Daijiworld Media Network - New York
New York, Jul 5: In a breakthrough for cardiovascular medicine, US researchers have developed an advanced artificial intelligence (AI) system that dramatically improves the accuracy of identifying patients at high risk of sudden cardiac death (SCD). The new model, named Multimodal AI for Ventricular Arrhythmia Risk Stratification (MAARS), outperforms existing clinical guidelines and could reshape how doctors assess cardiac risk.
Developed at Johns Hopkins University, MAARS integrates cardiac MRI scans with a wide array of patient health data to uncover subtle warning signs that traditional methods often miss, according to a report by Xinhua News Agency. The findings were published in the journal Nature Cardiovascular Research.
The study specifically focused on hypertrophic cardiomyopathy (HCM) — a common genetic heart condition and one of the leading causes of sudden cardiac death in young individuals.

"Right now, we're losing patients in the prime of their life because existing tools fail to identify who truly needs protection. At the same time, others are living with defibrillators they may never actually need," said senior author Natalia Trayanova, an expert in AI-based cardiology at Johns Hopkins.
While current US and European clinical guidelines have an average predictive accuracy of around 50%, the MAARS model achieved an impressive 89% overall accuracy. For patients aged 40 to 60 — the age group most at risk — the model was even more accurate, reaching 93%.
The secret to MAARS' success lies in its use of contrast-enhanced MRI scans, which highlight scar tissue in the heart — a key predictor of arrhythmias leading to sudden cardiac death. These scans are traditionally challenging for doctors to interpret, but MAARS uses deep learning to analyze subtle patterns and derive risk factors that human eyes might miss.
“This model could transform how we care for patients,” said Dr. Jonathan Chrispin, a co-author and cardiologist at Johns Hopkins. “Compared to current risk calculators, MAARS significantly enhances our ability to identify those who need intervention the most.”
Looking ahead, the research team plans to expand the model’s application to other serious heart conditions, such as cardiac sarcoidosis and arrhythmogenic right ventricular cardiomyopathy, potentially making it a game-changer in personalized cardiac care.