McLaren Health Care has launched an artificial intelligence powered cardiovascular screening program using the Carebricks platform from Bunkerhill Health to analyse routine chest computed tomography scans for incidental coronary artery calcium and aortic valve calcium. The program, deployed through the McLaren Heart and Vascular Institute, uses FDA-cleared algorithms to identify early cardiovascular risk from scans originally ordered for non-cardiac indications, positioning the health system among a small group of early adopters nationally.

How incidental imaging is shifting cardiovascular risk detection beyond cardiology-led pathways

The most consequential aspect of McLaren Health Care’s initiative is not the use of artificial intelligence itself, but the deliberate move to treat routine chest imaging as a scalable cardiovascular screening surface. Chest computed tomography scans are already among the most frequently ordered imaging modalities in U.S. healthcare, driven by lung cancer screening, infection evaluation, trauma assessment, and post-surgical follow-up. Historically, cardiovascular findings visible on these scans were often incidental, inconsistently reported, or clinically deprioritised unless symptoms were present.

Industry observers note that by operationalising incidental coronary artery calcium and aortic valve calcium detection across an entire health system, McLaren is effectively reframing how cardiovascular risk is surfaced. Instead of relying on dedicated cardiac imaging referrals, which remain constrained by access, cost, and clinician time, the program embeds risk detection directly into existing diagnostic workflows. This represents a structural shift from episodic, referral-based screening toward ambient, data-driven identification of disease progression signals.

Why coronary artery calcium and aortic valve calcium matter clinically

Coronary artery calcium scoring has long been recognised as a strong predictor of coronary heart disease risk, particularly in asymptomatic populations. Similarly, aortic valve calcium is increasingly used as a surrogate marker for the presence and progression of aortic stenosis, a condition that often remains clinically silent until advanced stages. What has limited broader adoption is not clinical relevance but operational friction, including the need for dedicated scans, specialist interpretation, and follow-up coordination.

Clinicians tracking the field believe that applying validated algorithms to existing chest scans removes several of these bottlenecks simultaneously. Detection becomes opportunistic rather than elective, interpretation becomes standardised rather than variable, and follow-up can be protocol-driven rather than dependent on individual clinician vigilance. This approach aligns with broader population health strategies that prioritise early intervention before symptom onset, particularly for diseases with long subclinical phases.

What is genuinely new versus incremental in this deployment

Artificial intelligence assisted image analysis is not new in cardiovascular imaging. What differentiates this program is the combination of FDA-cleared detection algorithms with workflow-level automation that extends beyond image interpretation. Bunkerhill Health’s Carebricks platform is positioned as a system of action rather than a standalone algorithm, enabling automated chart review, guideline alignment, and follow-up eligibility determination.

Regulatory watchers suggest that this distinction matters because many artificial intelligence imaging tools fail to translate into measurable clinical impact due to downstream execution gaps. Detection alone does not change outcomes if findings are not communicated, contextualised, and acted upon. By embedding algorithmic outputs into care pathways that trigger follow-up recommendations, McLaren is testing whether artificial intelligence can meaningfully reduce the operational burden that has historically limited preventive cardiology programs.

Comparison with existing cardiovascular screening models at major health systems

Dedicated cardiovascular screening initiatives at institutions such as Mayo Clinic and Cleveland Clinic have demonstrated clinical value but remain resource-intensive. These programs typically rely on specialised imaging, cardiology-led interpretation, and manual coordination of follow-up care. As a result, scalability across broader populations has been limited.

The McLaren Health Care model differs by decentralising detection and centralising action. Routine chest scans performed across multiple facilities become the raw input, while artificial intelligence standardises detection and triage. Industry analysts view this as a potential inflection point where cardiovascular screening shifts from centre-of-excellence models to system-wide population health infrastructure. If successful, the approach could lower per-patient screening costs and reduce disparities driven by referral access and specialist availability.

Regulatory clarity and remaining uncertainties around algorithm-driven screening

The use of FDA-cleared algorithms for incidental coronary artery calcium and aortic valve calcium detection provides a measure of regulatory confidence, particularly in an environment where artificial intelligence oversight remains fragmented. However, regulatory observers caution that clearance for detection does not equate to endorsement of automated care decisions. Health systems remain responsible for how findings are communicated to patients, how follow-up thresholds are defined, and how false positives or incidental findings are managed.

There is also ongoing debate around liability allocation when artificial intelligence flags findings on scans not originally intended for cardiovascular evaluation. While incidental findings are already a recognised medico-legal domain, scaling detection through automation increases volume and visibility, potentially amplifying both benefit and risk.

Adoption, reimbursement, and clinician workflow implications

One of the most immediate challenges for artificial intelligence driven screening programs is clinician acceptance. Cardiologists and primary care physicians alike must trust algorithmic outputs and feel confident that follow-up recommendations are clinically appropriate. Bunkerhill Health’s emphasis on workflow support rather than replacement appears designed to address this concern by positioning artificial intelligence as an augmentation layer.

Reimbursement remains a more complex question. While downstream diagnostic testing and interventions are reimbursable, the screening activity itself may not map cleanly to existing billing codes. Health economists suggest that near-term value justification may rely more on avoided acute events and improved outcomes than on direct reimbursement. This places pressure on health systems to measure and articulate return on investment through reduced hospitalisations and improved longitudinal care metrics.

Manufacturing and scalability considerations for AI-enabled screening platforms

From a technology perspective, scalability depends less on algorithm performance and more on integration resilience. Health systems operate heterogeneous imaging infrastructure, electronic health records, and governance processes. Platforms like Carebricks must adapt to local protocols while maintaining consistent performance across sites. Industry observers note that McLaren’s integrated system structure may provide a favourable test environment compared with more fragmented networks.

Data governance and patient consent also emerge as critical factors. Using retrospective scans from the past 12 months raises questions about notification thresholds and patient communication strategies. Transparency around how findings are generated and used will likely influence patient trust and program sustainability.

What clinicians, regulators, and industry observers will watch next

The most closely watched outcome will be whether early detection translates into earlier intervention and measurable outcome improvement. Metrics such as time to follow-up, initiation of preventive therapy, and progression rates of aortic stenosis will determine whether the program delivers on its promise. Clinicians will also scrutinise false positive rates and the burden of additional testing triggered by algorithmic findings.

Regulators are likely to monitor how incidental detection programs are governed as they proliferate. Clear standards around validation, auditability, and clinician oversight may emerge as adoption increases. For the broader industry, McLaren Health Care’s initiative serves as a case study in whether artificial intelligence can finally bridge the gap between detection and action in preventive cardiology.

Strategic implications for the broader healthcare artificial intelligence market

If successful, this deployment strengthens the argument that the next wave of healthcare artificial intelligence value will come from workflow orchestration rather than standalone diagnostics. Vendors that can demonstrate end-to-end impact, from detection to clinical action, may gain an advantage as health systems become more selective about technology investments.

The program also highlights a broader shift toward leveraging existing data assets more aggressively. Rather than generating new data through additional tests, health systems are increasingly looking to extract latent value from imaging and records already in place. This approach aligns with cost containment pressures and population health mandates, positioning artificial intelligence as an efficiency lever rather than an added expense.

  aortic valve calcium, artificial intelligence in healthcare, Bunkerhill Health, cardiovascular screening, Carebricks platform, coronary artery calcium, McLaren Health Care, medical imaging, preventive cardiology