egnite and Pathway Labs launch ECG AI pilot at Stamford Health, but can early detection improve patient care?

egnite, Inc., Pathway Labs and Stamford Health have launched a clinical pilot integrating FDA-cleared electrocardiogram artificial intelligence with cardiovascular care coordination to identify patients who may require further evaluation for structural heart disease. Announced on October 29, 2026, the initiative combines Pathway Labs’ EchoNext screening algorithm with egnite’s CardioCare platform at Stamford Health in Connecticut, extending AI analysis across routine 12-lead ECG recordings and subsequent clinical workflows. The programme will assess whether patients flagged by AI receive appropriate echocardiographic evaluation and specialist follow-up, rather than measuring screening performance in isolation. Edwards Lifesciences Corporation (NYSE: EW) is contributing funding to the pilot as part of a collaboration focused on earlier identification of structural heart disease. The central commercial and clinical question is whether combining automated detection with patient management can improve the progression from a potential cardiac finding to an appropriate treatment decision without creating excessive additional workload.

The collaboration introduces a connected approach to cardiovascular screening in which AI-generated findings are evaluated against a patient’s existing clinical information before care teams determine the appropriate next step. EchoNext examines standard ECG waveforms for patterns associated with several forms of structural heart disease, while CardioCare incorporates prior diagnoses, laboratory results and medical history to help prioritise patients for further assessment. The system is designed to track patients through imaging, specialist evaluation and treatment decisions, creating an opportunity to measure the effectiveness of the entire clinical pathway.

The pilot also marks the first live algorithm partnership within CardioCare’s broader strategy of integrating screening outputs from multiple sources. For Pathway Labs, it represents an opportunity to evaluate EchoNext in operational hospital workflows following US regulatory clearance in June 2026. For egnite, the programme extends an existing cardiovascular patient identification platform upstream into AI-assisted ECG screening.

How does the EchoNext and CardioCare integration change the identification of structural heart disease patients?

Structural heart disease includes abnormalities affecting heart valves, chambers and related cardiac structures. Conditions such as aortic stenosis, significant valve regurgitation and ventricular dysfunction can progress before patients receive specialist evaluation. Electrocardiograms are routinely used to assess electrical activity, rhythm disturbances and other cardiac findings, but they do not directly visualise structural abnormalities in the way echocardiography does.

EchoNext applies machine learning to digital 12-lead ECG waveforms to identify electrical patterns associated with suspected structural heart disease. The algorithm can generate an overall structural heart disease notification and additional indicators relating to specific cardiac abnormalities. Its purpose is to help clinicians identify patients who may benefit from further diagnostic investigation using established clinical methods.

The second stage involves egnite’s CardioCare platform. When EchoNext identifies a potentially relevant ECG finding, CardioCare evaluates that signal alongside longitudinal information already available in the patient’s medical record. This can include previous cardiac diagnoses, laboratory information and relevant clinical history, allowing the system to help prioritise patients for echocardiography rather than treating every AI notification as an identical case.

The combined process is intended to support clinical teams in distinguishing patients who may need additional evaluation from those whose conditions are already recognised or under appropriate management. The platform then tracks progress through imaging and specialist review, making it possible to examine whether an initial screening alert results in a completed clinical assessment.

This integration addresses an operational limitation of standalone screening algorithms. Identifying a potential abnormality can increase the number of patients requiring review, but it does not ensure that diagnostic appointments are scheduled, examinations are completed or treatment decisions are made. The pilot will evaluate whether connecting detection with established care management processes can improve that continuity.

What does the FDA clearance of Pathway Labs’ EchoNext algorithm establish about clinical performance and limitations?

EchoNext received 510(k) clearance from the US Food and Drug Administration on June 8, 2026, under submission K260332. The regulator classified the software as a Class II cardiovascular machine learning-based notification device. The clearance permits its use as an adjunct to standard clinical care for analysing compatible 12-lead ECG recordings and identifying patients who may warrant further evaluation for structural heart disease.

The authorised indications cover ECG recordings from adults aged 18 years and older. EchoNext is not intended for patients with implanted pacemakers and must not be used independently to establish or confirm a diagnosis. A negative result also cannot be used to rule out the need for further assessment when other clinical evidence indicates possible disease.

The regulatory submission identifies six additional categories of possible cardiac abnormalities that the algorithm can flag. These include reduced left ventricular ejection fraction, right ventricular systolic dysfunction, severe left ventricular hypertrophy, moderate or greater aortic stenosis, significant valvular regurgitation and pulmonary hypertension. These are screening indicators rather than confirmed diagnoses.

The FDA submission includes external clinical validation involving 102,219 patients across four healthcare systems in the United States and Canada. The evaluation compared ECG-based predictions with findings from echocardiographic assessments and reported sensitivity of 73.7% and specificity of 75.2% for the composite structural heart disease output.

Those figures indicate that the algorithm detected approximately 74% of patients with qualifying structural heart disease within the validation cohort, while correctly identifying approximately 75% of patients without the qualifying disease. The evaluation reported a positive predictive value of 61.6% in a population where structural heart disease prevalence was 34.8%.

Performance differed by condition. Sensitivity was 80.1% for left ventricular systolic dysfunction but 54.9% for moderate or greater aortic stenosis and 48.2% for severe left ventricular hypertrophy. Such variation is clinically relevant because an algorithm capable of identifying several conditions does not necessarily provide equivalent screening performance across each disease category.

The FDA clearance establishes that EchoNext met the regulatory requirements applicable to its authorised use. It does not establish that deploying the algorithm across routine hospital ECG volumes will improve patient survival, reduce hospitalisations or increase successful treatment rates. Those downstream outcomes require additional clinical evidence.

Why does Stamford Health’s experience with nearly 300 previously overlooked cardiac patients strengthen the pilot rationale?

Stamford Health provides an established clinical setting for assessing the combined technology because it has used egnite’s CardioCare platform since 2022. Arzhang Fallahi, director of structural heart at Stamford Health, reported that the organisation had identified and evaluated nearly 300 patients with significant heart disease who otherwise were not on a pathway towards treatment.

That figure relates to Stamford Health’s earlier use of CardioCare and should not be attributed to the newly launched EchoNext integration. It nevertheless provides evidence that the organisation has experience using digital systems to find patients requiring clinical attention. The new ECG capability could expand the range of information available for identifying additional patients.

Stamford Health conducts tens of thousands of ECG examinations annually, according to its clinical leadership. Applying EchoNext to eligible recordings creates an opportunity to identify possible structural abnormalities among patients who underwent ECG testing for various clinical reasons but have not necessarily received echocardiography.

The potential screening population is substantially larger than the group of patients already known to have clinically significant structural heart disease. However, the number of additional patients detected will depend on disease prevalence, ECG eligibility, algorithm performance and the extent to which existing clinical pathways have already identified relevant conditions.

David Hsi, chief of cardiology and co-director of the Heart & Vascular Institute at Stamford Health, indicated that the algorithm would expand the hospital’s existing AI activities in cardiac imaging. The deployment will consequently assess how ECG-based signals can complement established cardiovascular diagnostic and management programmes rather than operate as a separate screening service.

The hospital’s prior experience provides an operational foundation, but it does not guarantee comparable results from the new approach. The pilot must demonstrate whether AI-generated alerts contribute information that changes clinical management beyond what existing workflows would have achieved.

Can ECG-based artificial intelligence improve patient follow-up without overwhelming cardiology services?

The operational consequences of ECG-based screening extend beyond algorithm sensitivity and specificity. A healthcare system processing large ECG volumes may generate substantial numbers of alerts, including findings that do not result in confirmed structural heart disease. Each notification can require evaluation against existing records, clinician judgement and potentially additional diagnostic testing.

The positive predictive value observed during EchoNext’s regulatory validation reflects the disease prevalence and patient selection within that study. Because the validation population comprised patients who received both ECG and echocardiographic examinations within a defined period, its results cannot automatically be extrapolated to a broader population undergoing routine ECG testing.

In populations where qualifying disease is less prevalent, a larger proportion of positive screening alerts may not correspond to confirmed disease, depending on actual algorithm performance. The FDA submission illustrates this issue through condition-specific analyses. For moderate or greater aortic stenosis, the algorithm’s reported positive predictive value was 11.8% in the external validation cohort, highlighting the need for careful clinical interpretation of individual condition flags.

CardioCare is intended to introduce another layer of prioritisation by placing AI findings within a patient’s clinical context. This may help care teams avoid treating every abnormal signal as an independent reason for immediate imaging, although the effectiveness of that approach has not yet been established in the new pilot.

The programme will measure echocardiogram completion rates and the time between identification and specialist evaluation. These metrics will help determine whether the integrated process produces actionable patient referrals rather than simply increasing the volume of potential abnormalities entering the healthcare system.

Additional operational measures would strengthen the assessment, including the proportion of alerts requiring manual review, the number of new clinically meaningful diagnoses and the effect on imaging capacity. A system that identifies more potential cases but creates significant delays for existing patients may provide less benefit than one that accurately prioritises those most likely to need further care.

The financial implications are similarly dependent on workflow performance. Hospitals may face additional expenditure associated with software deployment, clinician review, diagnostic testing and care coordination. Improvements in early identification could support more timely treatment, but the overall economic effect would require evidence of appropriate clinical use and measurable reductions in avoidable downstream costs.

How could egnite’s network of more than 80 heart programmes support wider commercial adoption of EchoNext?

egnite’s CardioCare platform is currently deployed across more than 80 cardiovascular programmes and draws on information from a broader network of more than 1,200 cardiovascular care facilities. The company’s registry contains de-identified electronic health record data relating to approximately 15.7 million patients and 6 million echocardiography reports.

These figures describe egnite’s existing technology footprint and data resources, not the number of hospitals deploying EchoNext. Stamford Health is the first announced live partner for the new algorithm integration, and the October announcement does not identify confirmed deployment schedules for additional participating hospitals.

The existing network nevertheless gives the collaboration a potential route to broader adoption if the pilot demonstrates clinical and operational benefits. Hospitals already using CardioCare may be able to incorporate ECG-based screening into familiar patient management workflows, potentially reducing some integration barriers associated with standalone software.

For Pathway Labs, this distribution arrangement could accelerate the transition from a cleared screening product to routine clinical use. The technology originated from research involving Columbia University Irving Medical Center and NewYork-Presbyterian, with EchoNext’s underlying scientific approach described in a 2025 study published in Nature. Its development programme provides research and regulatory evidence, while the current partnership tests whether the software can deliver practical benefits across everyday healthcare operations.

For egnite, supporting third-party screening algorithms could broaden the relevance of CardioCare beyond its existing patient identification functions. Integrating additional data sources may strengthen its role as a clinical workflow platform, provided customers see measurable improvements in patient prioritisation and care coordination.

Neither company disclosed the commercial terms of its collaboration, the pricing of the combined offering or the expected revenue associated with deployment. The market opportunity therefore remains difficult to quantify despite the established size of egnite’s customer network.

Why is Edwards Lifesciences Corporation funding the pilot, and what could it mean for structural heart treatment markets?

Edwards Lifesciences Corporation is providing funding to support the programme as part of a collaboration with egnite and Pathway Labs focused on earlier identification of structural heart disease. The announcement does not disclose the funding amount, equity arrangements or any commitment to use particular Edwards Lifesciences products in subsequent patient treatment.

The involvement is strategically relevant because Edwards Lifesciences develops technologies for structural heart conditions, including transcatheter aortic valve replacement and mitral and tricuspid valve treatment. Earlier recognition of clinically significant disease could potentially increase the number of patients receiving timely specialist evaluation and being considered for appropriate treatment.

However, an AI-generated screening alert does not automatically create demand for a structural heart intervention. Patients must undergo diagnostic evaluation, establish the severity and clinical significance of their condition, and meet the relevant criteria for a specific treatment. Some may require monitoring or medical management rather than an interventional procedure.

The collaboration consequently sits upstream of Edwards Lifesciences’ established therapeutic markets. Its strategic value would arise if improved screening and care coordination help address patients who currently experience delays between disease recognition and appropriate evaluation. The pilot does not provide enough evidence to estimate additional treatment procedures or product revenue.

Edwards Lifesciences reported second-quarter 2026 revenue of $1.74 billion, an increase of 13.6% year on year. Transcatheter aortic valve replacement sales reached approximately $1.26 billion, while transcatheter mitral and tricuspid therapies contributed $195.9 million. These figures demonstrate the scale of the company’s established structural heart businesses relative to the early-stage digital screening initiative.

Edwards Lifesciences shares closed at $91.78 on October 28, 2026, a gain of approximately 1.1% for that session. A delayed market quotation at 10:34 a.m. Eastern Time on October 29 placed the shares near $94.11, approximately 2.5% above the previous close. No evidence establishes that this movement was caused by the ECG AI pilot, and the company has not disclosed a financial impact from its funding contribution.

The next major scheduled financial catalyst is Edwards Lifesciences’ third-quarter earnings release on November 4, 2026. The company’s structural heart sales performance, procedural growth and management guidance provide a more substantial basis for assessing near-term investor sentiment than the currently unquantified pilot investment.

What will the Stamford Health pilot need to demonstrate before ECG AI screening can scale across cardiovascular programmes?

The pilot’s primary value will come from measuring what occurs after an AI screening alert enters the clinical workflow. EchoNext has already received regulatory clearance, and its performance has been evaluated through substantial external clinical validation. The current initiative addresses a different question: whether integrating those findings with established patient management systems changes real-world care delivery.

Echocardiogram completion rates and time to specialist evaluation are two explicitly identified study measures. Stronger results would establish that an increased proportion of appropriately selected patients complete recommended testing and receive clinical assessment within shorter periods. These outcomes would provide more direct evidence of workflow effectiveness than the number of ECGs analysed or alerts produced.

The programme will also need to distinguish newly identified cases from patients whose conditions were already recognised. This would help determine whether the algorithm expands the number of patients receiving appropriate attention or predominantly reproduces information available through existing records. Data on alert burden, false-positive follow-up and diagnostic yield would clarify the operational cost of achieving any improvements.

Longer-term outcomes would require additional measurement. Earlier identification and completed imaging are intermediate achievements, while the ultimate clinical objective involves appropriate treatment and improved patient outcomes. A successful screening pathway should demonstrate that patients benefit from the information generated, not simply that additional tests are performed.

Rahul Sharma, senior medical adviser to egnite and a physician at Stanford Health Care, described the pilot as an opportunity to assess the full journey from detection through clinical follow-up and treatment. Pierre Elias, chief executive officer of Pathway Labs, likewise positioned the collaboration as a step towards evaluating EchoNext in real-world healthcare settings following its research and regulatory development.

The announcement does not provide a pilot completion date, planned patient enrolment, follow-up duration or predefined clinical success thresholds. These details will determine how comprehensively the results can establish the integration’s value and whether findings are suitable for comparison with other care settings.

egnite and Pathway Labs are testing a commercially relevant proposition: combining a specialised, FDA-cleared screening algorithm with a platform already used to coordinate cardiovascular care. Their approach could address the separation between identifying a possible cardiac abnormality and ensuring that a patient receives appropriate follow-up. The decisive evidence will be whether Stamford Health demonstrates higher rates of completed, clinically justified evaluations, shorter time to specialist review and improved treatment decisions without disproportionate increases in diagnostic workload.

What are the key takeaways from the egnite, Pathway Labs and Stamford Health ECG AI pilot?

• egnite, Pathway Labs and Stamford Health announced a pilot on October 29, 2026, integrating ECG-based AI screening with cardiovascular care coordination.
• The programme combines Pathway Labs’ EchoNext screening algorithm with egnite’s CardioCare platform to identify patients who may require structural heart disease evaluation.
• EchoNext received FDA 510(k) clearance in June 2026 for use alongside standard clinical care, not as an independent diagnostic tool.
• External validation involving 102,219 patients reported 73.7% sensitivity and 75.2% specificity for the algorithm’s overall structural heart disease output.
• Stamford Health has previously identified and evaluated nearly 300 significant cardiac cases using CardioCare since 2022, before the new EchoNext integration.
• The pilot will assess whether AI-generated alerts lead to completed echocardiograms, timely specialist reviews and appropriate care decisions.
• egnite’s existing network of more than 80 heart programmes offers a potential commercial expansion route if clinical benefits are demonstrated.
• Edwards Lifesciences Corporation is funding the pilot, although no investment amount or direct revenue contribution has been disclosed.
• Screening accuracy, false-positive workload and the clinical usefulness of new referrals will influence adoption across healthcare systems.
• The most meaningful future evidence will be measurable improvements across the entire pathway from disease identification to appropriate clinical management.