Can AI predict heart failure years earlier? Inside Anumana’s breakthrough AHA 2025 results

Anumana, the privately held health technology company co-founded by nference and Mayo Clinic, has unveiled breakthrough clinical data suggesting that artificial intelligence may be capable of detecting heart failure risk years before symptoms appear. At the 2025 American Heart Association (AHA) Scientific Sessions, the company presented a late-breaking featured science study showing that its ECG-AI platform, when applied to standard 12-lead electrocardiograms, can significantly outperform traditional clinical risk models in predicting near-term heart failure.

The findings, simultaneously published in the Journal of the American College of Cardiology, highlight the growing clinical role of AI in preventive cardiology. Anumana’s data-driven momentum comes as the company expands its platform capabilities beyond diagnostics into perioperative imaging and intraoperative decision support. These next-generation tools are being developed with fresh backing from Boston Scientific Corporation, which participated in Anumana’s recently completed Series C financing round.

The dual push, both scientific and strategic, marks a major inflection point in Anumana’s evolution from a diagnostics innovator into a comprehensive AI-powered cardiovascular solutions provider.

How does Anumana’s ECG-AI outperform standard models in identifying near-term heart failure risk?

The featured study titled “Enhanced Prediction of Incident Heart Failure Using Artificial Intelligence-Driven Analysis of 12-Lead Electrocardiogram Waveforms” drew from a pooled cohort of over 14,000 patients across three major U.S. longitudinal databases: the Framingham Heart Study, the Multi-Ethnic Study of Atherosclerosis, and the Cardiovascular Health Study. Investigators integrated Anumana’s proprietary ECG-AI model with the PREVENT-HF clinical risk equation and observed significant reclassification power. Up to 12.5 percent of individuals who were not flagged by clinical scores alone were newly classified into higher-risk tiers.

The data indicated that individuals with a positive ECG-AI screen were over 20 times more likely to develop heart failure within three years than those with a negative result. The model achieved an area under the curve (AUC) of 0.944, with sensitivity of 90.2 percent and specificity of 85.1 percent. These metrics, combined with its non-invasive and widely deployable modality, make ECG-AI a compelling candidate for population-wide screening programs.

The algorithm’s strength lies in its ability to extract early electrical signals of left ventricular dysfunction that precede symptomatic deterioration, thus offering the potential for preventive therapy and long-term outcome improvement.

What does the clinical community say about the broader implications of ECG-AI in cardiology?

Dr. Akshay S. Desai, Director of the Heart Failure Disease Management Program at Brigham and Women’s Hospital and lead investigator on the study, explained that the ability of ECG-AI to detect early cardiac dysfunction through subtle waveform analysis could shift the paradigm in heart failure care. By incorporating AI into standard ECG interpretation, clinicians may be able to initiate treatment years before symptom onset.

The study was conducted in collaboration with the National Heart, Lung, and Blood Institute’s HeartShare/AMP Heart Failure Program and used the BioData Catalyst research platform. This ensured access to deeply phenotyped, reproducible longitudinal data, supporting scientific rigor and regulatory reliability. The approach reflects a growing emphasis on early intervention strategies using AI-enhanced clinical tools.

Anumana executives also stressed that the new study advances their long-term mission to move from disease detection to disease prevention. By embedding ECG-AI into routine clinical workflows, the company aims to expand the role of artificial intelligence in frontline primary care and outpatient settings.

What role is Boston Scientific playing in Anumana’s expansion into perioperative cardiac AI?

Anumana’s recent Series C funding round, which included participation from Boston Scientific Corporation, marks a strategic pivot from diagnostic cardiology to perioperative and intraoperative support. The financing will be used to develop multimodal generative AI technologies that enhance procedural planning and real-time guidance in cardiac interventions.

These new offerings are designed for integration into electrophysiology, interventional cardiology, and structural heart procedures. Early targets include support systems for atrial fibrillation treatments such as cardiac ablation and left atrial appendage closure. The perioperative tools will offer predictive analytics and real-time visualization, facilitating more precise, data-informed care delivery.

Chief Executive Officer Maulik Nanavaty stated that Anumana’s diagnostic foundation provided a strong springboard for the new expansion. He added that the application of generative AI to clinical imaging and surgical decision-making represents a natural evolution of the company’s strategy. The integration of software-as-a-medical-device (SaMD) with hardware-enabled procedures could ultimately streamline workflows and reduce cognitive load for surgical teams.

What other studies were presented at AHA 2025 showcasing Anumana’s cardiovascular AI capabilities?

In addition to its featured study on heart failure, Anumana presented three additional abstracts that reinforced the versatility of ECG-AI across cardiovascular indications. In one multicenter study involving five U.S. health systems, the algorithm demonstrated 84 percent sensitivity and 72 percent specificity for pulmonary hypertension detection, further validating its generalizability across disease states.

A separate retrospective analysis revealed that in patients who were later diagnosed with pulmonary arterial hypertension or chronic thromboembolic pulmonary hypertension, more than 74 percent had at least one ECG flagged by the AI as positive well before the official diagnosis. This suggests the algorithm could play a critical role in shortening diagnostic delays and initiating earlier treatment pathways.

In another abstract, Anumana researchers used deidentified electronic health record data to examine the correlation between parity (lifetime number of pregnancies) and the risk of Takotsubo cardiomyopathy, a stress-induced cardiac condition. This work highlights the company’s broader push into AI-driven epidemiological and lifestyle-linked cardiovascular research.

Together, these abstracts position Anumana as a category-defining player not only in diagnostics but in predictive cardiology and population health analytics.

What is the current regulatory and market status of ECG-AI in the U.S.?

Anumana’s ECG-AI LEF algorithm is already cleared by the U.S. Food and Drug Administration and became eligible for reimbursement starting January 2025. This opens the door for integration into community clinics, telemedicine platforms, and general practitioner workflows. The tool can be deployed on standard 12-lead ECGs, which are already ubiquitous in primary care, making adoption relatively frictionless.

The availability of reimbursement creates a critical bridge between research and real-world implementation. It signals payer confidence in the clinical utility of the algorithm, which could accelerate adoption among health systems looking to reduce downstream costs through early intervention.

Anumana’s roadmap includes expansion into additional geographies, pursuit of new FDA clearances for its multimodal SaMD solutions, and potential partnerships that deepen the integration of AI across the entire cardiovascular care continuum.

What does investor sentiment suggest about Anumana’s direction and valuation outlook?

Investor sentiment toward artificial intelligence in cardiology has strengthened over the past 24 months, particularly following FDA approvals and publication of peer-reviewed clinical studies. Institutional investors increasingly see cardiovascular AI as a convergence point of structured data availability, validated endpoints, and measurable outcomes.

The addition of Boston Scientific Corporation to Anumana’s cap table is viewed by industry watchers as a strong validation of both product-market fit and long-term viability. By aligning its growth with a globally recognized cardiac device leader, Anumana is positioned to enhance its go-to-market strategy, secure regulatory support, and deepen clinical adoption.

While Anumana remains a private entity, its clinical milestones, investor mix, and expanding product portfolio suggest a trajectory that would be well-aligned with future IPO conditions. Analysts expect its valuation to rise substantially as further real-world data accrues and commercial deployment scales.

What does the future look like for AI-powered cardiac care platforms like Anumana?

Anumana is part of a growing class of digital health innovators aiming to embed AI into every layer of cardiac care, from diagnostics and screening to procedure guidance and longitudinal risk monitoring. By using AI to interpret electrocardiograms, guide catheter placement, and predict surgical outcomes, the company seeks to replace reactive models with anticipatory care strategies.

The near-term focus remains on expanding ECG-AI adoption and establishing real-world evidence frameworks. Long-term, the vision includes full multimodal integration combining waveform signals, imaging data, and procedural inputs into unified decision support tools.

If successful, this approach could reduce hospital readmissions, optimize resource use, and deliver precision care at scale. As cardiovascular disease remains the leading cause of mortality globally, Anumana’s AI platform may soon become a central force in shifting global outcomes.

Key takeaways from Anumana’s AHA 2025 data and strategic AI platform expansion

• Anumana presented late-breaking data at AHA 2025 showing that its ECG-AI algorithm significantly outperformed standard risk models in predicting near-term heart failure.
• The AI model reclassified up to 12.5% of individuals into higher-risk groups not flagged by traditional PREVENT-HF scores, with a 20x higher risk of developing heart failure within three years.
• Clinical validation was supported by pooled data from over 14,000 patients across major U.S. longitudinal cardiovascular cohorts.
• The ECG-AI algorithm achieved an AUC of 0.944 with 90.2% sensitivity and 85.1% specificity, demonstrating strong predictive accuracy.
• Anumana’s ECG-AI LEF algorithm is FDA-cleared and became eligible for U.S. reimbursement in January 2025, supporting adoption in primary and outpatient care settings.
• The company raised a new Series C funding round with participation from Boston Scientific Corporation, signaling strong institutional support and enabling expansion into perioperative cardiac care.
• Anumana is developing generative AI imaging tools and real-time intraoperative decision support for use in cardiac ablation and left atrial appendage closure procedures.
• Additional AHA 2025 presentations included successful AI-driven studies on pulmonary hypertension and stress-induced cardiomyopathy, showcasing the versatility of the platform.
• Institutional sentiment remains strong, with analysts viewing Anumana’s trajectory as aligned with broader shifts toward preventive and AI-enabled cardiovascular care.