Deep learning analysis of sinus-rhythm echocardiography for the identification of patients with atrial fibrillation.
- 1Leviev Cardiovascular Institute, Sheba Medical Center, Ramat Gan, Israel.
- 2Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv, Israel.
- 3Aisap.ai, Ramat Gan, Israel.
Abstract
Paroxysmal atrial fibrillation (AF) is underdiagnosed due to its intermittent and asymptomatic nature. We evaluated whether a deep neural network (DNN) applied to transthoracic echocardiograms (TTEs) acquired in sinus rhythm can discriminate between patients with and without documented AF and identify individuals at increased risk for subsequent AF documentation and adverse outcomes. We retrospectively analysed 85 686 TTE studies (2009-21). A DNN was applied to apical four-chamber and parasternal long-axis videos without Doppler. AF ground truth was based on International Classification of Diseases (ICD) codes, electrocardiogram, Holter, or prior echocardiographic documentation of AF. The dataset was split into training (60%), validation (20%), and an independent test set (20%). In the test cohort, we assessed discrimination performance. Among patients without prior AF, we evaluated associations between model classification and subsequent AF documentation and a composite outcome of all-cause mortality, heart failure hospitalization, or stroke. The test set included 17 121 patients (median age 66 years [interquartile range, IQR 54-75]; 57.6% male), of whom 3449 (20.1%) had prior AF. The model achieved an area under the receiver operating characteristic curve of 0.85, with sensitivity 0.80 and specificity 0.72. At the selected threshold, 6581 patients (38.4%) were classified as model-positive, including 3832 (58.2%) without prior AF. Among patients without prior AF, model-positive classification was associated with subsequent AF documentation [adjusted hazard ratio (HR) 1.64; 95% confidence interval (CI) 1.33-2.04] and with the composite outcome (adjusted HR 1.24; 95% CI 1.13-1.35). Deep learning applied to sinus-rhythm TTE demonstrates discrimination for AF and identifies an imaging-derived signal associated with subsequent AF documentation and adverse outcomes.
Related
- Early rhythm control in patients with atrial fibrillation and cancer: a subgroup analysis of the EAST-AFNET 4 trial
- Left Atrial Appendage Thrombus and Long-Term Clinical Outcomes in Anticoagulated Patients with Atrial Fibrillation and Atrial Flutter Referred for a Rhythm-Control Strategy: A Prospective Observational Cohort Study
- Echocardiography-based multimodal models for thromboembolic risk prediction in atrial fibrillation: a narrative review
- Cardiometabolic therapy for rhythm control: a case report of refractory atrial fibrillation resolving after initiation of semaglutide for obesity
- 2026 Asia Pacific Heart Rhythm Society (APHRS) Guidelines (Part III) on Rate and Rhythm Control in Atrial Fibrillation.
BibTeX
@article{cohen2026deep,
title = {{Deep learning analysis of sinus-rhythm echocardiography for the identification of patients with atrial fibrillation.}},
author = {Ido Cohen and Michael Fiman and Ranel Loutati and Ehud Schwammenthal and Kobi Faierstein and Lior Fisher and Elad Maor},
journal = {European heart journal. Digital health},
year = {2026},
month = oct,
doi = {10.1093/ehjdh/ztag168},
url = {https://europepmc.org/article/MED/42953053},
note = {CC BY},
}
RIS
TY - JOUR
TI - Deep learning analysis of sinus-rhythm echocardiography for the identification of patients with atrial fibrillation.
AU - Ido Cohen
AU - Michael Fiman
AU - Ranel Loutati
AU - Ehud Schwammenthal
AU - Kobi Faierstein
AU - Lior Fisher
AU - Elad Maor
PY - 2026
DA - 2026/10/21
JO - European heart journal. Digital health
DO - 10.1093/ehjdh/ztag168
UR - https://europepmc.org/article/MED/42953053
AB - Paroxysmal atrial fibrillation (AF) is underdiagnosed due to its intermittent and asymptomatic nature. We evaluated whether a deep neural network (DNN) applied to transthoracic echocardiograms (TTEs) acquired in sinus rhythm can discriminate between patients with and without documented AF and identify individuals at increased risk for subsequent AF documentation and adverse outcomes. We retrospectively analysed 85 686 TTE studies (2009-21). A DNN was applied to apical four-chamber and parasternal long-axis videos without Doppler. AF ground truth was based on International Classification of Diseases (ICD) codes, electrocardiogram, Holter, or prior echocardiographic documentation of AF. The dataset was split into training (60%), validation (20%), and an independent test set (20%). In the test cohort, we assessed discrimination performance. Among patients without prior AF, we evaluated associations between model classification and subsequent AF documentation and a composite outcome of all-cause mortality, heart failure hospitalization, or stroke. The test set included 17 121 patients (median age 66 years [interquartile range, IQR 54-75]; 57.6% male), of whom 3449 (20.1%) had prior AF. The model achieved an area under the receiver operating characteristic curve of 0.85, with sensitivity 0.80 and specificity 0.72. At the selected threshold, 6581 patients (38.4%) were classified as model-positive, including 3832 (58.2%) without prior AF. Among patients without prior AF, model-positive classification was associated with subsequent AF documentation [adjusted hazard ratio (HR) 1.64; 95% confidence interval (CI) 1.33-2.04] and with the composite outcome (adjusted HR 1.24; 95% CI 1.13-1.35). Deep learning applied to sinus-rhythm TTE demonstrates discrimination for AF and identifies an imaging-derived signal associated with subsequent AF documentation and adverse outcomes.
SN - 2634-3916
ER -