Midjourney's body scanner: what we know and what remains unproven

The demonstration looks like a film scene: you step into water, pass through a ring of sensors and, about 60 seconds later, a 3D map of your body appears. It is tempting to fill in the rest: no radiation, faster than MRI and able to find problems before symptoms begin.

There is a great deal of work between an impressive demo and a useful medical test. The Midjourney Scanner is an ultrasound-tomography prototype. The company has shown images and a business plan; the sources reviewed do not include independent clinical trials demonstrating disease detection, better outcomes or net benefit from screening healthy people.

The idea deserves attention. The medical claim still needs evidence. This guide separates the two so a futuristic image does not make the decision for you.

What was announced, and what was not

According to coverage of the June 2026 presentation, the system lowers a person on a platform through water and gathers ultrasound waves from a ring of sensors. Computing then reconstructs cross-sections and volumes. The prototype reportedly uses 40 Butterfly Ultrasound-on-Chip modules, while its first stated target is body-composition mapping.

The presentation also described a scan lasting about 60 seconds and a San Francisco site planned before the end of 2027. These are project descriptions and launch targets. They are not results from a comparative study.

Layer	What we have	What is missing
Announcement	Prototype, demonstration images and initial intended use	Full technical publication for the product
Performance	Claims about speed and resolution	Independent accuracy and repeatability
Diagnosis	Possible future applications	Sensitivity, specificity and indication-level comparison
Screening	Promise of observing the body without X-rays	Net benefit, follow-up and cost in a defined population
Regulation	Acknowledgement that medical uses require clearance	Verifiable device, manufacturer, intended use and pathway

A prototype can be excellent and remain a prototype. Caution does not mean dismissing it. It means not lending it results it does not have yet.

Why it is not MRI or CT

All three technologies create internal images, but they use different signals. MRI uses magnetic fields and radiofrequency. CT uses X-rays. Ultrasound sends sound waves and reconstructs information reflected by or transmitted through tissue.

The name “Ultrasonic CT” describes a tomographic geometry, not radiation-free CT. Surrounding an area with sensors and calculating slices from many angles may reduce some of the operator dependence of handheld ultrasound. It does not make the modalities interchangeable.

A similar-looking image does not guarantee the same tissue contrast or the same ability to assess lung, brain, spinal cord, bone or small lesions. Each intended use needs its own comparison against an appropriate reference standard.

The science is moving forward, but it does not validate this product

In April 2026, a Nature Biomedical Engineering study demonstrated whole cross-sectional human ultrasound tomography. The team used a 512-element circular receiver and a rotating transmitter to image the abdomen and thighs, reporting strong agreement with sequential MRI. It also explored abdominal fat thickness and needle localisation.

This is important progress. It was nevertheless a different system, with different hardware and research objectives. The study did not calculate the Midjourney scanner's sensitivity for cancer, evaluate a spa service or show that whole-body screening improves health.

This distinction matters in biohacking and health technology: a principle working in a laboratory does not automatically validate every product that uses the principle.

Why it uses water, and where ultrasound struggles

The water is not just visual theatre. Air between sensor and skin reflects much of the signal, which is why conventional ultrasound uses gel. Immersion improves coupling and makes it easier to transmit and receive from many angles.

Physics still applies:

• Gas: air-filled lung and bowel gas make it difficult to see structures behind them.
• Bone: it absorbs, scatters and refracts the signal, complicating adult brain, spinal cord and bone-interior imaging.
• Depth and anatomy: body size, implants, motion and geometry affect image quality.
• Resolution: improving detail may require frequencies that penetrate less deeply.
• Reconstruction: an algorithm can produce a convincing image while artefacts and uncertainty remain.

The FDA notes that ultrasound does not use ionising radiation and has a strong safety record when used prudently by trained providers. It also notes that ultrasound energy can slightly heat tissue or produce cavitation. “No X-rays” does not mean “no dose, no limits and suitable for every purpose”.

Removing radiation does not remove the screening problem

Imagine the scanner finds a small nodule. It may be important, benign, an artefact or something that would never cause harm. The report cannot settle that by itself. It can start a chain of new imaging, contrast, biopsy, specialists, cost and weeks of worry.

The NCI summarises four risks of screening: false positives, false negatives, overdiagnosis and treatment that does not improve survival. The FDA finds no demonstrated benefit for whole-body CT in asymptomatic people, while the ACR does not recommend total-body MRI for people without symptoms or risk factors because evidence is insufficient and non-specific findings can trigger unnecessary procedures.

Those positions do not assess this ultrasound prototype and should not be treated as a ban on it. They do illustrate a shared rule: imaging becomes a screening programme only when a defined population receives more benefit than harm.

A 2025 review of whole-body MRI in 9,024 asymptomatic people found a 1.57% confirmed cancer detection rate, but also frequent incidental findings, unstandardised protocols and no long-term outcome or cost-effectiveness data. Seeing more and helping more remain different questions.

Body composition: a reasonable hypothesis, not a recommendation

Measuring muscle, subcutaneous fat or abdominal distribution looks like a more plausible starting use than diagnosing dozens of diseases. Ultrasound already measures body structures in selected settings, and a surrounding system may reduce some limitations of a handheld probe.

But plausible is not yet better. The product would need comparison with DXA, MRI, regional ultrasound, circumference measures and functional tests. We need to know how much changes when one person is scanned twice, how performance varies across different bodies and whether the information produces better decisions.

For now, biomarkers chosen for the decision they can change, strength, aerobic fitness and medical history provide a stronger framework than collecting images out of curiosity.

Eight questions before paying or participating

1. What is the exact intended use? Body composition, research and diagnosis are different claims.
2. What is its regulatory status? Ask for the device, manufacturer and clearance for this indication, not a similar one.
3. Which reference standard was used? It must match the organ and purpose.
4. What errors does it make? Sensitivity, specificity, false positives, false negatives and inconclusive results.
5. Is it repeatable? Tracking needs to separate real change from system noise.
6. Who interprets it? There should be clinical responsibility and a route for urgent or uncertain findings.
7. What comes after the advertised price? Include confirmation imaging, appointments and possible procedures.
8. What happens to the 3D volume? Ownership, access, AI training, retention, portability and deletion.

If you have a symptom, a family history or a defined risk, the route remains a preventive assessment with a clinical question. A general scan does not replace the indicated test for that question.

Frequently asked questions about the Midjourney scanner

What exactly is the Midjourney body scanner?

It is an ultrasound-imaging prototype that reportedly uses water, a ring of sensors and computational reconstruction to produce body maps. An announcement and demonstration images are not the same as a validated diagnostic test.

Is it an MRI or a CT scan?

No. MRI uses magnetic fields and radiofrequency; CT uses X-rays; this system uses ultrasound waves. Similar-looking images do not establish the same tissue contrast, scope or diagnostic accuracy.

Can it detect cancer?

The reviewed sources contain no published clinical evidence showing cancer detection with this product or net benefit from screening healthy people. Independent studies would be needed for each organ, condition and population.

Why does it use water?

Water couples the sensors to the body and avoids air gaps that would reflect much of the ultrasound energy. It also helps collect signals from many angles.

Does no ionising radiation mean no risk?

No. Avoiding ionising radiation is an advantage, but physical limits, possible ultrasound-energy effects, false positives, false negatives, incidental findings, anxiety and downstream tests remain.

What is the most plausible near-term use?

Body composition and tracking changes in muscle or fat are plausible research uses. The product still needs to prove accuracy, repeatability and usefulness against existing methods before it can be recommended.

Is it cleared by the FDA?

Do not assume that it is. Coverage of the presentation said medical applications would need the relevant clearances. Before use, verify the exact device name, manufacturer, intended use and regulatory status in the official database.

What should I ask before having a scan?

Ask for the intended use, validation against a reference standard, error rates, who interprets the result, what happens after an uncertain finding, total follow-up cost and how your data are stored, shared and deleted.

Sources

1. Lawler R. “Midjourney goes from generating cat images to full-body ultrasound scans”. The Verge. 2026. Launch coverage.
2. Garrett A, Xu J, Oh D, et al. “Whole cross-sectional human ultrasound tomography”. Nature Biomedical Engineering. 2026. Paper.
3. U.S. Food and Drug Administration. “Ultrasound Imaging”. Guidance.
4. U.S. Food and Drug Administration. “Full-Body CT Scans: What You Need to Know”. Guidance.
5. American College of Radiology. “ACR Statement on Screening Total Body MRI”. 2023. Statement.
6. National Cancer Institute. “Cancer Screening Overview”. PDQ summary.
7. Martins da Fonseca L, et al. “Whole-body MRI for opportunistic cancer detection in asymptomatic individuals: a systematic review and meta-analysis”. European Radiology. 2025. Paper.
8. U.S. Food and Drug Administration. “510(k) Clearances”. Database and regulatory overview.

This article is informational and does not replace medical assessment. Do not use a general scan to delay care for a symptom or to replace an indicated test.