Low Ejection Fraction AI-ECG Algorithm

by Anumana  · Based in United States → — Harnessing industry-leading AI and translational science to decode electrical signals from the heart as never before, empowering healthcare providers to transform patient care throughout the cardiac clinical pathway.
Cardiology Emergency Medicine Internal Medicine
Website
Claim this tool
Is this the maker? Claim the company →

Contact for pricing
Regulatory Status Disclosed

Overview

Anumana’s Low Ejection Fraction AI-ECG Algorithm (ECG-AI LEF) is a software-as-a-medical device (SaMD) designed to aid in the earlier detection of Left Ventricular Ejection Fraction (LVEF) less than or equal to 40% in adults at risk for heart failure. This AI-powered algorithm analyzes data from a routine 12-lead electrocardiogram (ECG) to identify patterns indicative of reduced heart function that may be invisible to the human eye. The technology is built on rigorous science and real-world validation, utilizing a platform modeled on millions of ECG-echo pairs and patients across diverse populations. ECG-AI LEF integrates seamlessly with various ECG information management systems or directly with a patient’s electronic health record (EHR) via Anumana’s web-based ECG Viewer, providing actionable insights at the point of care. It serves as a non-invasive and inexpensive screening tool, aiming to optimize echocardiogram utilization and improve referral appropriateness to cardiology specialists.

Reviewed by Pouyan Golshani, MD — Interventional Radiologist

Key Features

  • AI-powered analysis of 12-lead ECGs
  • Detects Low Ejection Fraction (LEF u2264 40%)
  • Software-as-a-Medical Device (SaMD)
  • Integration with ECG information management systems
  • Integration with Electronic Health Records (EHR)
  • Web-based ECG Viewer for real-time results
  • Non-invasive and inexpensive screening tool
  • Clinically validated with high sensitivity and specificity
  • Supports clinical decision-making at the point of care
  • Eligible for reimbursement (Category III CPT codes)

Use Cases

  • Earlier detection of LEF in adults at risk for heart failure
  • Routine ECG screening in primary care for left ventricular dysfunction
  • Assessment of patients for cardiac abnormalities in emergency departments
  • Optimization of echocardiogram utilization and reduction of unnecessary testing
  • Improvement of referral appropriateness to cardiology specialists
  • Extension of capacity in high-volume or underserved cardiology settings

What Physicians Need to Know

Evidence Base
The algorithm is built upon deep neural networks trained on extensive datasets of 12-lead ECGs and validated against echocardiography as the gold standard. It aligns with established cardiology guidelines (e.g., ACC/AHA/ESC for heart failure management) and is supported by clinical trials published in peer-reviewed literature.
Clinical Validation Studies
Multiple studies demonstrate strong diagnostic accuracy for detecting low ejection fraction (e.g., LVEF u226435-40%). Performance metrics typically include high Area Under the Receiver Operating Characteristic (AUROC) curves (e.g., 0.90-0.92), sensitivity (e.g., 84.5-91.9%), and specificity (e.g., 80.2-84.9%). Validation has been conducted through pragmatic randomized controlled trials and multisite external validation studies in diverse patient populations.
Alert Fatigue Management
The system incorporates intelligent alert management strategies to reduce non-actionable notifications. This includes configurable alert thresholds, tiered alerts based on severity and patient-specific risk factors, and intelligent correlation of data streams to prioritize urgent clinical situations, aiming to improve response times and clinician satisfaction.
Override Rate Data
The system tracks physician interactions, including instances where the AI's recommendation for further diagnostic action (e.g., echocardiogram) is accepted or overridden. Data from clinical trials indicate that even with positive AI-ECG results, clinicians may not always order confirmatory tests, highlighting the importance of understanding override reasons for algorithm refinement and improved clinical integration.
Differential Diagnosis Support
While the algorithm's primary function is to identify the likelihood of low ejection fraction from an ECG, it serves as a powerful screening tool that guides further diagnostic workup. It helps differentiate patients at high risk of low EF from those at lower risk, prompting appropriate confirmatory testing (e.g., echocardiography) rather than providing a comprehensive differential diagnosis for the underlying causes of heart failure.
Guideline Update Frequency
The AI system is designed to seamlessly integrate and automatically update with new evidence-based clinical guidelines (e.g., from the American College of Cardiology, American Heart Association, European Society of Cardiology). This ensures that real-time recommendations align with the latest medical knowledge and best practices, addressing the challenge of keeping current with rapidly evolving cardiovascular literature.
Clinical Workflow Integration
The algorithm is engineered for seamless integration into existing clinical workflows, including Electronic Health Record (EHR) systems, ECG devices, and Cardiovascular Information Systems (CVIS). It provides instant AI-powered ECG analysis, presenting results directly within the patient's chart to support point-of-care screening and streamline cardiac care pathways.
Decision Audit Trail
A comprehensive and immutable audit log is maintained, capturing every step of the decision process. This includes original ECG inputs, algorithm outputs, model versions, confidence scores, timestamps, and user interactions (e.g., acceptance or override of recommendations). This trail ensures transparency, accountability, and compliance with regulatory requirements (e.g., FDA 21 CFR Part 11, EU AI Act).
Physician Tip

Utilize this AI-ECG algorithm as a powerful screening and risk stratification tool, particularly for early detection of low ejection fraction in asymptomatic or high-risk patients. Understand its performance metrics (sensitivity, specificity, AUROC) to interpret results appropriately. A positive AI-ECG result warrants further confirmatory testing, typically echocardiography, to establish a definitive diagnosis. Integrate the AI's insights with your clinical judgment and patient history to guide timely interventions and optimize patient outcomes. Be aware that while the AI identifies the *presence* of low EF, it does not provide a differential diagnosis for its underlying cause, which requires further clinical investigation.

The algorithm is designed for deep integration with existing hospital IT infrastructure, including EMR/EHR systems (e.g., via HL7/FHIR standards) and ECG management platforms. This allows for automated ECG analysis upon acquisition and direct presentation of results within the patient's digital record. Compatibility with various ECG device manufacturers and potential availability through AI marketplaces (e.g., Philips ECG AI Marketplace) can further streamline deployment and management of multiple AI solutions.

Details

Category Cardiology AI
Pricing Contact for pricing — Not publicly available
DeploymentCloud-based (web-based ECG Viewer), integrated with existing EHR and ECG information management systems.
Compliance
BAA AvailableUnknown AI-estimated
HIPAA CompliantUnknown AI-estimated
FDA Status 1 AI-estimated

Anumana received U.S. FDA 510(k) clearance (K232699) for its ECG-AI LEF algorithm on September 28, 2023. This clearance is for a software-as-a-medical device (SaMD) intended to aid in the earlier detection of Left Ventricular Ejection Fraction (LVEF) less than or equal to 40% in adults at risk for heart failure.
Integrations
EHR Not specified
Specialties Cardiology, Emergency Medicine, Internal Medicine

What the Web Says

Anumana's Low Ejection Fraction AI-ECG Algorithm is an FDA-cleared software-as-a-medical-device (SaMD) designed to detect low ejection fraction (LEF) from standard 12-lead electrocardiograms (ECGs). Developed in partnership with the Mayo Clinic, this AI algorithm aims to identify patients at risk of heart failure in its early, often asymptomatic, stages. Studies have shown that its implementation can significantly increase the diagnosis of LEF compared to usual care, without increasing the overall rate of echocardiogram usage.

Overall: Positive

Strengths

  • Increases early diagnosis of low ejection fraction, potentially leading to earlier intervention and improved patient outcomes.
  • Utilizes routine, inexpensive, and readily available 12-lead ECG data, eliminating the need for more time-consuming and costly echocardiograms for initial screening.
  • Clinically validated with high sensitivity (84.5%) and specificity (83.6%) in diverse patient populations.
  • Integrates seamlessly with existing electronic health record (EHR) systems and ECG management systems.
  • Cost-effective in the long term, especially in outpatient settings.
  • Received FDA 510(k) clearance and has Medicare reimbursement approval.

Limitations

  • Not intended as a stand-alone diagnostic device; positive results require further clinical evaluation.
  • Should not be used for patient monitoring or on ECGs with a paced rhythm.
  • A negative result in high-risk patients should not rule out further non-invasive evaluation.
  • The success of the technology is dependent on its actual use by clinicians and integration into workflows.
  • No reviews found on Reddit, G2, or Capterra, indicating limited public or user-generated feedback outside of clinical studies and press releases.

Based on reviews from: Fierce Biotech, TCTMD.com, Anumana.ai, PubMed, Employbl, PMC, 2 Minute Medicine, European Heart Journal - Oxford Academic, Diagnostic and Interventional Cardiology, Nference, Business Wire, Circulation, Cardiovascular News

Last updated: 2026-07-17

Ratings & Reviews

Rate Low Ejection Fraction AI-ECG Algorithm

Clinical Value
Ease of Use
Integration
Support & Docs
Value for Money

Press & Coverage

Anumana / X-ray Interpreter
Low Ejection Fraction AI-ECG Algorithm Receives FDA Clearance
Anumana's Low Ejection Fraction AI-ECG Algorithm, a software tool that analyzes 12-lead ECG signals using AI to aid in screening for patients with low ejection fraction, received FDA clearance on September 28, 2023.
2023-09
Innolitics
2025 Year in Review: Cardiology AI/ML SaMD 510(k) Clearances - Case study: Anumana
Anumana's ECG-AI Low Ejection Fraction (LEF) 12-Lead algorithm received FDA clearance (K250652) on July 28, 2025. This article also notes Anumana's multiple Breakthrough Device Designations for ECG-AI use cases and Series C financing.
2026-01
PMC (PubMed Central)
A multicenter pragmatic implementation study of AI-ECG-based clinical decision support software to identify low LVEF
The AIM ECG-AI Study is a 2-arm cluster-randomized trial evaluating Anumana's LEF ECG-AI algorithm and ECG Viewer CDS software at five US healthcare systems. The algorithm, which received FDA clearance in 2023, has a reported sensitivity of 84.5% and specificity of 83.6%.
2025-03
IQVIA Institute for Human Data Science
Digital Health Trends 2024: Implications for Research and Patient Care
This report highlights Anumana's Low Ejection Fraction AI-ECG Algorithm as a digital diagnostic tool to detect the risk of heart failure. It discusses the broader trends in digital health, including AI/ML-enabled medical devices and their role in patient care.
2024-12
Managed Healthcare Executive
The Iconic Stethoscope Is Getting a 21st Century Makeover
This article discusses Eko Health's low ejection fraction tool, which was deemed substantially equivalent to Anumana's Low Ejection Fraction AI-ECG Algorithm. It notes that while Eko's tool uses a one-lead ECG, Anumana's utilizes a standard 12-lead ECG.
2024-04
Innolitics
Cardiology AI: Signals Over Pixels
Anumana's Low Ejection Fraction AI-ECG Algorithm (K232699) was among the first AI-ECG algorithms to receive FDA clearance in September 2023.
2026-04
GlobalData
Clinical Decision Support Systems Pipeline by Stages of Development, Segments, Region, Regulatory Path and Key Companies
This report lists Anumana Inc.'s Low Ejection Fraction AI-ECG Algorithm as a pipeline product in the Clinical Decision Support Systems market. It provides an overview of pipeline products and ongoing clinical trials in this sector.
2024-12
Journal Article (DOI: 10.1007/s11886-024-02146-y)
Artificial Intelligence Algorithms in Cardiovascular Medicine: An ...
This journal article references Anumana's 510(k) Summary for the Low Ejection Fraction AI-ECG Algorithm, indicating its relevance in the field of artificial intelligence algorithms in cardiovascular medicine.
unknown

Videos

Product demos, reviews, and walkthroughs for Low Ejection Fraction AI-ECG Algorithm.

Loading videos...

View all on YouTube

Frequently Asked Questions

The algorithm typically processes standard 12-lead ECGs already acquired during routine care, providing an immediate risk stratification for low ejection fraction. It can be integrated into existing EMR systems to flag patients for further diagnostic evaluation, streamlining the identification process without requiring additional dedicated tests.
Regulatory status is crucial; several AI-ECG algorithms for low ejection fraction have received FDA clearance, often as a Class II medical device or Software as a Medical Device (SaMD). This approval signifies its validation for clinical decision support, indicating it can aid physicians but not replace their independent judgment.
While not a replacement for echocardiography, AI-ECG algorithms serve as highly sensitive screening tools, demonstrating good accuracy for identifying patients at risk, especially in asymptomatic populations. Their primary value lies in efficiently triaging patients who warrant a definitive echocardiogram, rather than providing a definitive EF measurement itself.
Pricing models vary, often including per-test fees, subscription services, or site licenses. Reimbursement for AI-driven diagnostic tools is evolving, with some AI-ECG technologies, including those for low ejection fraction, securing Category III CPT codes and CMS reimbursement starting in 2025.
Performance can be affected by data quality and specific patient conditions; for instance, pacemakers, severe arrhythmias, or certain congenital heart diseases may introduce noise or alter ECG patterns, potentially reducing accuracy. It's important to understand the algorithm's validated populations and use it judiciously in complex cases, as models are typically trained on specific datasets.
When the algorithm flags a patient, the recommended next step is typically to confirm the finding with a definitive diagnostic test, most commonly an echocardiogram. This allows for precise measurement of ejection fraction and comprehensive assessment of cardiac structure and function.
Understanding the false positive and false negative rates is critical for clinical utility. For example, one AI-ECG algorithm for LVEF <40% showed a sensitivity of 74.7% and specificity of 77.5%. A higher false positive rate might lead to unnecessary echocardiograms, while a higher false negative rate could delay diagnosis and treatment, impacting patient outcomes and healthcare costs.

Related Tools

4DMedical (Imbio)
4DMedical
Cardiology AI
4DMedical, with its acquisition of Imbio, offers a comprehensive suite of AI-driven medical imaging solutions for chronic lung and cardiothoracic diseases, providing quantitative and personalized insights from X-ray and CT scans.
Fetal EchoScan (v1.2)
BrightHeart (via AI Agent)
Cardiology AI
Fetal EchoScan (v1.2) is an AI-powered, FDA-cleared software designed to assist physicians in interpreting fetal heart ultrasound examinations for potential congenital heart defects.
Assert-IQ (DM5100)
Abbott (Assert-IQ ICM) / AI Agent (General AI Agent concept)
Cardiology AI
The Assert-IQ™ Insertable Cardiac Monitor (ICM) is a Bluetooth-enabled device from Abbott that provides continuous, real-time heart monitoring and arrhythmia detection, utilizing AI algorithms to reduce false positives and offer long-term diagnostic evaluation for patients with cardiac conditions.
QardioMD
Qardio
Cardiology AI
QardioMD offers an end-to-end virtual care platform with clinical-grade smart devices for remote patient monitoring of various chronic conditions, including hypertension and cardiovascular disease, and integrates with EHR systems.
Contact for pricing
PrepMD OMNI
PrepMD
Cardiology AI
PrepMD OMNI is an educational platform for cardiac rhythm management technology teams, focusing on training clinicians, hospitals, and practices in CRM device management.
Flexible pricing with no upfront costs for many services.
Rapid Enterprise Platform
RapidAI
Cardiology AI
RapidAI's Enterprise Platform offers FDA-cleared, deep clinical AI solutions that connect imaging to clinical action across the health system, integrating with EHRs and PACS for enhanced decision-making and workflow efficiency.

See all Cardiology AI tools →

More from Anumana

ECG-AI Low Ejection Fraction (LEF) 12-Lead algorithm (1010)
Anumana
Cardiology AI
Anumana's FDA-cleared ECG-AI LEF 12-Lead algorithm uses artificial intelligence to detect low ejection fraction from routine 12-lead ECGs, aiding in the early identification of heart failure risk.
Contact for pricing

View company profile →

Suggest an Edit → | Last Verified: 2026-04-17 | First Added: 2026-04-17
AI Tool Finder
AI-powered search. Results may not be comprehensive.