14492
clinical study

From Pixels to Prevention: AI-Driven Coronary Calcium Detection and Patient Management


Materials & Methods

A two-part AI system integrating automated image-based coronary artery calcium (CAC) scoring with natural language processing (NLP)–derived report mining was deployed to identify and manage incidental CAC on routine non-gated, non-cardiac CT studies (May 2024–April 2025). Outpatient studies in patients under 80 years with AI-assessed CAC ≥ moderate were included, while those with existing cardiology care or limited life expectancy were excluded. Positive NLP detections were routed to a navigator-driven management platform for cardiology referral and follow-up. Outcomes included identification rates, referral completion, consultations, ancillary testing and treatment changes.

Results

Among 1,756 outpatient CT scans analyzed by the image-based AI tool, the NLP system identified 951 patients with CAC findings; 266 (27.9%) were excluded because of existing cardiology care. Of the 69 referred patients, 61 (88.4%) completed cardiology visits. Medication adjustments occurred in 68.8% (42/61) of patients while an equal proportion underwent ancillary cardiac testing (stress tests, echocardiography). Six patients (14.3%) proceeded to left heart catheterization, with two requiring revascularization (PCI or CABG).

Conclusions

AI-enabled CAC detection and structured management is feasible and clinically impactful. This workflow facilitates opportunistic cardiovascular disease prevention, identifies care gaps and drives meaningful changes in patient management — ranging from medication optimization to invasive evaluation. Early evidence supports scalable implementation across outpatient imaging programs.

Disclaimer: This links to a third-party website, which may require credentials to review.

Explore the Latest AI Insights, Trends and Research