Midjourney, the company best known for AI image generation, has launched a healthcare division — Midjourney Medical — and announced a full-body ultrasound scanner it calls "Ultrasonic CT". The device aims to image the whole body in about 60 seconds using sound waves and water, with no radiation and no magnetic fields. The important caveat: this is a first-generation prototype with no regulatory clearance. Midjourney says it will launch by offering body-composition maps rather than diagnosis, with a first San Francisco location planned for late 2027.
Key takeaways
- Midjourney announced a new health division and a full-body ultrasound scanner on 17–18 June 2026.
- It's marketed as "Ultrasonic CT", but it is not a CT scan — it uses ultrasound, not X-rays or radiation.
- It's a Gen-1 prototype with no FDA clearance; Midjourney says it will start with body-composition maps, not diagnosis.
- The underlying imaging comes from Butterfly Network under a licensing deal — not from Midjourney's generative AI.
- The headline numbers (superior to MRI, 50,000 scanners) are Midjourney's own, unverified prototype claims.
What is Midjourney Medical?
Midjourney Medical is a new division of the AI image-generation company, announced in mid-June 2026 and presented as a long-term bet on diagnostic imaging. Its first hardware product is the Midjourney Scanner, demonstrated live at a San Francisco event. Founder David Holz described it as the first new whole-body imaging method in decades and framed its image quality as comparable to — in the company's words, in some ways superior to — MRI. Those are the company's framings, made at a launch, not independently verified findings.
What is "Ultrasonic CT" and how does it work?
The full name is "Fullbody Ultrasonic Computational Tomography", shortened to "Ultrasonic CT". The name is genuinely misleading, so it's worth being precise: despite the "CT" label, there is no X-ray and no ionising radiation. It is ultrasound. The proposed experience is unusual — you step onto a platform and descend into a shallow pool of water, passing through a ring of underwater ultrasonic sensors. The sensors fire sound waves through the body from many angles, and a compute cluster reconstructs the returning waves into cross-sectional images of muscle, fat, bone and organs in around 60 seconds.
What exists now versus what's only claimed?
This is the distinction that matters for any clinical reader.
| Exists now | Claimed or planned |
|---|---|
| A working Gen-1 prototype, demonstrated live | Image quality "superior to MRI" |
| Body-composition mapping (the stated launch capability) | Diagnostic capability (pending FDA submissions) |
| A licensing deal with Butterfly Network | ~50,000 scanners worldwide over six years |
| A public announcement and marketing site | A billion full-body scans per month |
| — | A San Francisco "Midjourney Spa" opening late 2027 |
Midjourney itself is explicit that diagnostic claims need regulatory approval: it says it is "starting by just giving you detailed body composition maps" and will submit test results to the FDA for increased capabilities over time. In other words, on day one this is not a diagnostic device.
Who is behind the technology?
The imaging physics isn't Midjourney's own. The scanner is built on a co-development and exclusive licensing agreement with Butterfly Network, signed in November 2025 (a deal worth up to around $74 million over five years, per Butterfly's regulatory filing). The prototype uses 40 Butterfly Ultrasound-on-Chip modules — transducers fabricated on silicon chips using semiconductor manufacturing — with future generations expected to use more. Notably, none of the licensed technology is generative AI: the part of Midjourney that made it famous has nothing to do with how this scanner forms images. Butterfly, for its part, has said it will help support any claims with valid clinical data "as applicable" — a measured position.
What should clinicians take from this?
For clinicians, the interesting question isn't whether the device is exciting — it clearly is. It's whether scan outputs are clinically validated, interpretable, appropriately governed, and integrated into a patient pathway. A beautiful interface and a 60-second scan don't answer any of those. The likely near-term reality is that some patients will pay for body-composition maps and a few will arrive at their GP asking what an unfamiliar output means — which is a clinical-reference and communication problem long before it's a diagnostic one.
That's the gap where a guideline-grounded reference helps: Ask iatroX is free, and lets clinicians check what UK guidance (NICE, CKS, SIGN and the SmPC) actually recommends when an incidental or unfamiliar finding lands in front of them — without implying any automated diagnosis. The scanner is a hardware story; the clinical question it creates is about evidence and follow-up.
Frequently asked questions
Is the Midjourney scanner a real medical device? It's a real, demonstrated prototype, but it has no regulatory clearance for diagnosis. Midjourney says it will launch offering body-composition maps and will seek FDA clearance for further capabilities over time.
Is "Ultrasonic CT" the same as a CT scan? No. Despite the name, it uses ultrasound (sound waves), not the X-rays of a conventional CT scan, so there's no ionising radiation involved.
Does Midjourney's generative AI power the scanner? No. The imaging is based on Butterfly Network's ultrasound-on-chip technology under licence; it's unrelated to Midjourney's image-generation models.
When and where will it be available? Midjourney plans a first San Francisco location — a "Midjourney Spa" — at the end of 2027, with a long-term ambition of roughly 50,000 scanners worldwide. Those are stated goals, not current availability.
Can it replace MRI or CT? That's Midjourney's framing, not an established fact. Ultrasound and MRI answer different clinical questions, and the "superior to MRI" claim is unverified — see our separate analysis on whether it could replace MRI.