Abstract:Background: Left ventricular (LV) systolic dysfunction is associated with a >8-fold increased risk of heart failure and a 2-fold risk of premature death. The use of ECG signals in screening for LV systolic dysfunction is limited by their availability to clinicians. We developed a novel deep learning-based approach that can use ECG images for the screening of LV systolic dysfunction. Methods: Using 12-lead ECGs plotted in multiple different formats, and corresponding echocardiographic data recorded within 15 days from the Yale New Haven Hospital between 2015 and 2021, we developed a convolutional neural network algorithm to detect an LV ejection fraction <40%. The model was validated within clinical settings at Yale New Haven Hospital and externally on ECG images from Cedars Sinai Medical Center in Los Angeles, CA; Lake Regional Hospital in Osage Beach, MO; Memorial Hermann Southeast Hospital in Houston, TX; and Methodist Cardiology Clinic of San Antonio, TX. In addition, it was validated in the prospective Brazilian Longitudinal Study of Adult Health. Gradient-weighted class activation mapping was used to localize class-discriminating signals on ECG images. Results: Overall, 385 601 ECGs with paired echocardiograms were used for model development. The model demonstrated high discrimination across various ECG image formats and calibrations in internal validation (area under receiving operation characteristics [AUROCs], 0.91; area under precision-recall curve [AUPRC], 0.55); and external sets of ECG images from Cedars Sinai (AUROC, 0.90 and AUPRC, 0.53), outpatient Yale New Haven Hospital clinics (AUROC, 0.94 and AUPRC, 0.77), Lake Regional Hospital (AUROC, 0.90 and AUPRC, 0.88), Memorial Hermann Southeast Hospital (AUROC, 0.91 and AUPRC 0.88), Methodist Cardiology Clinic (AUROC, 0.90 and AUPRC, 0.74), and Brazilian Longitudinal Study of Adult Health cohort (AUROC, 0.95 and AUPRC, 0.45). An ECG suggestive of LV systolic dysfunction portended >27-fold higher odds of LV systolic dysfunction on transthoracic echocardiogram (odds ratio, 27.5 [95% CI, 22.3-33.9] in the held-out set). Class-discriminative patterns localized to the anterior and anteroseptal leads (V2 and V3), corresponding to the left ventricle regardless of the ECG layout. A positive ECG screen in individuals with an LV ejection fraction ≥40% at the time of initial assessment was associated with a 3.9-fold increased risk of developing incident LV systolic dysfunction in the future (hazard ratio, 3.9 [95% CI, 3.3-4.7]; median follow-up, 3.2 years). Conclusions: We developed and externally validated a deep learning model that identifies LV systolic dysfunction from ECG images. This approach represents an automated and accessible screening strategy for LV systolic dysfunction, particularly in low-resource settings.

Electrocardiograph analysis for risk assessment of heart failure with preserved ejection fraction: A deep learning model

Deep Learning-Derived 12-Lead Electrocardiogram-Based Genotype Prediction for Hypertrophic Cardiomyopathy: a Pilot Study.

Deep Learning-based 12-Lead Electrocardiogram for Low Left Ventricular Ejection Fraction Detection in Patients

Deep learning assessment of left ventricular hypertrophy based on electrocardiogram

Model for classification of heart failure severity in patients with hypertrophic cardiomyopathy using a deep neural network algorithm with a 12-lead electrocardiogram

Prediction of Left Ventricular Ejection Fraction Changes in Heart Failure Patients Using Machine Learning and Electronic Health Records: a Multi-Site Study.

Deep learning-based prognostic model using non-enhanced cardiac cine MRI for outcome prediction in patients with heart failure

Identification of High-Risk Imaging Features in Hypertrophic Cardiomyopathy using Electrocardiography: A Deep-Learning Approach

Large Language Model-informed ECG Dual Attention Network for Heart Failure Risk Prediction

Artificial intelligence-enabled electrocardiographic screening for left ventricular systolic dysfunction and mortality risk prediction

ECG-AI: electrocardiographic artificial intelligence model for prediction of heart failure

Deep Learning-Enabled Assessment of Left Heart Structure and Function Predicts Cardiovascular Outcomes

ECG-guided non-invasive estimation of pulmonary congestion in patients with heart failure

Detection of Left Ventricular Systolic Dysfunction From Electrocardiographic Images

Dynamic risk stratification of worsening heart failure using a deep learning-enabled implanted ambulatory single-lead electrocardiogram

Identification of Hypertrophic Cardiomyopathy on Electrocardiographic Images with Deep Learning

Development and validation of a deep learning model to screen hypokalemia from electrocardiogram in emergency patients

Electrocardiogram-Based Heart Age Estimation by a Deep Learning Model Provides More Information on the Incidence of Cardiovascular Disorders

Preventing unnecessary echocardiograms in de novo patients referred to the cardiology outpatient clinic using electrocardiogram-based deep learning

Deep Learning to Estimate Left Ventricular Ejection Fraction From Routine Coronary Angiographic Images

A smartphone based AI model to detect left ventricular systolic dysfunction on 12-lead ECG