Diagnostics & imaging AI · Clinical AI governance
Ferrum Health
Ferrum Health is a clinical AI governance platform for health systems. It helps teams deploy, manage, and monitor multiple AI models and shows how each one performs at each site.
Strong fit
- Health systems using many clinical AI models that need oversight and monitoring
- Imaging and clinical leaders who track whether models get worse over time, how they perform at each site, and their return
- Teams that want one way to test and launch both purchased and in-house models
Weak fit
- Buyers who want a single radiology detection tool
- Clinics with no plan to use AI
- Organizations that will not assign clinical owners for AI after launch
Bottom line
Ferrum Health is worth considering for health systems that run, or plan to run, several clinical AI tools and need one way to govern, monitor, and deploy them. It works best when each model has a clinical owner. Buyers who want a single AI tool that detects findings should look at Aidoc, Rad AI, or Viz.ai.
Score breakdown
Weights: Outcomes 35% · Product 30% · Implementation 20% · Pricing clarity 15%.
Ferrum Health's customers are health systems that use several clinical AI models, both purchased and built in-house. They use the platform to manage those models in one place, watch whether they get worse over time, test how they perform at each site, and launch new ones in a controlled way, while patient data stays in the environments they choose. Because it is a system for running AI tools, Ferrum works alongside detection companies such as Aidoc and Rad AI more than it competes with them. Pricing is quoted by the number of models and sites, the monitoring needed, and whether it runs in the cloud or on-premise.
Customers value testing models at each site and connecting several models more easily. Ferrum is smaller than the leading imaging AI companies and has less public proof of results than established detection vendors. Among the diagnostic imaging AI tools we review, it scores 6.8 overall, level with deepc and below Aidoc and ScreenPoint Medical (both 7.9). Aidoc, Rad AI, and Viz.ai fit buyers purchasing a clinical AI application, and Ferrum fits when the need is running many of those tools safely.
Competitor landscape
| Vendor | Overall | Ease of implementation |
|---|---|---|
| Aidoc | 7.9 | 7.5 |
| ScreenPoint Medical | 7.9 | 7.6 |
| Harrison.ai | 7.7 | 7.5 |
| Qure.ai | 7.4 | 7.1 |
| Riverain Technologies | 7.3 | 6.9 |
| CARPL.ai | 7.3 | 7.6 |
| Subtle Medical | 7.2 | 6.8 |
| DeepTek | 7.1 | 6.8 |
| AZmed | 6.9 | 6.7 |
| Enlitic | 6.9 | 6.5 |
| Avicenna.AI | 6.9 | 6.6 |
| Ferrum Health | 6.8 | 6.5 |
| deepc | 6.8 | 7.0 |
| Koios Medical | 6.7 | 6.3 |
Pricing
| Item | Detail |
|---|---|
| Model | A platform for overseeing and launching clinical AI, sold to health systems on quoted prices. |
| What usually drives cost | The number of models and sites, how much monitoring you need, and whether you run it in the cloud or on-premise. |
| What to ask before you buy | The cost to oversee the AI models you have now, with monitoring fees and fees for adding new models listed separately. |
| Published pricing | Quoted for each health system. Ferrum does not publish a price list. |
Prerequisites for purchase
| Need | Why it matters |
|---|---|
| What you need to get Ferrum Health to function | |
| Clinical AI models you use now, or a plan to launch some soon | An oversight platform needs models to oversee. |
| Clinical and IT leaders who review AI performance | Performance alerts need someone responsible for acting on them. |
| A decision on cloud, on-premise, or your current setup | The deployment choice has to be made before launch can start. |
| A way to reach data that keeps patient information in approved places | Rules on where data must stay need to be spelled out. |
| Measures of success for model safety, performance, and return | Agreed measures give the team a reason to use the platform. |
| What will maximize your value | |
| Start by monitoring models in daily use | Proof on models in daily use builds trust faster than starting only with new ones. |
| Review performance alerts with clinical department owners | IT alone cannot judge the effect on patient care. |
| Add every new model the same controlled way | Models connected outside the platform escape monitoring. |
| Show each site's results to the clinicians who use the tools | Clinicians change how they use a tool when they see how it performs at their site. |
| Add more models only after monitoring is reliable | Adding models before monitoring is reliable multiplies risk. |
| Deal-breakers | |
| You only want a single radiology detection algorithm. | |
| You have no AI models and no plan to deploy any. | |
| Clinical leaders will not review performance evidence. | |
| Security will not approve a controlled way to launch models. | |
| You expect Ferrum to fully replace detection products like Aidoc. | |
Value creation time frame
| # | Stage | Typical range |
|---|---|---|
| 1 | From signing to kickoff | 3-8 weeks for security review, patient data access, and listing your models |
| 2 | From kickoff to first use | 8-16 weeks to monitor or launch the first models |
| 3 | From first use to steady results | 3-6 months before running many models feels routine |
Methodology
| Weight | Factor | What it measures |
|---|---|---|
| 35% | Customer outcomes | Whether buyers get measurable operational or clinical-workflow results after go-live |
| 30% | Product | Capability depth, reliability, and fit for the job the category buys |
| 20% | Implementation | How hard it is to stand up, integrate, train, and stabilize |
| 15% | Pricing clarity | Whether a buyer can model total cost without a mystery quote |
| Label | Meaning |
|---|---|
| Highly recommend | Strong outcomes and product with manageable caveats |
| Recommend | Solid fit for the right buyer; know the tradeoffs |
| Conditional | Only with a specific use case or heavy caveats |
| Not recommended | Avoid for most buyers in this category |
Read our full methodology for how we weight scores and assign recommend labels.