Findings from a new report by the American Heart Association (AHA) suggest that Ultromics’ EchoGo Heart Failure artificial intelligence (AI) algorithm could identify patients with heart failure and preserved ejection fraction (HFpEF) up to nine months earlier than standard clinical care.
The Oxford, UK-headquartered company’s AI tool is designed to automate the analysis and measurement of echocardiograms. By eliminating manual tracing, the tool has been developed to standardise diagnoses and reduce operator variability in the identification of HFpEF and other conditions such as cardiac amyloidosis.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
With the aim of helping health systems better understand the potential clinical and economic value of cardiovascular and stroke-focused AI tools, the AHA’s Impact Report was created by using independent clinical and image data through its AI Assessment Lab.
Designed to evaluate cardiovascular and stroke AI algorithms, the AHA’s lab and ensuing report were developed based on an independent dataset curated by real-world data and clinical AI platform Dandelion Health.
The AHA’s report findings were based on modelled data on a per-10,000-patient basis over five years.
Of the most significant findings, the report found that Ultromics’ EchoGo could have led to the identification of HFpEF an average of 263 days earlier than standard clinical care among patients who would have otherwise experienced delayed diagnosis.
The results also indicate that early intervention could result in 477 lives saved per 10,000 patients over five years.
On the economic side, the report’s findings suggest that health systems could realise cost-savings of up to $1.9m over five years when using EchoGo, with savings amounting to approximately $1,800 per patient from both a health system and payer perspective.
HFpEF is viewed as one of the most challenging forms of heart failure to diagnose given that symptoms tend to be non-specific and may not show up on standard imaging tests. With an estimated 30%–50% of patients unaware they have a heart failure diagnosis, and many not identified until the disease has already progressed in severity, this factor is exacerbated because normal ranges are predominantly derived from white male populations. This recognised diagnostic bias means HFpEF is often underdiagnosed in women and individuals from other ethnic backgrounds.
Ultromics’ chief commercial officer, Roger Owens, commented: “Since they’re so novel, AI diagnostics are evaluated differently than traditional cardiovascular technologies. Right now, many hospitals struggle to adopt them since they need clear evidence of how these models are trained, validated, and integrated into real clinical workflows before they can trust them in practice.
“The AHA’s AI Assessment Lab provides exactly what the field has been missing: clinically meaningful validation, based on independent data, that helps bridge the gap between theory and real-world application.
“For EchoGo Heart Failure, that kind of external validation is critical to building institutional trust –demonstrating not only that the model works, but that it works consistently in a diverse dataset, can be interpreted alongside physician judgment, and is ready to support decision-making at scale.”
