FDA clears an AI tool that flags pulmonary hypertension risk from a standard ECG
The clearance is the third in a growing family of machine-learning notification algorithms built to spot serious cardiac conditions from a routine 12-lead ECG, a test most patients already get.
The FDA cleared a new machine-learning tool on March 28 designed to flag patients who may have pulmonary hypertension based on nothing more than a standard 12-lead ECG — the same routine heart-rhythm test used for a wide range of cardiac evaluations [s1].
What was cleared
The device, called the ECG-AI Pulmonary Hypertension (PH) 12-Lead algorithm, was cleared through the FDA's 510(k) pathway under the classification "Pulmonary Hypertension Machine Learning-Based Notification Software" [s1]. It was reviewed by the FDA's cardiovascular advisory committee and classified as a Class II device, cleared on the basis that it is substantially equivalent to an existing predicate device rather than requiring the more rigorous premarket approval pathway [s1]. The manufacturer, Cambridge, Massachusetts-based Anumana, Inc., submitted the application on July 29, 2025, and received the clearance decision eight months later [s1].
As "notification software," the tool is designed to analyze the electrical signal of a standard ECG and flag a pattern associated with elevated pulmonary hypertension risk for a clinician's attention — not to independently diagnose the condition. Pulmonary hypertension, elevated blood pressure in the arteries supplying the lungs, is notoriously difficult to detect early because its symptoms — shortness of breath, fatigue — overlap with far more common conditions, and confirming it typically requires specialized testing like echocardiography or right-heart catheterization that isn't part of routine primary care.
Part of a broader pattern from one manufacturer
This is not Anumana's first ECG-based AI clearance. FDA records show the company first cleared a similar tool for detecting reduced ejection fraction — a marker of weakened heart-pumping function — in September 2023, then received a follow-on clearance for an updated version of that same algorithm in July 2025 [s2]. The March 2026 pulmonary hypertension clearance is the company's third distinct ECG-AI notification algorithm, following the same basic model: train a machine-learning system to detect a specific cardiac condition's signature in a standard 12-lead ECG waveform, then seek clearance for it as a notification tool that flags results for physician review [s2].
The pattern reflects a broader shift in cardiology-focused AI: rather than requiring new equipment, these algorithms are built to extract additional diagnostic signal from a test already embedded in routine care, on the premise that the ECG waveform contains more clinically useful information than is typically extracted by a human reader.
What the clearance does and doesn't establish
A 510(k) clearance is not, by itself, evidence of how well a device performs in everyday clinical use — it establishes that the FDA found the device substantially equivalent to a legally marketed predicate, a lower bar than the clinical trials required for full premarket approval [s1]. The FDA's public 510(k) summary record does not include the algorithm's sensitivity, specificity, or the population it was validated against, so those performance characteristics cannot be verified from what the agency has made public [s1]. Readers should treat the clearance as a regulatory milestone — the tool is now legally marketed in the US — rather than a clinical verdict on its accuracy.
Sources
- 510(k) Premarket Notification K252360 — Pulmonary Hypertension Machine Learning-Based Notification Software — US Food and Drug Administration, 2026-03-28
- 510(k) Premarket Notification database — Anumana, Inc. clearance history — US Food and Drug Administration, 2026-03-28
Sources
- 510(k) Premarket Notification K252360 — Pulmonary Hypertension Machine Learning-Based Notification Software — US Food and Drug Administration , March 28, 2026
- 510(k) Premarket Notification database — Anumana, Inc. clearance history — US Food and Drug Administration , March 28, 2026
The FDA has cleared 1,357 AI medical devices. Three were tested on patient outcomes.
A researcher who expected the evidence base to be thin says even she was surprised by how thin. Most cleared devices never appear in a registered clinical trial at all.
FDA opens comment on how to regulate AI medical devices that generate their own answers
A new discussion paper proposes a two-axis risk framework and a physician-training analogy for evaluating generative AI devices. It is not a rule, and the agency is asking what one should look like.
FDA creates a new device category for AI that reads skin wounds without touching them
The agency granted its first authorization for a software-aided adjunctive diagnostic device in wound assessment, a category built for tools that analyze a wound optically.
In December, Washington asked two opposite questions about medical AI
One agency asked how to measure AI devices after they are deployed. Another asked how to speed adoption up. The offices that would answer the first question are losing staff.