AI ECG Tool Detects Cardiac Amyloidosis for Early Screening

Studies estimate that roughly 13–15% of patients in heart failure clinics may have undiagnosed disease, and diagnosis has historically relied on late-stage, resource-intensive imaging. Earlier detection can redirect patients away from ineffective or harmful therapies and toward options proven to improve survival. A newly patented artificial intelligence algorithm now analyzes standard 12‑lead electrocardiograms to flag cardiac amyloidosis as a potential screening signal.

AccurKardia, Inc. has been granted US patent vering a proprietary machine‑learning‑based system that identifies cardiac amyloidosis from a routinely performed 12‑lead electrocardiogram (ECG). The intellectual property spans major amyloidosis subtypes, including light chain (AL) amyloidosis and wild‑type and hereditary transthyretin (ATTR) amyloidosis. Positioned within a broader ECG biomarker portfolio, the patent underscores efforts to expand the role of routine ECG from rhythm analysis toward disease screening.

The system is described as using explainable, feature‑based machine learning built on annotated ECG parameters rather than a “black box” model. This design is intended to support interpretability for clinicians, transparency for regulators, and adaptability for downstream deployment at scale. By grounding classification in physiologically meaningful ECG features, the approach is framed to align with existing clinical workflows.

The cardiac amyloidosis detection algorithm is currently for research use only and has not been cleared by the U.S. Food and Drug Administration (FDA) for use in the United States. AccurKardia’s AccurECG 2.0 platform is FDA‑cleared as a Class II software as a medical device for fully automated, near real‑time ECG interpretation (510(k) K252361). Additional investigational ECG‑based biomarkers in the company’s pipeline include AK‑AVS for aortic stenosis risk assessment and AK+ Guard for hyperkalemia risk assessment, both of which have received FDA Breakthrough Device Designation and are for research use only.

“Cardiac amyloidosis hides in plain sight, and because the symptoms are quite similar to other causes of heart failure, we have historically relied on expensive, late-stage imaging to diagnose what is already advanced disease, often after conventional medical therapy has failed to improve symptoms,” said Dr. Jason Lazar, executive vice dean, chair of the Department of Medical Education and Director of Non-invasive Cardiology at SUNY Downstate. “A reliable ECG-based screening signal, leveraging information the human eye simply cannot extract, has the potential to redefine when and how we intervene, particularly as therapeutic options continue to expand. Simply put, earlier diagnosis leads to much better outcomes.”

“Disease-modifying amyloidosis therapies are among the most important advances in cardiology in a generation, but their impact is gated by our ability to find the right patients in time,” said Juan C. Jimenez, co-Founder and CEO of AccurKardia. “This patent establishes the foundation for closing that gap. We are turning a test that is already performed millions of times per year globally into a screening biomarker, deployable without new hardware or procedures, seamlessly integrating our capabilities into existing workflows.”

Related Links
AccurKardia