Midjourney, the AI research lab known for its image-generation tools, is pivoting into medical hardware. The company announced a project called Midjourney Medical, a full-body scanner that produces high-resolution internal images in less than 60 seconds using sound waves instead of X-rays or magnets.
How the scanner works
The device relies on a ring of half a million underwater ultrasonic sensors. Those sensors fire sound waves through the body, and thousands of computers reconstruct a three-dimensional anatomical model from the echoes. The company says the process runs at nearly 100 times the speed of a conventional MRI — and delivers comparable resolution without the radiation of a CT scan or the confined tube of an MRI machine.
Ambitious rollout targets
Midjourney plans to place 50,000 units worldwide by 2031, with a combined capacity to handle 1 billion imaging sessions per month. That scale would dwarf current MRI availability: the United States, for example, has roughly 12,000 MRI machines. The company is funding development entirely from its existing product revenue — it carries no external investors.
AI’s role in the shift
Researchers describe AI as a structural shift in medicine, not just an incremental improvement. Machine-learning models already outperform human pattern recognition in radiology, genomics, and pathology. Some medical-ethics observers argue that failing to deploy AI-assisted diagnosis could soon amount to negligence — a standard that would accelerate adoption of tools like the Midjourney scanner.
Experts suggest that AI-driven drug discovery and early detection, combined with faster imaging, could enable cures for most diseases within 10 to 15 years. The claim is bold, but the pace of progress in medical AI has been steadily increasing.
What comes next
Midjourney has not announced a prototype timeline or regulatory filings. The scanner will need clearance from agencies such as the U.S. Food and Drug Administration before it can be sold. The company has not said whether it will manufacture the devices itself or license the design. Those questions will determine whether the 2031 target is achievable — or just an aspiration.
