Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
May 21, 2026npj Digital MedicineOpen Access

AI model detects hospital-diagnosed HF from 12-lead ECGs with an AUC of 0.86.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Heart failure diagnosis is resource-intensive and leads to severe underdiagnosis, while diagnostic codes have limited validity for training supervised deep learning models directly.

Does a deep learning model using electrocardiograms accurately detect heart failure?

Population

25,300 development, 43,727 prospective test, and 161,352 MIMIC-IV external validation cohort patients

Comparison

Deep learning model for HF detection from electrocardiograms vs reference standards

Design

Model development and validation study

Key result

An artificial intelligence model using 12-lead electrocardiograms detected hospital-diagnosed heart failure with an AUC of 0.86, which improved to 0.96 when validated with strict NT-proBNP thresholds.

Authors

ESElias StenhedeJRJesper RavnHSHenrik Schirmer

Discussion

Loading...

Member takes

Overview

Supports EHR-based heart failure prediction in large cohorts; leaves open whether it improves outcomes before clinical adoption.

Key Points

  • This research aims to improve heart failure diagnosis by implementing a deep learning model that analyzes electrocardiograms.
  • Developed a neural network trained on a cohort of 25,300 patients and validated on 43,727 patients.
  • Utilized NT-proBNP levels to refine labelling during training and improved model accuracy.
  • Conducted external validation using the MIMIC-IV cohort of 161,352 patients.
  • Achieved an AUC of 0.96 (95% CI 0.951–0.963) with age-adjusted NT-proBNP thresholds.
  • Model predictions validated against echocardiographic assessments demonstrated accuracy in capturing diastolic and systolic function.
  • Released the model as open source for wider accessibility.

Study Design

Type

Observational (n=69,027)

Multicenter

Yes

Structured PICO

Does a deep learning model using electrocardiograms accurately detect heart failure?

P
Population
230,379 patients across a development cohort (n=25,300), an independent test cohort (n=43,727), and an external validation cohort (MIMIC-IV, n=161,352)
I
Intervention
Deep learning neural network for heart failure detection from electrocardiograms, trained using diagnosis codes validated by NT-proBNP levels
O
Outcome
Area under the receiver operating characteristic curve (AUC) for detecting hospital-diagnosed heart failuresurrogate

Main Result

Effect estimate: AUC 0.86 (95% CI 0.847-0.864)

A deep learning model trained on ECGs with NT-proBNP-validated labels can accurately detect heart failure, offering a scalable tool for addressing underdiagnosis.

Limitations

  • Lack of a definitive gold standard for heart failure diagnosis in routinely collected data.
  • ICD-10 codes have limited validity and act as a noisy proxy for the underlying heart failure syndrome.
  • Imputation of missing values in the H2FPEF grading algorithm may have led to underestimation of scores.
  • Some analyses were specified post hoc.
  • The study population had limited sociodemographic diversity, with less than 20% being of non-Caucasian descent.

Cite This Study

Stenhede et al. (2026) conducted an observational in Heart failure (n=69,027). Artificial intelligence-enabled electrocardiogram model was evaluated on Detection of hospital-diagnosed heart failure (AUC 0.86, 95% CI 0.847-0.864). An artificial intelligence model using 12-lead electrocardiograms detected hospital-diagnosed heart failure with an AUC of 0.86, which improved to 0.96 when validated with strict NT-proBNP thresholds.

synapsesocial.com/papers/6a0ea188be05d6e3efb60580https://doi.org/10.1038/s41746-026-02774-4
View Full Paper
Ask AI
Bookmark
Share