The Challenge: From FDA Clearance to Getting Paid
At Ventra Health, we are committed to delivering cutting-edge insight into the evolving intersection of artificial intelligence and healthcare reimbursement. Clients frequently ask for our guidance on opportunities to increase revenue within their practice. We are of the opinion that AI can have a net positive (after associated costs) impact on practice revenue currently. This net positive impact will grow exponentially within a couple of years. Providers should work to develop formal internal infrastructure to evaluate clinical AI applications and the corresponding reimbursement landscape.
This briefing provides a consolidated update on U.S.-approved clinical radiology AI applications that have achieved Medicare and/or commercial reimbursement through one or more of the three eligible pathways. Here is the complete description of the eligible pathways:
| Approval by at least one Medicare Administrative Contractor (MAC) |
| Medicare’s regional contractors can issue Local Coverage Determinations (LCDs) to cover new services. Gaining an LCD means Medicare will reimburse the AI service (at least in those regions) even without a national policy. Several AI imaging tools have secured MAC coverage in multiple regions, a strong validation that payers find them “reasonable and necessary.” |
| Inclusion in the CMS New Technology Add-on Payment (NTAP) program: |
| This CMS IPPS program provides extra inpatient reimbursement for new, costly technologies that show substantial clinical improvement. An NTAP is temporary (up to 3 years) but signals that a hospital can get up to 65% of the new tech’s cost on top of usual DRG payment. In 2020, an AI stroke detection tool became the first software to earn an NTAP, a milestone for AI in healthcare |
| Assignment of a Category III CPT (Current Procedural Terminology) code for emerging technologies: |
| The AMA assigns Category III CPT codes to emerging technologies to track usage and outcomes. These temporary codes (formatted with a four-digit number and a “T”) do not guarantee payment but are the first step toward Category I codes with formal reimbursement. AI developers often pursue Category III codes to enable billing for their service; if evidence and utilization grow, they can later petition for a permanent Category I code (with relative value units and pricing). For example, HeartFlow’s cardiac AI and Koios’s ultrasound AI both started with Category III codes – crucial for eventually securing broader reimbursement. |
The FDA has cleared numerous AI applications for clinical use, over 1,000 as of early 2025, with the majority in radiology. Here is a full list. However, it is important to note that FDA clearance is necessary for an AI tool’s clinical use, but it does not address payment. Reimbursement remains the critical hurdle for widespread adoption, and the reimbursement decisions are made separately by CMS and individual MACs.
Accordingly, many radiology AI solutions lack dedicated billing codes or coverage, creating a “wild west” of ad-hoc payment. A few pioneering applications, however, have broken through via special pathways, demonstrating how AI can achieve Medicare and commercial reimbursement. Below we highlight notable radiology AI applications that have navigated these hurdles and achieved reimbursement status through at least one of the pathways above.
AI Applications Achieving Reimbursement Milestones
As of April 2025, several AI applications in clinical radiology have been approved for reimbursement by various Medicare Administrative Contractors (MACs) and the Centers for Medicare & Medicaid Services (CMS). The table below summarizes leading AI radiology tools and the reimbursement pathways they have secured.
| AI Company | MAC Coverage | CPT Code (for AI) |
| Koios DS Breast and Koios DS Thyroid (Koios Medical) | Not yet covered by a specific Coverage Article | Yes – Category III CPT 0689T & 0690T for quantitative ultrasound tissue characterization (AI-driven) introduced in 2022. Early use shows these codes can increase reimbursement by 30% for professional and 10% technical fees. Source: auntminnie.com |
| HeartFlow FFR-CT Analysis | 5 MACs (38 states) cover noninvasive FFR-CT since 2021. | Yes – Category I CPT code 75580 in 2024. Sources: cms.gov, heartflow.com |
| HeartFlow Plaque Analysis (HeartFlow) | 4 MACs (Noridian, CGS, NGS, Palmetto, WPS) final Coverage Article effective 2024. | Yes – Category III codes 0623T–0626T for AI coronary plaque analysis. Shared with Cleerly/Elucid. Source: citoday.com |
| Cleerly Coronary Analysis (Cleerly) | Included under the same 2024 MAC coverage policy for AI plaque analysis, coronary CTA. | Yes – Uses Category III CPT 0623T–0626T (automated plaque quantification from CCTA). Sources: citoday.com, cardiovascularbusiness.com |
| Elucid PlaqueIQ (Elucid) | Included under the same 2024 MAC coverage policy for AI plaque analysis, coronary CTA. | Yes – Uses Category III CPT 0623T–0626T for plaque analysis (first FDA-cleared plaque morphology AI). Source: businesswire.com |
| Viz LVO Stroke (Viz.ai) | No separate MAC policy; AI is used within emergency stroke workflow. | No – No dedicated CPT yet for AI triage; usage tracked via ICD-10-PCS code 4A03X5D for inpatient billing. Source: auntminnie.com |
Understanding clinical radiology AI reimbursement developments is increasingly crucial. AI tools that have payment pathways can immediately impact service lines, revenue cycle, and patient care models. Case in point is Koios Medical’s AI-powered software, Koios DS. Koios DS has received FDA clearance and is eligible for commercial reimbursement through CPT codes 0689T and 0690T for quantitative ultrasound tissue characterization. These Category III codes are designated for emerging technologies and represent an initial step toward full Medicare reimbursement.
Outlook & Summary
The AI applications in the table above prove that AI can and will graduate to full reimbursement when backed by evidence and advocacy. Recent coverage decisions demonstrate Medicare’s growing openness to paying for AI applications that enhance diagnostic precision.
This document highlights the importance of staying abreast of and proactively evaluating new AI applications. Ventra will continue our commitment to support providers in optimizing reimbursable AI adoption strategies as they become available, maintaining our commitment to help providers navigate revenue cycle transformation through data-driven strategies and deep regulatory insight.
Our expert analysis helps clients evaluate technological investments through a reimbursement lens, empowering sustainable innovation. We anticipate more AI applications will pursue the pathways forged by early innovators. By staying abreast of future changes, practices can be early adopters of AI technologies that are backed by reimbursement, ensuring both better patient outcomes and sustainable operations. As the reimbursement landscape for radiology AI rapidly takes shape, Ventra Health will continue to monitor and inform on these and all AI reimbursement trends.
