Lung Cancer Screening Software for Radiology Groups and What Each Layer Costs
A screening program needs four layers of software and only one is nodule AI. The 20 FDA OEB clearances, the recent ones to shortlist, a worked per study versus flat license example and the RFP questions.
By the Radiological.ai team
November 2026 · 7 min read
Worklist
Structured report
DraftRun the assistant to draft this report for review.
Decision support. The radiologist reviews and signs every study.
Decision support for qualified clinicians. Radiological.ai does not provide a diagnosis and is not a substitute for professional judgment.
The short answer: A lung cancer screening program needs four pieces of software, and only one of them is the AI most vendors sell. The pieces are eligibility and order capture, a cleared nodule detection or CAD tool, a reporting layer that produces a Lung-RADS report with a nodule table, and tracking that closes the loop on every follow-up recommendation. Twenty nodule CAD clearances sit under FDA product code OEB, so detection is the most crowded layer and usually the cheapest to swap.
FDA figures come from the openFDA 510(k) database, product code OEB, queried November 21, 2026. Coverage rules are from the CMS national coverage determination for lung cancer screening (NCD 210.14). Prices on this page are our own published prices. Every other vendor price is quoted, not published, and is not repeated here.
What does a screening program actually need software for?
Medicare covers annual low-dose CT screening for eligible people, and the CMS coverage determination sets conditions on both the patient and the imaging facility. That turns a plain chest CT service into a program with paperwork, and each piece of paperwork lands on a different tool.
| Program need | Software layer | Who sells it | What to check |
|---|---|---|---|
| Eligibility, shared decision visit, order | EHR order set or screening navigator | Your EHR vendor, or a screening platform | Does the order carry age, pack years and quit date into the exam record |
| Spot and measure nodules on the CT | Nodule CAD or detection AI | Coreline, Fujifilm, Infervision, Riverain, Siemens, V5med and others | The indication in the K number, and whether it fits your PACS |
| Write the report | Reporting and structured Lung-RADS template | Your reporting platform, or a reading assistant | Nodule table, category, prior comparison in the template |
| Follow up and registry data | Tracking and registry submission | Screening platform or a registry such as the ACR Lung Cancer Screening Registry | A recommendation that nobody tracks is the usual program failure |
The mistake we see in vendor shortlists is buying the second row and assuming the other three come with it. They do not. A CAD license places a marker on the nodule. It does not write the Lung-RADS category, it does not compare against last year, and it does not chase the patient who was told to return in six months.
How crowded is the nodule detection layer?
The FDA record shows 20 clearances under product code OEB since 2003, across 16 applicant names, with Coreline Soft, Medicsight and two Siemens entities holding two each. Counting names overstates the field because several are the same company renamed or acquired. The recent ones are the ones to shortlist.
| K number | Cleared | Product | Applicant |
|---|---|---|---|
| K254075 | May 27, 2026 | Synapse Lung Nodule AI | Fujifilm Corporation |
| K251203 | December 3, 2025 | AVIEW Lung Nodule CAD | Coreline Soft Co., Ltd. |
| K240554 | May 16, 2025 | InferRead Lung CT.AI | Infervision Medical Technology Co., Ltd. |
| K242919 | March 27, 2025 | V5med Lung AI | V5med, Inc. |
Only four OEB clearances were issued in the last two years of the record. That is a slow category, which is good news for a buyer: the products are mature, the differences between them are small, and nobody has pricing power. Read each Indications for Use statement, because the wording on patient population and scan type differs. Our Coreline Soft alternatives page lays out the Coreline record, and ClearRead CT alternatives does the same for Riverain, including single site VA orders.
What does lung cancer screening software cost?
Most detection vendors quote per site, per study or as a flat license, and none of the OEB vendors above publishes a price. The shape of the cost matters more than the headline. A per study fee scales with your program, a flat license does not, and the break even moves with volume. We worked through that math in lung nodule AI software cost.
Worked example, with illustrative numbers you should replace with your quote. A two radiologist practice reads 150 screening LDCTs a month, 1,800 a year. A per study quote at $6 a scan is $10,800 a year, and the same quote at 450 scans a month is $32,400. A flat license of $18,000 loses at the lower volume and wins at the higher one. Ask every vendor for both numbers at your current volume and at double.
The reporting layer is priced differently. Radiological.ai Practice includes X-ray and CT at $199 per reading radiologist per month billed annually, or $399 month to month. Two readers on annual billing is $4,776 a year, and it does not scale with scan count. That covers drafting and measurement support, not a replacement for a cleared detection license if your program needs one on paper.
Do you need a cleared tool, or a reading assistant?
It depends on what the program has to prove. The CMS determination does not require CAD software, so a program can run on structured reporting, a Lung-RADS template and a tracking list. If your health system, payer contract or accreditation reviewer expects a cleared nodule tool, buy one from the table above and choose on integration and cost per scan.
If the goal is faster, more consistent reads across the whole list, a reading assistant helps with more than lung CT. Radiological.ai flags suspected findings, including suspected pulmonary nodules on chest CT, places measurements on nodules the radiologist confirms, pulls the matching nodule from the prior study and drafts the structured Lung-RADS report with the nodule table filled in. It is decision support with no FDA clearance claim. The radiologist confirms every nodule, assigns the category and signs. The AI lung nodule detection workflow shows what that looks like on a study.
Questions to put in the RFP
- Which K number covers the module you are quoting? Product names change faster than clearances.
- What is the price at 1x and 2x our volume, per study, per site and flat, with the renewal uplift cap in writing.
- Does the result return to PACS and the report, or does the radiologist retype the nodule table.
- Who owns the follow up list, and what happens to a patient whose six month recommendation has no scheduled exam.
- What happens to the data when the contract ends.
For the rest of the stack, the AI in lung cancer screening guide covers how the read itself changes, and follow-up recommendation tracking software covers the fourth row of the table. The FDA cleared radiology AI vendor list explains how to verify any K number yourself in about four minutes. The ACR also publishes the Lung-RADS assessment categories your template must follow.
Ready to see the reading side on a chest CT? Check the published plans and start with the number of radiologists who read the screening list.
See Radiological.ai read a study
The assistant flags suspected findings for review, prioritizes the worklist so urgent studies surface first, and drafts the structured report into your template. You review, edit and sign every study.