Butterfly Network (BFLY): Transforming Imaging With Midjourney

Medicine has made significant progress in the past few decades, largely thanks to advances in biochemistry, pharmaceuticals, and biotechnology, giving us mRNA vaccines, advanced cancer therapies, monoclonal antibodies, stem cell therapies, gene therapies, etc.

However, another segment of medicine has been somewhat more stagnant: medical imaging. Overall, medical imaging has become cheaper and more convenient for some technologies, such as endoscopy, but the image quality and the volume of medical imaging data have not significantly improved.

This is a pressing concern, as many patients with serious conditions are at first misdiagnosed, but 80% of such diagnostic dilemmas can be solved with simple imaging.

One issue is that high-resolution imaging technology like MRI is inherently expensive, as they rely on superconducting magnets needing to be cooled to close to absolute zero with liquid helium.

Another is that cheaper solutions like ultrasound are still relying on decade-old sensor technology that has only improved incrementally.

It is with that in mind that Butterfly Network was created. The company looked to replace the old piezoelectric ultrasound sensors with silicon chips instead.

“With our Ultrasound-on-Chip™ technology, we integrated thousands of transducer elements directly onto the circuits that control them at the wafer level. This innovation allows us to pack massive processing power into a chip the size of a postage stamp, eliminating the need for piezoelectric crystals—a groundbreaking shift in the industry.”

The company has now leveraged this unique technology into a partnership with the AI company Midjourney to create full-body scanners capable of, in time, creating a 3D model of a patient in just 60 seconds, potentially replacing MRI and other ultrasound imaging technologies entirely.

Butterfly Network Overview

Butterfly Network History

Butterfly Network was founded in 2011 by world-renowned scientist and serial entrepreneur Dr. Jonathan Rothberg. He previously founded CuraGen, one of the first genomics companies, in 1991, which IPOed in 1999 and grew to a $5B valuation by 2000. He later founded Ion Torrent in 2007, creating at the time the smallest and cheapest DNA decoder on the market, which was acquired in 2010.

So this is with an extremely solid background in advanced medical machine experience that Butterfly Network was founded with a mission: bringing to ultrasound technology the advances of silicon manufacturing.

The key idea is that with advanced methods to modify silicon wafers, the same method used to create computers’ and data centers’ memory and chips, ultrasound could be created and measured.

So rather than relying on vibrating lead-based crystals (piezoelectric technology), silicon ultrasound systems rely on micro-electromechanical systems (MEMS) and PMUT (Piezoelectric Micromachined Ultrasound Transducers).

Source: Butterfly Network

The motivation for this focus on media imaging after previous experience in molecular biology was driven by Dr. Rothberg’s daughter’s personal struggle with kidney issues, leading him to want to find a way to make medical imaging universally accessible.

The company name is taken from Dr. Rothberg’s meeting with MIT professor Max Tegmark at a physics conference, who was utilizing a data-processing architecture called a “butterfly network” to connect radio telescopes.

The company produced its first “Ultrasound-on-Chip” prototype in 2015, and its first commercial-ready ultrasound device in 2018, ahead of its 2021 IPO. The initial Butterfly iQ ultrasound sensor was followed by iQ+ in 2020 and iQ3 in 2024.

Source: Butterfly Network

Butterfly Network By The Numbers

Butterfly Network is still a small, lean startup by employee count, with only 220 full-time employees globally. This is also the result of a staff reduction effort since peak sizing in 2021 (~463 employees), in order to secure a good enough cash runway before hitting the acceleration of current sales and partnership deals (see below “Butterfly Network’s Future).

The company has been growing its revenues from $73.4M in 2022 to an expected $117M-$121M in 2026, and reduced its net loss from $136.5M to $21M-$25M for the same period, thanks to a strong gross margin of>65%.

Currently, Butterfly Networks holds more than 600 patents related to its ultrasound devices.

The company benefits from the fact that ultrasound imaging is a booming technology with an increasing volume of research for the detection of many health problems, and even for use in innovative treatments like transcranial focused ultrasound therapies. Handheld and Point-of-Care Ultrasound  (POCUS) ultrasound publications have increased 2.4x since 2019, and emerging ultrasound application publications 6.5x.

Source: Butterfly Network

Butterfly Network Technology

The way silicon-based ultrasound imaging works is a three-step process:

1. Wave Generation: A silicon system applies an alternating voltage across a microscopic, suspended silicon or metal membrane. This causes the membrane to vibrate, generating high-frequency sound waves that travel into the body.
2. Echo Reception: As those sound waves hit different tissues (muscle, bone, tendons, organs, or blood), they echo back to the silicon chip. The returning sound waves vibrate the membrane, which the chip instantly converts into digital electrical signals.
3. Image Processing: These raw electrical signals are sent to a processor. The system calculates exactly how long the echoes took to return to map out the depth and density of the body’s internal structures.

So while this is roughly the same process as with piezoelectric crystals, the reliance on microscopic silicon elements allows for a much greater level of control and precision of the resulting image.

Butterfly Network uses as many as  9,000 silicate elements that vibrate to generate and receive sound waves, placed onto a single chip no larger than a postage stamp.

This gives this technology a few key advantages. The first is that it can span a vast range of frequencies, allowing a single probe to scan the whole body.

The second is the ease of wafer-level integration of sensors with electronics, helping miniaturization.

Lastly, because it uses the same technology used for other sensors and silicon chips, it can be instantly scalable by relying on world-class foundries like TSMC (TSM ). This is a radical departure from manufacturing methods that rely on dicing fine piezoelectric crystals and hand-soldering thousands of wires, which kept analog systems expensive and low volume.

Overall, this novel design enables the combination of phased, curved, and linear arrays within a single probe without compromising image quality, while incorporating advanced imaging capabilities in a handheld device typically found in high-end cart-based systems.

Butterfly Network Product Line-Up

For now, the core of the company is its line of ultrasound sensors to be used by doctors in hospitals, clinics, private practices, etc. Currently, the latest version of this product is the Butterfly iQ3.

While this initial handheld and point-of-care (POCUS) is already a $2B Total Addressable Market (TAM), giving Butterfly a lot of space to keep growing, the potential for ultrasound-on-chip technology is a lot wider, spread over all healthcare and surgical markets, with care facilities and hospital car ultrasounds another $26B TAM, and further potential in other health-related markets.

Source: Butterfly Network

The ultrasound-on-chip technology will likely not be limited to just Butterfly Network devices. Instead, the company is adopting a hybrid approach to also license its technology to selected medical device partners through the Butterfly Embedded program. The concept is important to expand the applications of this innovation to its full potential.

“Through our DSK and APIs, innovators can built entirely new imaging applications and extend what ultrasound can do.”

The same approach is taken with Butterfly Garden, an AI marketplace for AI-powered apps for Butterfly devices, with already 30+ companies, usually narrowing down on diagnostic and custom AI models to analyze the ultrasound data for specific use cases: cardiac, articulations, veins, birth defects, etc.

Source: Butterfly Network

The company integrates its ultrasound images in its Compass AI operating system, integrating the handheld device, Butterfly Cloud, third-party ultrasound systems, and integration to hospital IT systems like patient files and image databanks.

Midjourney Full-Body Scanners

In itself, the performance and characteristics of the ultrasound-on-chip technology are impressive, but might seem “only” like the next step in ultrasound sensors. However, the flexibility, size, price point, and scalability of this technology are now potentially enabling a complete revolution in diagnostics.

This is because it is going to be used by Midjourney’s recently unveiled full-body scanners, which use ultrasound to create a 3D model of a person’s entire body in less than 60 seconds, in spa-like conditions, far away from medical and hospital settings.

“Ultrasonic CT lets us aim for whole-body imaging that’s in many ways superior to even MRI machines, but the scan takes as little as 60 seconds. There is no radiation, no powerful magnetic fields—just sound and water and 60 seconds.”

The level of detail is so much higher than that of MRI that if this scanner fulfills all its promises, it might make other scanner technology obsolete in a very short time. Especially as it is very expensive, and it typically takes 60 to 90 minutes to do a full-body MRI.

Source: Midjourney

The sensors in the scanner are half a million tiny squares each the size of a fine grain of sand, and each capable of acting as both a tiny speaker and a tiny microphone. These sensors create structured ultrasonic waves, whose ripples are recorded millions of times per second. They are clustered into a ring of 40 arrays of ultrasound emitter/receivers.

Midjourney is planning to quickly deploy its scanner in a test facility in SanFranscico in late 2027, before the deployment in 2028 of the scanners and associated spas to more cities, and an upgrade to the 3^rd generation of the system and a fleet of over 50,000 scanners worldwide by 2031.

This would enable Midjourney, powered by Butterfly Network’s technology, to perform up to one billion full-body scans a month. This should be enough to cover a huge percentage of the global population, or enough to give regular, monthly scans to a billion people.

“The collaboration is enabled through Butterfly Embedded™, Butterfly’s Ultrasound-on-Chip™ licensing and co-development business initiative, formerly known as Octiv. Butterfly previously filed the terms of its agreement with Midjourney in a Form 8-K on November 17, 2025, which disclosed up to $74 million in expected payments to Butterfly over a five-year term”

The resulting image of the body can then be annotated automatically by AI to identify organs, veins, bones, etc, with a resolution as small as half a millimeter (0.02 inches) and reconstructed as a whole body model using at least 21 AI servers with 2 petaflops of compute power.

Source: Midjourney

This absurdly large volume of data is why such an endeavor has not been tried before. Until very recently, this would have overwhelmed any supercomputers and gone way beyond our capacity to process the data into a useful format.

But with the past years’ progress in computing hardware and AI, the time has come.

The first obvious application of high-resolution, full-body scanners is detecting anything obviously abnormal. This can be abnormal growth like cancers, but also potentially aneurysms, problems in a given organ, etc.

“We think it’s completely possible that with enough early imaging in the future, the world could avoid 30% of all deaths and 50% of all healthcare costs.”

But the wider applications would be in fitness and the prevention of diseases in general. Regular scans will not just detect a pre-existing problem, but also directly assess changes in our bodies. It will measure with a precision never achieved before how children grow, how we gain or lose weight, how we age, etc.

“Collectively, we can begin to change our relationship with our bodies and start to ask questions like: if we can catch things early, can we change our lifestyles to correct them? And seeing our bodies change over time, alongside our actions, how much can we improve our health, our minds, and our lives?”

Butterfly Network’s Future

Butterfly Network seems to be exiting the startup / interesting tech phase and quickly growing into one of the hottest medical imaging companies in the world, as reflected by its booming stock price since the announcement of its deal with Midjourney.

For its existing product line-up, it is quickly entering into new markets across the Americas and Asia, including high-growth regions such as Brazil, with iQ3 expansion in multiple countries.

Continued advancement of next-generation Apollo semiconductor architecture is also designed to significantly increase data processing and compute performance.

Significant growth in the software pipeline since launching the next-generation Compass AI solution, it has signed its ninth Butterfly Embedded by mid-2026, and more  Butterfly Gardens are to be expected as well, diversifying the revenue stream.

New applications where ultrasounds were never significantly used before, like deep brain imaging in partnership with Aleph Neuro, could create further unexpected markets, proving the technology is not just an upgrade on existing ultrasound systems, but a radically new and more potent imaging method.

“Aleph explained its approach as delivering “MRI-level detail of the brain” without drilling or the bulk of an MRI machine — addressing what it sees as the two fundamental hardware bottlenecks in the field.”

And of course, the deal with Midjourney could prove even more transformative, in theory, able to make yearly or even monthly full-body ultrasound scans part of our routine health checks.

“Aleph’s announcement follows last week’s full-body ultrasound CT scanner unveiling from another Butterfly Embedded™ partner, Midjourney. Together, these milestones underscore the expanding impact of Ultrasound-on-Chip™ across the future of health and technology.”

And this is before even more science-fiction-like applications of ultrasounds, like bioprinting new organs or wearable diagnostic tools, might switch to ultrasound-on-chip systems as well.

Overall, this large open-ended potential can make it difficult to value the company, as so much will depend on the size and success of markets that do not exist yet, from full-body regular scanning to brain imaging.

However, as the company now has multiple paths to profitability, its risk profile is a lot lower, and it is likely that mainstream investors will progressively be attracted to it, mirroring the process that occurred for companies like Intuitive Surgical when it moved from new medtech to mainstream operating procedure (ISRG ) (follow the link for our report on this company).

Latest Butterfly Network (BFLY) Stock News and Developments