Related Experiment Videos
An ECG biomarker for sudden cardiac death discovered with deep learning
Ziad Obermeyer1,2, Alexander Schubert3, James Ross4
1School of Public Health, University of California, Berkeley, Berkeley, CA, USA. zobermeyer@berkeley.edu.
Insights
Sudden cardiac death risk can be better predicted using deep learning on electrocardiograms (ECGs). This new AI model identifies high-risk patients missed by current methods, potentially improving defibrillator use and saving lives.
Area of Science:
- Cardiology
- Artificial Intelligence
- Biomedical Engineering
Background:
- Sudden cardiac death (SCD) remains a significant cause of mortality.
- Current risk prediction using left ventricular ejection fraction (LVEF) is insufficient, missing many at-risk individuals and incorrectly flagging others.
- There is a critical need for improved predictive biomarkers for SCD.
Purpose of the Study:
- To develop and validate a deep learning model using electrocardiograms (ECGs) for more accurate prediction of sudden cardiac death risk.
- To identify novel ECG-based biomarkers for SCD prediction.
- To assess the clinical utility of the developed model in guiding defibrillator implantation decisions.
Main Methods:
- Applied deep learning to a large dataset of ECGs linked to death certificates from a Swedish region.
- Validated the model in independent datasets from a US health system and a Taiwanese hospital registry.
- Utilized a generative model to visualize and understand the ECG waveform morphology identified by the predictive model.
Main Results:
- The deep learning model identified a high-risk group (2.2% of patients) with a 7.0% annual SCD rate, significantly higher than the LVEF-identified group (4.6%).
- 86.1% of high-risk patients identified by the model were not flagged by LVEF.
- Implanted defibrillators in high-risk patients were associated with a 54.4% reduction in mortality compared to expected rates.
- External validation confirmed the model's ability to predict ventricular arrhythmias and arrhythmic cardiac arrests.
Conclusions:
- Deep learning applied to ECGs offers a powerful new tool for identifying individuals at high risk of sudden cardiac death.
- The model significantly outperforms LVEF in risk stratification and identifies a distinct high-risk population.
- The identified ECG biomarker is visually apparent, robust, and potentially reveals new insights into SCD mechanisms, suggesting improved clinical decision-making for defibrillator therapy.
Abstract:
Sudden cardiac death is, in theory, preventable with defibrillators. But every year, many patients die without defibrillators because doctors fail to predict their risk1. The only predictive biomarker in wide use, cardiac left ventricular ejection fraction (LVEF), misses most sudden cardiac deaths2, and flags many low-risk patients for futile defibrillators that never fire3,4. Here we apply deep learning to a dataset linking all electrocardiograms (ECGs) in a Swedish region to death certificates. The resulting model isolates a high-risk group (2.2% of the sample) with a 7.0% annual rate of sudden cardiac death, higher than those with reduced LVEF (1.9% of the sample; 4.6% annual rate). Notably, 86.1% of the model's high-risk patients were not flagged by LVEF. High-risk ECG patients with defibrillators implanted were 54.4% less likely to die than expected, suggesting a mortality benefit. We externally validate the model in a US health system, in which it predicts ventricular arrhythmias that cause sudden death; and a Taiwanese hospital registry, in which it specifically predicts future arrhythmic cardiac arrests. To visualize the waveform morphology 'discovered' by the predictive model, we pair it with a generative model of the ECG waveform. Together, they reveal a biomarker that is easily visible and robustly predicts sudden cardiac death, but has not to our knowledge been previously described. Tying the biomarker's shape to electrophysiological first principles, we form and preliminarily test a new hypothesis on the mechanism of sudden cardiac death.
Related Concept Videos
01:20Blood Studies for Cardiovascular System I: Cardiac Biomarkers
The essential diagnostic tools for detecting myocardial necrosis and monitoring individuals suspected of having acute coronary syndrome (ACS) include:
Troponins
Troponins, particularly cardiac troponins I and T, are the most precise and sensitive markers of myocardial injury. They are detectable within 4-6 hours of myocardial injury and remain...
01:29Electrocardiogram
Three major waveforms are present in a typical ECG recording: the P wave, the QRS complex, and the T...
01:23Cardiopulmonary Resuscitation III: AED Use
01:25Correlation between ECG and Cardiac Cycle
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...