AI was meant to replace radiologists, it hasn't. Imaging AI is past the demo stage. Hundreds of tools are cleared, and the companies that run the scanners and the reading rooms are buying the best ones. The open question is who pays for each read. This report gets you up to date on the whole market.
12 sections to help founders and investors understand the medical imaging AI market with clarity
Before the numbers, we set the limits of the market: what we count, what we leave out, and which markets sit right next to it.
A tool that flags a brain bleed on a CT scan is in. So is software that makes an MRI faster or sharper. A system that only stores images, or a camera that inspects parts in a factory, is out.
With that box drawn, you can compare companies and numbers without guessing what's inside.
Medical imaging AI is maturing, and the companies that run the scanners and the reading rooms are buying it.
We've found a lot of interesting signals. RadNet agreed in March 2026 to buy French imaging AI company Gleamer for up to roughly $270M. And Sectra agreed to buy Oxipit, whose ChestLink tool can set aside certain normal chest X-rays in Europe. AI is moving into mainstream imaging platforms and workflows.
Here, we'll look at why imaging groups and software platforms are buying AI teams, which uses get paid first, and what that leaves for the startups that are still independent.
Imaging AI market sizes online rarely match. Some include all radiology software. Others count the scanners too.
We built our own first-principles estimate instead, then checked it against other sources. Every assumption is written down, so you can see how we reach the number and why it differs from the rest.
We also show a high and a low case, depending on how fast hospitals start paying for AI.
Scan volumes keep growing, and there aren't enough radiologists to read them all. Every delay in the reading room is a patient waiting for an answer.
We compare imaging departments, outpatient clinics and other buyers of medical imaging AI. We rank what hurts most, from scan backlogs and delayed reports to expensive equipment time, and what could trigger a purchase.
That tells you which problems a radiology chief will pay to fix, and which ones only look good in a research paper.
Detection models, triage tools, automatic measurements, faster MRI reconstruction, AI that drafts the report: imaging AI covers far more than reading a scan.
We break down the stack, from the scanners and image archives to the platforms that run dozens of models at once and the reporting tools radiologists use all day. We also flag the bottlenecks, like images from different scanner brands, slow links to hospital systems, and models that only cover one finding.
Then we look ahead, at foundation models that read many types of scans and draft a full report.
Per-scan fees, yearly licenses, marketplace deals through image software vendors, AI bundled with the scanner, and a few payment codes from insurers: imaging AI gets paid in several ways.
We compare imaging tools paid for by hospitals, scanner makers and insurers. Some save time without separate insurance payments. Others help imaging centers see more patients. A few have reimbursement codes that support per-use fees.
You'll see which models still make money once the big platforms bundle AI, and which ones turn into a free feature.
Some brakes are well known: most AI reads aren't paid for by insurers, hospital budgets are tight, and every new tool needs its own clearance.
Others are easier to miss. A model trained on one scanner can slip on another. Radiologists may trust a tool too much, or ignore it. Platform vendors can give AI away for free. And who is liable when AI misses a cancer is still unclear.
In this section, we map each risk, rate its likelihood, and show which ones could slow hospitals down the most.
A growth rate only matters if you know what drives it. So after the numbers, we explain the forces behind them.
Some drivers are easy to see: more scans every year, too few radiologists, and screening programs that keep expanding. Others work slowly, like payment codes for AI, platforms that make it easy to switch on a new model, and evidence from large trials. Some of them push each other forward.
Before building or investing, you should know what makes a radiology team keep using a tool every day, and who will pay for it once the pilot ends.
Money still flows into imaging AI, and more and more exits are sales to imaging groups and software platforms.
We'll compare funding for tools that spot one condition, broader imaging platforms and models that read many kinds of scans. We'll also look at reimbursement, scanner partnerships and which companies investors see as acquisition targets.
We'll look at recent rounds and acquisitions to see which imaging AI companies can stay independent, and which ones are being built to be bought.
Lunit and Qure.ai have become the poster children of imaging AI, but the market is much wider than two names.
Some of the most interesting players never try to read the whole scan. They speed up MRI exams, measure plaque in heart arteries, check that a mammogram was taken correctly, or turn a radiologist's dictation into a finished report. Patients never hear their names, but radiology teams use them every day.
Part of the reason they're easy to miss: many sell through scanner makers or software marketplaces, so their name sits behind someone else's logo. Others are cleared for one narrow finding, which makes them hard to compare with a broad platform. And a tool that reads lung nodules and one that cuts MRI time don't even have the same buyer.
Below, you'll find the market grids and ecosystem maps we've built, from detection to reporting, so you can see who does what and where the gaps are.
In imaging AI, being accurate is not the same as being used. A model can match a radiologist on a test set and still sit on the shelf if nobody pays for it.
So we studied why imaging AI startups fail. Some build a tool for one finding and can't sell enough of it. Others depend on one platform to reach customers, lose to a scanner maker that bundles a similar feature, or run out of cash waiting for a payment code.
If you're building, you'll see why a paying buyer, a place in the radiologist's daily work and a path beyond one finding matter as much as accuracy. If you're investing, you'll know what to check about reimbursement, distribution and the risk of being bundled away.
HeartFlow, Rad AI, Subtle Medical… a few imaging AI companies have found a way to get paid.
HeartFlow made heart CT analysis a reimbursable test. Rad AI focused on radiology reports. Subtle Medical helps imaging teams capture faster, clearer scans. Each solved a different part of the workflow, with different levels of commercial proof.
We'll look at how these teams found someone willing to pay, fit into the radiologist's day, and grew beyond their first product.
It's for founders, investors, radiology leaders, imaging center operators, scanner and imaging software vendors, and strategic teams who want a clear view of the medical imaging AI market.
It's useful if you're choosing which finding or step to build for, deciding whether to sell direct or through a platform, weighing an acquisition, or comparing reimbursement, regulation, funding and competition across the market.
It has 190+ pages. It is designed to be comprehensive enough to support serious market evaluation, while still being fast to read and easy to use.
We prioritize signal, structure, and visual clarity over unnecessary length.
We define the medical imaging AI market as AI software that analyzes or improves medical images for clinical use.
We include image interpretation, triage, measurement, reconstruction and workflow tools for radiology and other medical imaging.
We exclude general image storage, non-medical computer vision and diagnostic products that do not use imaging.
Yes. All key data points and assumptions are explained and sourced in the document.
All sources are tier-1: company filings, funding databases, public datasets, expert interviews, industry reports, etc.
Yes. The report was last updated on October 27, 2026 with the latest market signals.
Yes. This is a one-time purchase. There is no subscription required to access the document.
Yes. You can use NEWMARKET for 20% off when purchasing two reports, or NEWMARKETMAX for 30% off when purchasing five or more.
If a market you care about is not currently on the list, email us at team@newmarketpitch.com and we may be able to cover it.
No problem. You can reach out directly and we respond within 24 hours.
Imaging AI can spot a problem on a scan in seconds, but someone still has to pay for that read. This report brings the whole market together: size, buyers, technology, business models, funding, competitors, risks and where the next openings are.
