AI Algorithm Used to Predict Odds of Hypertrophic Cardiomyopathy

A recent study conducted by researchers at Mount Sinai Fuster Heart Hospital has resulted in the creation of an AI algorithm constructed to spot signs of hypertrophic cardiomyopathy (HCM).

This program, named Viz HCM, is capable of quickly and specifically identifying patients suffering from HCM to flag them as high risk during appointments. The US Food and Drug Administration (FDA) had previously approved Viz HCM for detection of the condition on an electrocardiogram (ECG).1

What differentiates the results of this study from its existing capabilities is that Viz HCM can now provide a general percentage chance of a given patient contracting HCM. This will help undiagnosed patients have a better idea of their own risk of disease, as well as prioritizing that risk for healthcare professionals to address.1

“This is an important step forward in translating novel deep-learning algorithms into clinical practice by providing clinicians and patients with more meaningful information,” Joshua Lampert, MD, Director of Machine Learning at Mount Sinai Fuster Heart Hospital told NEJM AI. “Clinicians can improve their clinical workflows by ensuring the highest-risk patients are identified at the top of their clinical work list using a sorting tool. Patients can be better counseled by receiving more individualized information through model calibration which improves interpretability of model classification scores.”1

HCM is a common heart disease globally, affecting all genders and races. Often inherited rather than contracted, it is also a leading reason for heart transplantation. Estimates have suggested that around 750,000 Americans are affected by HCM, although only a small portion are clinically diagnosed. Additionally, while several effective management and treatment strategies have emerged over the years, most patients are unaware that they are carrying the disease until the effects have already begun to manifest.2

HCM is also a direct cause of heart failure and arrhythmias, potentially inducing sudden death. Its primary effect on the heart is thickening the ventricular wall, which can apply pressure to blood vessels and potentially obstruct blood flow. Symptoms can range from completely asymptomatic to palpitation, syncope, and in extreme cases, sudden cardiac death.3

Viz HCM was tested on almost 71,000 patients who had received an electrocardiogram between March 2023 and January 2024. The algorithm identified 1522 patients as having a positive flag for HCM; this was then reviewed by researchers to determine which patients had a confirmed diagnosis. After applying model calibration to Viz HCM, researchers determined conclusively that the calibrated probability of having HCM was correlated to the actual chance of a patient having the disease.1

This model could potentially allow for early recognition and response to HCM, allowing cardiologists to identify and explain the individual risk to each patient where prior models simply flagged them. Additionally, it could forestall symptoms of HCM before they arise, thereby preventing adverse outcomes such as sudden death or symptoms of obstructed blood flow.1

“This study reflects pragmatic implementation science at its best, demonstrating how we can responsibly and thoughtfully integrate advanced AI tools into real-world clinical workflows,” said co-senior author Girish N. Nadkarni, MD, MPH, Chair of the Windreich Department of Artificial Intelligence and Human Health and Director of the Hasso Plattner Institute for Digital Health. “It’s not just about building a high-performing algorithm—it’s about making sure it supports clinical decision-making in a way that improves patient outcomes and aligns with how are is actually delivered.”1

References

1. Mount Sinai Hospital. AI algorithm can help identify high-risk heart patients to quickly diagnose, expedite, and improve care. Eurekalert! April 22, 2025. Accessed April 22, 2025. https://www.eurekalert.org/news-releases/1081019?

2. Maron BJ, Desai MY, Nishimura RA, et al. Diagnosis and Evaluation of Hypertrophic Cardiomyopathy: JACC State-of-the-Art Review. J Am Coll Cardiol. 2022;79(4):372-389. doi:10.1016/j.jacc.2021.12.002

3. Zhang Y, Adamo M, Zou C, et al. Management of hypertrophic cardiomyopathy. J Cardiovasc Med (Hagerstown). 2024;25(6):399-419. doi:10.2459/JCM.0000000000001616