HQ Team
July 7, 2025: A new artificial intelligence model developed by researchers at Johns Hopkins University can predict whether patients are likely to experience a cardiac arrest much better than doctors, who rely on limited clinical symptoms.
Current clinical guidelines used by doctors across the US 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,” said Natalia Trayanova, a researcher focused on using artificial intelligence in cardiology.
The researchers’ multimodal artificial intelligence system predicted individual hypertrophic cardiomyopathy patients’ risk for sudden cardiac death by analysing a variety of medical data and records.
Hypertrophic cardiomyopathy is most often caused by abnormal genes in the heart muscle. These genes cause the walls of the heart chamber (left ventricle) to become thicker than normal.
The thickened walls may become stiff, and this can reduce the amount of blood taken in and pumped out to the body with each heartbeat.
Scarring of heart
People with hypertrophic cardiomyopathy develop fibrosis, or scarring, across their heart, and it is the scarring that elevates their risk of sudden cardiac death.
A heart attack occurs when blood flow to the heart is blocked, causing damage to heart tissue — the heart usually keeps beating. A cardiac arrest happens when the heart suddenly stops beating due to an electrical problem, causing loss of consciousness and no pulse, and requires immediate cardiopulmonary resuscitation and defibrillation to survive.
Many patients with hypertrophic cardiomyopathy will live normal lives, but a percentage are at significantly increased risk for sudden cardiac death. It’s been nearly impossible for doctors to determine who those patients are.
Hypertrophic cardiomyopathy is one of the most common inherited heart diseases, affecting one in every 200 to 500 individuals worldwide, and is a leading cause of sudden cardiac death in young people and athletes.
MRI images analysed
After analysing medical records, for the first time, researchers explored all the information contained in the contrast-enhanced MRI images of the patient’s heart.
While doctors haven’t been able to make sense of the raw magnetic resonance images, the AI model zeroed right in on the critical scarring patterns, according to a statement.
“People have not used deep learning on those images,” Trayanova said. “We can extract this hidden information in the images that is not usually accounted for. We can predict with very high accuracy whether a patient is at very high risk for sudden cardiac death or not.”
The researchers tested the model against real patients treated with the traditional clinical guidelines at Johns Hopkins Hospital and Sanger Heart & Vascular Institute in North Carolina.
Compared to the clinical guidelines that were accurate about half the time, the AI model was 89% accurate across all patients, according to the statement.
93% accuracy
It was 93% accurate for people between 40 and 60 years old — the population among hypertrophic cardiomyopathy patients is most at risk for sudden cardiac death.
The researchers said the new model could save many lives and also spare many people from unnecessary medical interventions, including the implantation of unneeded defibrillators.
The AI model can also describe why patients are high risk so that doctors can tailor a medical plan to fit their specific needs.
New types of heart diseases
“Our study demonstrates that the AI model significantly enhances our ability to predict those at highest risk compared to our current algorithms and thus has the power to transform clinical care,” says co-author Jonathan Chrispin, a Johns Hopkins cardiologist.
The team plans to further test the new model on more patients and expand the new algorithm to use with other types of heart diseases, including cardiac sarcoidosis and arrhythmogenic right ventricular cardiomyopathy.
An AI model is a computer program trained on large datasets to recognise patterns, make predictions, or make decisions autonomously without human intervention.
It uses complex algorithms and mathematical techniques to analyse data and extract meaningful insights, enabling it to perform tasks that typically require human intelligence, such as reasoning, problem-solving, natural language processing, and image recognition.
