Scientists Have Developed An AI Model To Help Predict Sudden Cardiac Risks 'Most Doctors Can't See'

Every year, thousands of seemingly healthy people—often young, active, and without obvious warning signs—die suddenly due to cardiac arrest. For decades, doctors have struggled to reliably identify which patients with heart conditions are at high risk and who might be unnecessarily undergoing invasive interventions. That may be about to change.In a breakthrough that could transform how we predict—and prevent—sudden cardiac death, scientists at Johns Hopkins University have developed an artificial intelligence model that vastly outperforms current clinical standards in identifying people most at risk. Their new system, known as MAARS (Multimodal AI for Arrhythmia Risk Stratification), not only forecasts risk with up to 93% accuracy in vulnerable age groups, but also explains why someone is high risk—something most algorithms fail to do.The focus of the study is hypertrophic cardiomyopathy (HCM), one of the most common inherited heart conditions. It affects around 1 in 200 to 500 people globally and is a leading cause of sudden cardiac death in athletes and young adults. While most individuals with HCM live normal lives, a subset is at high risk for lethal arrhythmias—heart rhythm disturbances that can cause the heart to stop without warning. And here’s the catch: right now, doctors only have a 50-50 shot at predicting who will be affected.“Currently we have patients dying in the prime of their life because they aren’t protected,” said Dr. Natalia Trayanova, senior author of the study and a leading figure in AI cardiology research. “And others are putting up with defibrillators for the rest of their lives with no benefit.”Trayanova is referring to implantable cardioverter defibrillators (ICDs)—tiny devices inserted into the chest that deliver electric shocks to correct abnormal heart rhythms. They save lives in the right patients but come with physical, emotional, and financial burdens when used unnecessarily.The need for a more precise, personalized tool has never been greater.What the AI Model Does Differently?Published in Nature Cardiovascular Research, the new model represents a significant departure from traditional clinical guidelines used across the US and Europe.MAARS doesn’t rely on a single data source. Instead, it analyzes a multimodal spectrum of information—ranging from electronic health records and patient histories to contrast-enhanced cardiac MRI images that reveal scarring, or fibrosis, within the heart.Scarring is a key factor in determining sudden death risk in HCM. But interpreting these raw images is extremely challenging for even seasoned cardiologists. That’s where AI has the edge.“People have not used deep learning on those images,” Trayanova explained. “We are able to extract this hidden information in the images that is not usually accounted for.”The AI essentially spots dangerous patterns in the heart’s scar tissue that the human eye—and even conventional software—can’t see.How Accurate Is It?In clinical tests involving real-world patients from Johns Hopkins Hospital and Sanger Heart & Vascular Institute in North Carolina, the results were staggering:Traditional clinical guidelines correctly predicted sudden cardiac death risk only 50% of the time.The MAARS model achieved an overall accuracy of 89%.In patients aged 40 to 60—a group particularly vulnerable to undetected risk—the model was 93% accurate.What makes this even more valuable is its ability to provide explanations. The system doesn't just say "this patient is high risk"—it breaks down the why, giving cardiologists critical information to tailor treatment plans.“This significantly enhances our ability to predict those at highest risk compared to our current algorithms,” said co-author Dr. Jonathan Crispin, a Johns Hopkins cardiologist. “It has the power to transform clinical care.”Why This Could Be Transformative?MAARS isn't the first AI model from Trayanova’s lab. In 2022, her team built another tool that provided survival predictions for patients with prior heart attacks, known as infarcts. But this latest model breaks new ground by tackling one of the most elusive forms of cardiac risk—arrhythmias caused by scarring in inherited heart conditions. The potential benefits are wide-ranging:Lives saved by identifying at-risk patients who might otherwise be missed.Better quality of life for patients who avoid unnecessary ICD implantation.More personalized treatment plans based on detailed, AI-generated insight.Importantly, the model was trained and validated across diverse demographics, showing consistent performance regardless of age, gender, or ethnicity.The researchers aren’t stopping here. They plan to expand MAARS to include other forms of arrhythmia-related heart diseases, such as cardiac sarcoidosis and arrhythmogenic right ventricular cardiomyopathy—conditions that also carry a high risk of sudden death but suffer from diagnostic ambiguity.They’re also working to test the model in larger, more varied populations to move it closer to clinical adoption.A New Era in Cardiology?Artificial intelligence has long been hyped as the future of medicine. But MAARS is more than hype—it’s a working proof of concept that shows how deep learning can complement medical expertise, not replace it.AI may soon become your cardiologist’s most powerful diagnostic tool—one that sees what even the best-trained human eyes might miss. And when lives are on the line, that kind of clarity could mean everything.

Scientists Have Developed An AI Model To Help Predict Sudden Cardiac Risks 'Most Doctors Can't See'

Quick Stories

1 In 8 Babies Born Premature In India, 17% Having Low Birth Weight— Health Survey Blames Air Pollution

New 'Frankenstein' Covid Variant Spreads Across 38 Countries: Unique Covid Symptom Surfaces In UK Outbreak

Nipah Outbreak Alert In Kerala: One Death Suspected, 345 Under Surveillance As Three Districts On High Alert

Health Conditions A-Z

Inner Child: Being Left Out And Rejected In Childhood Becomes A Social Seed For Deeper Connections

What Is Innotox? The ‘Korean Botox’ People Are Injecting At Home—And Why Doctors Are Worried

Loneliness Is Linked To 100 Deaths Every Hour, Reveals WHO