AI predicts patients likely to die of sudden cardiac arrest
Researchers in the Trayanova Lab at Johns Hopkins University, led by ADVANCE member Dr. Natalia Trayanova, as well as ADVANCE members from the Johns Hopkins School of Medicine and other collaborators, have developed a new AI tool that significantly improves prediction of sudden cardiac death in patients with hypertrophic cardiomyopathy. The tool, MAARS (Multimodal AI for Arrhythmic Risk Stratification), combines contrast-enhanced cardiac MRI with electronic health records to detect hidden scar patterns linked to fatal arrhythmias.
Current clinical guidelines used by doctors across the United States and Europe to identify the patients most at risk for fatal heart attacks have about a 50% chance of identifying the right patients, “not much better than throwing dice,” Trayanova says. The team’s model achieved an accuracy of 89% overall and 93% for patients aged 40–60, significantly outperforming current guideline-based methods.
In this work, published in Nature Cardiovascular Research, MAARS not only identifies patients at high risk but also explains the underlying factors contributing to that risk, enabling more personalized and effective treatment decisions. The team is now extending the tool to other cardiac conditions, with further clinical trials underway.
JHU HUB Announcement: (click here)
The paper may be accessed here: https://rdcu.be/eusbo
