AI tool improves accuracy in detecting heart disease

A team of researchers at Mount Sinai Hospital in New York has successfully calibrated an AI tool to more accurately assess the likelihood of hypertrophic cardiomyopathy (HCM) in patients.

By assigning specific probability scores, the AI model now offers clearer guidance to clinicians and patients regarding disease risk.

HCM, a thickening of the heart muscle that affects around one in 200 people globally, can lead to serious complications such as heart failure or sudden cardiac death.

The Viz HCM algorithm, already approved by the US Food and Drug Administration, previously provided vague classifications like ‘suspected HCM.’ Thanks to model calibration, clinicians can now give patients more precise estimates—for instance, a 60% probability of having the condition.

Researchers ran the algorithm on nearly 71,000 patients who had undergone electrocardiograms between March 2023 and January 2024. Out of these, 1,522 were flagged by the AI, with further review of medical records and imaging confirming diagnoses.

The results validated that the newly calibrated probabilities closely reflected real-world outcomes, improving the tool’s accuracy and practical utility.

Experts say this advancement enhances clinical workflows by helping doctors prioritise patients based on their actual risk levels.

Beyond technological innovation, the study marks a step forward in integrating AI responsibly into everyday clinical practice—making healthcare more personalised, interpretable, and effective.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!