arrhythmias, was selected for annotation by one of 17 cardiologist consensus panels. The mean sensitivity of the AI model for the identification of critical arrhythmias was 98.6% (95% confidence interval (CI) = 97.7-99.4), as compared to 80.3% (95% CI = 77.3-83.3%) for the technicians. False-negative findings were observed in 3.2/1,000 patients for the AI model versus 44.3/1,000 patients for the technicians. Accordingly, the relative risk of a missed diagnosis was 14.1 (95% CI = 10.4-19.0) times higher for the technicians. However, a higher false-positive event rate was observed for the AI model (12 (interquartile range (IQR) = 6-74)/1,000 patient days) as compared to the technicians (5 (IQR = 2-153)/1,000 patient days). We conclude that the DeepRhythmAI model has excellent negative predictive value for critical arrhythmias, substantially reducing false-negative findings, but at a modest cost of increased false-positive findings. AI-only analysis to facilitate direct-to-physician reporting could potentially reduce costs and improve access to care and outcomes in patients who need ambulatory ECG monitoring.</p>}},
  author       = {{Johnson, L S and Zadrozniak, P and Jasina, G and Grotek-Cuprjak, A and Andrade, J G and Svennberg, E and Diederichsen, S Z and McIntyre, W F and Stavrakis, S and Benezet-Mazuecos, J and Krisai, P and Iakobishvili, Z and Laish-Farkash, A and Bhavnani, S and Ljungström, E and Bacevicius, J and van Vreeswijk, N L and Rienstra, M and Spittler, R and Marx, J A and Oraii, A and Miracle Blanco, A and Lozano, A and Mustafina, I and Zafeiropoulos, S and Bennett, R and Bisson, J and Linz, D and Kogan, Y and Glazer, E and Marincheva, G and Rahkovich, M and Shaked, E and Ruwald, M H and Haugan, K and Węcławski, J and Radoslovich, G and Jamal, S and Brandes, A and Matusik, P T and Manninger, M and Meyre, P B and Blum, S and Persson, A and Måneheim, A and Hammarlund, P and Fedorowski, A and Wodaje, T and Lewinter, C and Juknevicius, V and Jakaite, R and Shen, C and Glotzer, T and Platonov, P and Engström, G and Benz, A P and Healey, J S}},
  issn         = {{1546-170X}},
  language     = {{eng}},
  month        = {{02}},
  publisher    = {{Nature Publishing Group}},
  series       = {{Nature Medicine}},
  title        = {{Artificial intelligence for direct-to-physician reporting of ambulatory electrocardiography}},
  url          = {{http://dx.doi.org/10.1038/s41591-025-03516-x}},
  doi          = {{10.1038/s41591-025-03516-x}},
  year         = {{2025}},
}