Is DeepHealth Owned by RadNet? What Vendor Ownership Means for Independent Radiology Groups

The short answer: Yes. DeepHealth, Inc. is a wholly owned subsidiary of RadNet, Inc., which trades on the NASDAQ under RDNT and operates outpatient imaging centers across the United States. That matters when you buy DeepHealth software because your reporting vendor and a competitor for your outpatient referrals are the same corporate parent. It is not a reason to rule the product out, and vertical integration brings real advantages. It is a reason to ask about data use, segregation and exit terms explicitly, in writing, before you sign rather than at your first renewal.

Updated August 2026.

This question started coming up in evaluations for a straightforward reason: DeepHealth got a lot bigger, quickly. RadNet acquired the French detection vendor Gleamer in March 2026 and folded it into DeepHealth. Three months later, on June 10, 2026, DeepHealth announced that Reporting Pro was commercially available after being introduced at RSNA 2025. A company that many US radiology groups thought of as a screening AI shop now sells the environment reports are written in.

So an independent practice sitting through a Reporting Pro demo is doing two evaluations at once. One is the ordinary software evaluation. The other is a question nobody teaches you to ask: what does it mean that the company selling me my reporting workflow also owns imaging centers down the road?

Is DeepHealth owned by RadNet?

Yes, and the relationship is direct rather than a minority stake or a partnership. DeepHealth is described in RadNet's own announcements as a wholly owned subsidiary, and it is the umbrella under which RadNet has consolidated its health informatics products, now including the Gleamer detection line acquired in March 2026. RadNet itself is one of the larger operators of outpatient diagnostic imaging centers in the United States, which is the part that makes this more than a corporate trivia question.

Worth being precise about what this is and is not. It is not a hidden ownership structure; RadNet publicizes it. It is not evidence of any improper handling of customer data, and nothing here suggests otherwise. It is simply a structural fact with contractual consequences, and structural facts are best handled in the contract rather than in the hallway after the demo.

What is DeepHealth Reporting Pro?

Reporting Pro is DeepHealth's AI reporting product. Per the launch announcement, it brings speech recognition, AI-generated clinical findings, measurements, AI-generated impressions, quality assurance and structured reporting into a single workflow, and it is available across imaging modalities including X-ray, ultrasound, CT, PET/CT and MRI. DeepHealth describes it as designed to integrate with any existing PACS and RIS, and names template migration from legacy systems as a capability, which is a direct pitch to groups leaving an incumbent reporting platform.

One line in the announcement is worth reading carefully, because it defines the product boundary. DeepHealth describes the seamless experience, connecting image management, AI findings, workflow orchestration and reporting, as what you get when Reporting Pro is deployed together with the DeepHealth Diagnostic Suite. Reporting Pro on its own is where the report gets written. Workflow orchestration, meaning what rises to the top of the reading list, sits in the wider suite. If your evaluation is about turnaround time rather than dictation ergonomics, establish early which of those two you are actually being quoted for. Our DeepHealth alternative page sets Reporting Pro against PowerScribe One row by row, including where each one wins.

At launch, commercial availability was stated as the United States and the United Kingdom, with Australia, South Africa and select European markets planned. RadNet framed the launch against a projected 15% radiologist shortage in the United States by 2029, which is the same staffing pressure most reporting AI is sold on.

Does it matter if your software vendor is also a competitor?

It depends entirely on who you are, and the honest answer has three cases.

There is a real counterargument, and it deserves stating. A vendor that runs imaging centers reads studies every day, which means its product decisions are informed by actual radiology operations rather than by a roadmap written in a software company. Dr. Jason Sinner, a radiologist and medical director at RadNet, is quoted in the launch material saying the product significantly reduces reporting times. That kind of internal proving ground is genuinely valuable and most software vendors do not have one. Vertical integration is not automatically a problem. It is a trade, and the terms of the trade belong in the contract.

What should you ask before buying software from a competitor?

These are the seven questions worth putting in writing. None of them are hostile, and a well-run vendor will answer all of them without drama.

1. Is our data segregated, and how? Ask about the technical boundary between your instance and the vendor's own operations, not just the legal one.
2. What is used for model training or product development? Get the answer at the level of report text, images and metadata separately. These are often treated differently.
3. Can our operational data be aggregated into benchmarks? Turnaround times, volumes and case mix are competitively meaningful even when de-identified. If aggregation is permitted, ask whether you can opt out.
4. Who inside the parent company can see anything? Ask for the access control answer, including support staff and analytics teams.
5. What happens if the competitive picture changes? If the parent acquires a center in your market mid-contract, does anything in your agreement change? Usually the answer is no, which is exactly why you want to know now.
6. What are the exit terms? Specifically: what format your templates and report data come back in, how long you have to retrieve them, and what it costs. This is the question that decides how much leverage you have at renewal.
7. Does the pricing assume the wider suite? Ask what capability you lose if you buy the reporting product alone, and get the answer before the quote.

Answers to questions like these tend to be scattered across a master agreement, a business associate agreement and two or three addenda, which is how obligations get missed at renewal time. Whatever you use to track contractual obligations and their review dates, get these commitments recorded there rather than leaving them in an email thread, because the person who negotiated them is rarely the person handling the renewal three years later.

Is DeepHealth a good alternative to PowerScribe?

For some groups, yes. The PowerScribe 360 timeline is what makes this urgent: annual renewals and maintenance end on August 31, 2026 and full support ends on August 31, 2027, which has pushed a large share of US radiology into a reporting migration on someone else's schedule. Reporting Pro is a credible option in that shortlist, and its template migration pitch is aimed squarely at groups with a decade of AutoText they do not want to rebuild.

The comparison to run is not really DeepHealth against Radiological.ai. It is DeepHealth against PowerScribe One and against whatever else makes your shortlist, judged on scope, integration burden, fully loaded first-year cost and the ownership question above. We keep a neutral version of that first comparison on the DeepHealth alternative page, and the wider set of switching decisions is grouped on the radiology AI alternatives hub. If the migration itself is what worries you, the PowerScribe 360 migration checklist covers sequencing, and radiology report templates and vendor lock-in covers what actually exports cleanly and what you will rebuild by hand.

Where Radiological.ai fits

We are not an imaging provider. We do not operate centers, we do not read studies, and we do not compete with our customers for referrals, so the questions in this article have short answers when they are pointed at us. The assistant flags suspected findings for your review, prioritizes the worklist so suspected urgent studies surface first, and drafts the structured report into your own template across X-ray, CT and MRI. It connects to the PACS you already run rather than asking you to adopt a stack.

It is decision support. It does not diagnose, and the responsible radiologist reviews, edits and signs every report. If the reporting side is what you are evaluating, structured radiology reporting covers drafting into your own section templates, and worklist prioritization covers the part that Reporting Pro leaves to the wider suite.

Buy the product that fits your reading list. Just make sure you know who owns it, and write down what they told you.