AI’s growing impact in heart failure detection

Professor Paul Leeson  Professor of Cardiovascular Medicine at the University of Oxford, and co-founder and CMO at Ultromics.

Rapid advancements in artificial intelligence (AI) have the potential to revolutionize healthcare by seamlessly integrating into clinical practice. Embracing AI enables healthcare professionals to unlock new possibilities in diagnostics, particularly for complex conditions like heart failure, detecting disease faster and more accurately than current clinical practice.

Professor Paul Leeson, a renowned expert in cardiovascular imaging, shares insights on innovation from Ultromics. Leeson is a professor of cardiovascular medicine at the University of Oxford, director of the Oxford Cardiovascular Clinical Research Facility, and co-founder and Chief Medical Officer (CMO) of Ultromics.

"AI is revolutionizing heart failure detection, bringing transformative changes to cardiology, with AI, vast amounts of data can be analyzed efficiently, aiding in identifying diseases at their nascent stages."

"This not only speeds up the diagnostic process but also enhances accuracy, allowing for early interventions and assisting doctors in making further care decisions," Leeson said.

Globally, up to 64 million people are living with heart failure and up to 50% of these cases are Heart Failure with Preserved Ejection Fraction (HFpEF) ^[1,2]. HFpEF is notoriously difficult to detect. Patients with HFpEF are often undiagnosed due to challenges in its detection caused by multiple comorbidities, conflicting guidelines, and difficulties in detecting imaging markers and diastolic function ^[3]. If left untreated, HFpEF can lead to serious complications that require hospitalization. Early detection can help patients avoid hospitalization, but it’s often hard to diagnose until the heart failure becomes serious.

"There certainly are unmet needs for heart failure, one of them is to detect heart failure early before patients are at risk of worsening heart failure and hospitalization," Leeson remarked.

Data from echocardiography provides a cornerstone in the management of heart failure. However, echocardiogram videos are inherently difficult to interpret. Clinical risk scores currently used to detect HFpEF present their own set of challenges, and patients are often classified as indeterminate ^[4]. This problem is further compounded by the national shortage of sonographers in the US and up to 64% of HFpEF cases may be undiagnosed ^[2,5].


Utilizing AI to Improve HFpEF Detection

Clinical studies have shown that AI is very accurate at identifying HFpEF, in both retrospective and real-world settings ^[4,6]. In a pivotal study, published in JACC Advances, it was demonstrated that EchoGo^® Heart Failure had an overall sensitivity of 88% and a specificity of 82%, reducing indeterminate diagnosis from 64% to 7% ^[4]. Another study, published in the Journal of Cardiac Failure, showed the utility of using EchoGo^® HF to detect HFpEF in a real-world echo clinic setting ^[6]. These studies have earned the device breakthrough device designation and subsequent clearance by the FDA, and reimbursement from CMS.

"Ultromics' EchoGo® stands out among AI algorithms for its unique ability to identify disease from a single apical four-chamber view using echocardiography."


"Developed with Mayo Clinic, the technology can detect signals of diastolic dysfunction that are not visible to sonographers and classifies whether HFpEF is present or not. Through its advanced AI capabilities, Ultromics' EchoGo^® Heart Failure revolutionizes diagnostic precision, ushering new opportunities for timely and accurate diagnosis and expedites effective treatment options now available to patients," Leeson noted.

In 2022, Empagliflozin received FDA approval for treating HFpEF, following the EMPEROR trial results that demonstrated 21% lower risk of hospitalizations. ^[7]

Seamless, Scalable

EchoGo^® Heart Failure is now available for sites in the United States and offers a unique solution for clinicians in the market. 

"As a cardiologist, what's exciting about EchoGo^® to me is its simplicity and that it's ready to be universally deployed to any practice," Leeson says. "It doesn’t require any hardware or devices – it’s a SaaS platform and connects through the cloud via a wireless connection. This is an elegant solution that allows clinicians to access powerful AI tools rapidly."

"Part of our decision to develop EchoGo^® using a SaaS platform was to improve the speed with which the product could be scaled to ensure it could be easily implemented into a variety of clinical settings, especially in the community," Leeson added.

The use of a SaaS model is also appealing as it allows for enhanced collaboration between clinicians and remote access to patient reports, ensuring patients receive the best possible care in minimal time and allowing doctors to access these reports from multiple locations. EchoGo^® automatically integrates into the user’s PAC system and clinicians automatically receive reports. 

EchoGo^® Heart Failure is now reimbursable in both inpatient and outpatient settings under new codes that came into effect at the end of last year. This will increase the accessibility of the Ultromics platform and economic savings for healthcare institutions, which Ultromics customers have already started to benefit from.

The way forward

EchoGo^® Heart Failure represents a significant advancement in the detection and management of heart failure. By leveraging AI, the technology can support healthcare professionals with precision detection of heart failure subtypes, starting with HFpEF.

The collaboration with Mayo Clinic and the rigorous validation of the algorithm underscores its reliability and efficacy. The economic and clinical benefits are substantial, promising to reduce healthcare costs while improving patient care.

As EchoGo^® Heart Failure becomes more integrated into clinical practice, it stands to transform the current workflow, making early and accurate diagnosis more accessible and effective. 

Learn more about EchoGo® Heart Failure: https://www.ultromics.com/products/echogo-heart-failure

References:

1. Groenewegen A, Rutten FH, Mosterd A, Hoes AW. Epidemiology of heart failure. European Journal of Heart Failure [Internet]. 2020;22(8):1342–56.
2. Pfeffer MA, Shah AM, Borlaug BA. Heart Failure With Preserved Ejection Fraction In Perspective. Circulation Research. 2019;124(11):1598–617.
3. Ho JE, Redfield MM, Lewis GD, Paulus WJ, Lam CSP. Deliberating the Diagnostic Dilemma of Heart Failure With Preserved Ejection Fraction. Circulation. 2020;3;142(18):1770–80.
4. Akerman AP, Mihaela Porumb, Scott CG, Beqiri A, Agisilaos Chartsias, Ryu AJ, et al. Automated Echocardiographic Detection of Heart Failure With Preserved Ejection Fraction Using Artificial Intelligence. JACC. 2023;2(6):100452–2.
5. Won D, Walker J, Horowitz R, Bharadwaj S, Carlton E, Gabriel H. Sound the Alarm: The Sonographer Shortage is Echoing Across Healthcare. Journal of ultrasound in medicine. 2024
6. Karnik A, Jankowski M, Narang A. Unmasking HFpEF With Artificial Intelligence: A Disruptive Opportunity for Disease Detection. Journal of cardiac failure. 2024
7. Anker SD, Butler J, Filippatos G, Ferreira JP, Bocchi E, Böhm M, et al. Empagliflozin in Heart Failure with a Preserved Ejection Fraction. New England Journal of Medicine [Internet]. 2021;385(16).