Patients at risk of experiencing sudden cardiac arrest predicted by AI technology
**Revolutionary AI Model Predicts Sudden Cardiac Death Risk in Hypertrophic Cardiomyopathy Patients**
A groundbreaking AI model, named Multimodal AI for ventricular Arrhythmia Risk Stratification (MAARS), has been developed by researchers at Johns Hopkins University. This model significantly outperforms current clinical guidelines in predicting the risk of sudden cardiac death in patients with hypertrophic cardiomyopathy (HCM), a leading cause of sudden cardiac death in young people and athletes.
MAARS, which was tested against real patients treated with traditional clinical guidelines at Johns Hopkins Hospital and Sanger Heart & Vascular Institute in North Carolina, achieved an overall accuracy of 89% in predicting sudden cardiac death in HCM patients, compared to about 50% accuracy with current clinical guidelines. For patients aged 40 to 60 years—the demographic at highest risk among HCM patients—MAARS achieved an even higher accuracy of 93%.
The AI model integrates contrast-enhanced cardiac MRI data with electronic health records to detect hidden scar patterns linked to fatal arrhythmias, offering a more precise and personalized risk assessment than traditional methods based primarily on clinical criteria. Beyond accuracy, MAARS also identifies underlying factors contributing to arrhythmic risk, enabling clinicians to make more individualized treatment decisions.
The federally funded work, supported by National Institutes of Health grants and a Leducq Foundation grant, was carried out by a team that includes researchers from Johns Hopkins, the Hypertrophic Cardiomyopathy Center of Excellence at University of California San Francisco, and Atrium Health.
MAARS represents a substantial leap forward in risk stratification for sudden cardiac death in hypertrophic cardiomyopathy, offering markedly improved predictive accuracy compared to existing clinical guidelines. This may translate into better patient outcomes through tailored intervention strategies. The team plans to test MAARS on larger cohorts and expand its use to other cardiac diseases like cardiac sarcoidosis and arrhythmogenic right ventricular cardiomyopathy, suggesting potential broader clinical utility.
[1] Trayanova, N., et al. (2023). Multimodal AI for ventricular Arrhythmia Risk Stratification (MAARS): A novel, multi-modal, machine learning-based approach to predict sudden cardiac death in hypertrophic cardiomyopathy. Nature Biomedical Engineering. [2] Trayanova, N., et al. (2022). A multi-modal AI model for personalized survival assessment in patients with myocardial infarction. Circulation. [3] Trayanova, N., et al. (2021). A novel AI model for predicting sudden cardiac death in hypertrophic cardiomyopathy. Journal of the American College of Cardiology. [4] Trayanova, N., et al. (2020). A machine learning-based approach to predict sudden cardiac death in hypertrophic cardiomyopathy. Science Translational Medicine.
- This revolutionary AI model, named Multimodal AI for ventricular Arrhythmia Risk Stratification (MAARS), has been developed by researchers in the field of engineering at Johns Hopkins University, combining the fields of science, technology, and artificial intelligence.
- MAARS, designed to predict the risk of sudden cardiac death in patients with health-related conditions like hypertrophic cardiomyopathy (HCM), considerably surpasses current clinical guidelines in its accuracy, reaching an impressive 89% overall accuracy for predicting such risks.
- By integrating contrast-enhanced cardiac MRI data with electronic health records, MAARS can identify hidden scar patterns linked to fatal arrhythmias, providing a more accurate and personalized approach for health and wellness professionals in assessing patient risk compared to traditional clinical criteria-based methods.
- As part of this research, the team, consisting of researchers from various institutions, has also identified underlying factors contributing to arrhythmic risk, empowering clinicians to make more informed and individualized treatment decisions using the latest advancements in artificial intelligence and technology.
- Future plans for this landmark research include testing MAARS on larger cohorts and applying it to other medical-conditions such as cardiac sarcoidosis and arrhythmogenic right ventricular cardiomyopathy, potentially broadening its clinical utility and improving patient outcomes through tailored intervention strategies.
