Tuesday, June 30, 2026

A Sweeping New AI to Detect Heart Conditions Is Coming to OpenEvidence

A Sweeping New AI to Detect Heart Conditions Is Coming to OpenEvidence

Pathway Labs' FDA-Cleared EchoNext AI Spots Structural Heart Disease in Electrocardiograms

A new artificial-intelligence tool can now read a routine 12-lead electrocardiogram and flag structural heart disease that the human eye typically misses.

The tool, called EchoNext, was built by Pathway Labs using ECG-echocardiogram pairs originally developed at NewYork-Presbyterian and Columbia University.

It just received the world's first FDA clearance for multicondition AI detection in cardiology, covering six structural heart conditions.

Perhaps more importantly for everyday practice, EchoNext is now being deployed inside OpenEvidence, the clinical decision-support platform already used by roughly half of practicing U.S. physicians.

That distribution strategy, rather than the algorithm itself, may be what changes daily practice fastest.

What EchoNext Actually Detects

The model screens for six categories of disease that ordinary ECG interpretation cannot reliably catch: right-sided heart failure, left-sided heart failure, valve disease, severe hypertrophy consistent with infiltrative cardiomyopathy, pulmonary hypertension, and broader structural or congenital abnormalities.

All of these conditions share a common problem: they are common, often silent, and usually diagnosed only after a patient becomes symptomatic or develops overt heart failure.

The standard 12-lead ECG is inexpensive, fast, and already ordered millions of times per year, which is exactly why a screening layer built on top of it is attractive.

Unlike echocardiography, the ECG requires no sonographer, no scheduling delay, and no incremental imaging cost.

Figure 1. EchoNext vs. Cardiologist Detection of Structural Heart Disease
EchoNext AI
77%
Cardiologists (unassisted)
64%

Detection rate of structural heart disease in a head-to-head validation study

In a head-to-head validation, EchoNext identified 77% of structural heart problems, compared with 64% for cardiologists reading the same tracings unaided.

The model was trained on more than 700,000 paired ECGs and echocardiograms, then externally validated across more than 20 hospitals and 500,000 patients.

A peer-reviewed report in Nature Medicine described the first documented case in which EchoNext flagged previously undiagnosed heart failure that ultimately led to a heart transplant.

Why Distribution Through OpenEvidence Matters

An FDA clearance that sits unused inside a single health system changes very little for the average outpatient cardiologist or internist.

By embedding EchoNext directly into OpenEvidence, Pathway Labs is effectively skipping years of piecemeal hospital-by-hospital implementation.

Workflow-wise, a clinician will be able to upload an image of an ECG into the platform and receive an algorithmic structural-disease risk prediction in return.

That puts a screening capability previously confined to research-grade academic centers into community clinics, urgent cares, and small cardiology practices.

Figure 2. From Routine ECG to Actionable Screening Signal
Routine 12-lead ECG obtained
Uploaded into OpenEvidence
EchoNext risk prediction returned
Echocardiogram or referral if flagged

The Six Cleared Indications at a Glance

Condition FlaggedClinical Relevance
Left-sided heart failureOften subclinical until exertional symptoms develop
Right-sided heart failureFrequently under-recognized on routine exam
Valve diseaseMay precede audible murmur or symptom onset
Severe hypertrophy / infiltrative cardiomyopathy patternRaises suspicion for conditions such as amyloidosis or hypertrophic cardiomyopathy
Pulmonary hypertensionCommonly diagnosed late, after right-heart strain develops
Other structural / congenital abnormalityCaptures residual structural disease outside the above categories

Company and Platform Snapshot

EntityRoleOwnership Status
Pathway LabsDeveloper of EchoNext algorithmPrivately held; no public ticker
OpenEvidenceClinical decision-support distribution platformPrivately held; no public ticker

Pathway Labs also disclosed an $8.5 million seed financing round led by AlleyCorp and Breyer Capital to fund health-system expansion, clinical team growth, and continued research.

Neither company is publicly traded, so there are no equity tickers or per-test consumer pricing to report at this time.

For physician-investors tracking the AI-in-cardiology space more broadly, exposure currently runs through diversified healthcare-technology vehicles rather than single-name positions in this particular tool.

Practical Considerations Before Adoption

EchoNext is a screening aid, not a diagnostic replacement for echocardiography.

A flagged result should prompt confirmatory imaging rather than a standalone clinical decision.

Reimbursement and CPT coding pathways for AI-augmented ECG interpretation remain unsettled and should be clarified with billing compliance before routine use.

Documentation should clearly distinguish an AI-generated screening flag from a formal cardiologist interpretation to avoid medicolegal ambiguity.

Case Vignette

A 58-year-old patient presents to a primary care clinic for an annual physical with no cardiac complaints.

A routine ECG is obtained and uploaded into the practice's clinical decision-support platform.

The AI screening tool flags a pattern consistent with severe hypertrophy and possible infiltrative cardiomyopathy.

The patient is referred for echocardiography and subsequent cardiac MRI, which confirm early transthyretin amyloid cardiomyopathy at a stage when targeted therapy can still meaningfully alter the disease course.

Bottom Line

EchoNext brings FDA-cleared, AI-driven structural heart disease screening to the most commonly ordered cardiac test in medicine.

Its integration into a decision-support platform used by roughly half of U.S. physicians could extend early detection capability well beyond academic cardiology centers.

Confirmatory imaging, clear documentation, and attention to billing pathways will remain essential as adoption spreads.

This article is intended for physician education and general clinical awareness only. It does not constitute medical, legal, billing, or investment advice, and should not substitute for independent clinical judgment, institutional policy review, or consultation with appropriate compliance, legal, or financial professionals.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.