From Research to Routine: The Billion Dollar Rise of Cardiac CT AI

Getting a new diagnostic technology from a research lab into routine clinical use is a long, hard slog. It’s a path littered with regulatory traps and financial dead ends. For coronary CT analysis, especially the AI-driven functional tools, the journey was plotted through key regulatory wins, coverage decisions, and a mountain of long-term evidence. This is the story of how coronary CT analysis went from a niche tool for researchers to a standard of care for cardiology program leads making daily referral decisions.

The Genesis of Non-Invasive FFRct Analysis: De Novo Clearance

The first real breakthrough for advanced coronary CT analysis was the De Novo clearance for non-invasive fractional flow reserve CT (FFRct). HeartFlow got this clearance, and it showed the FDA recognized a completely new way to assess coronary artery disease that went beyond just looking at anatomical stenosis. The De Novo pathway is made for these kinds of low-to-moderate-risk devices that don’t have a predecessor, introducing a new technology or a new use. For HeartFlow, this wasn’t just checking a regulatory box. It created an entirely new classification for SaMD (Software as a Medical Device) that could extract functional data from a standard coronary CT angiography (CCTA) scan. This was a world away from traditional invasive FFR. Suddenly, cardiologists had a tool to see the hemodynamic significance of a lesion without having to run a catheter up to the heart, which immediately changed how patients were stratified and who got referred for invasive procedures.

Anchoring Routine Referral: CMS Local Coverage and UK Health Technology Appraisal

FDA clearance is one thing, but getting paid is what drives adoption in the real world. HeartFlow’s journey hit its next major milestone with a series of CMS Local Coverage Determinations (LCDs) for FFRct analysis. These LCDs gave CT-derived FFR a path to national reimbursement, which was the green light healthcare systems across the US needed to integrate it into standard practice without taking a financial hit. For the cardiology programs themselves, the LCDs provided the financial security to invest in the software and build it into their diagnostic algorithms for stable ischemic heart disease. At the same time, the technology was getting a nod internationally. A United Kingdom health technology appraisal, HTG429, gave a strong, independent thumbs-up to the clinical and cost-effectiveness of FFRct. Appraisals like this are huge for adoption in nationalized healthcare systems because they provide a complete review of all the evidence. Together, the CMS coverage in the US and the UK appraisal sent a powerful signal to cardiologists everywhere that CT-derived FFR analysis was a clinically valuable, economically sound diagnostic tool. This one-two punch of regulatory clearance followed by broad reimbursement shows exactly how scientific validation and market access have to work together.

The Long View: One-Year PLATFORM Follow-up and Evidence Evolution

To move from a cool new idea to a routine referral, you need more than just initial approvals. You need long-term evidence that the tool actually helps patients. The one-year PLATFORM follow-up for CT-derived FFR delivered just that. The PLATFORM study was originally set up to look at the diagnostic and economic effects of FFRct, but the extended follow-up gave everyone a window into the long-term outcomes for patients managed with this strategy. These kinds of long-term studies are what establish a diagnostic’s lasting impact on patient management, clinical events, and healthcare costs. This data gave referring cardiologists proof they could see, reinforcing their confidence that FFRct could accurately sort patients who needed revascularization from those who would do just fine on medical therapy. Building up this kind of real-world evidence (RWE) is what cements any diagnostic tool’s place in the clinic, especially one that uses advanced AI. And as coronary CT angiography programs got more experience, AI plaque quantification, which started as a research-only technique, began to enter the conversation for routine referrals. The very same datasets that supported FFRct analysis also became the foundation for developing risk prediction models, making CCTA even more useful for a total cardiovascular risk assessment.

Expanding the AI Footprint: Independent Plaque Analysis and Ambulatory Monitoring

The success of FFRct opened the door for more AI in the cardiac diagnostics market. Building on the now-established value of CCTA, companies like Cleerly created their own independent plaque analysis tools. Cleerly’s software uses AI to directly quantify and characterize the plaque burden in the coronary arteries from CCTA scans, giving cardiologists detailed information about the disease itself, not just the stenosis it causes. This analysis adds another layer of information, moving past a purely functional assessment to give a deeper look at the disease biology. Being able to precisely measure different plaque types (like non-calcified or low-attenuation plaque) adds another dimension to risk stratification and helps create more personalized treatment plans. The regulatory path was also clearer now. The original FDA De Novo for FFRct, plus the 510(k) clearances that followed for tools on the same modality, created a common, if demanding, gate for these imaging tools to enter the clinic, as seen in the FDA 510(k) database. This pattern of building evidence and getting regulatory approval wasn’t limited to imaging. You can see it in other areas of cardiac monitoring too. iRhythm Technologies, for example, applied a similar evidence-based approach to ambulatory rhythm capture with its AI-powered analysis of long-term ECG data. While it’s a different technology from CCTA, iRhythm’s path shows the same trend: AI diagnostics proving their clinical worth to become a standard part of the cardiac referral chain. The expansion across these different tools shows how AI is building more connected platforms for managing cardiovascular health.

The Referral Shift: From Research to Routine

So what’s the result of all these milestones? The initial De Novo clearance, the complete CMS coverage, the strong long-term clinical data, and the expansion into other AI analyses have totally changed coronary CT referral pathways. A tool that started in research labs is now something interventional and imaging cardiologists can’t do without in their daily practice. For the program leads, the clarity from the CMS LCDs and the sustained evidence from studies like PLATFORM took the guesswork out of using CT-derived FFR and advanced plaque analysis. It gave them confidence. The ability to non-invasively check both the anatomy and functional significance of a lesion, and get a detailed plaque workup, means they can be much more precise about who needs an invasive procedure. The result is fewer unnecessary catheterizations and better patient outcomes. The playbook is now clear for any new cardiac AI tool trying to get from the lab to the clinic: get clearance, secure coverage, and follow up with long-term data like the PubMed search for FFRct clinical outcomes shows. As AI models get smarter and our ability to generate real-world evidence improves, this kind of technology will only become more foundational to modern cardiology. You can check the CMS National Coverage Determinations for the latest.