Chipiron.

Here is a fact about the MRI machine that should bother you more than it does: the scanner is frequently more expensive than the room it lives in, and the room is expensive on purpose. A conventional MRI needs a very strong magnet, which needs liquid helium to stay cold, which needs a sealed and shielded space so the outside world's magnetism doesn't ruin the picture. Add it up and you get a device that costs on the order of a million-plus dollars, weighs like a small truck, and lives in a fixed room in a big hospital. The physics works beautifully. The economics work only for institutions.

Chipiron, a Paris-based startup founded in 2020, has taken the position that this is a solvable problem rather than a law of nature. Its bet is counterintuitive in the way good deeptech bets usually are: instead of building a stronger magnet, build a better listener. Run the machine at an ultra-low magnetic field - the company targets below roughly 30 millitesla, about a thousand times weaker than the 1.5 or 3 tesla scanner in a hospital - and then recover the signal you'd otherwise lose with an extraordinarily sensitive quantum sensor and a lot of software.

The trick to "a thousand times weaker but still clinically useful" is that a weaker field produces a weaker signal, and a weaker signal is easy to drown in noise. Chipiron's answer is a patented low-temperature SQUID - a Superconducting QUantum Interference Device, one of the most sensitive detectors of magnetic fields that physics has produced. The company describes its low-Tc SQUID volume gradiometer as the first SQUID antenna able to capture MRI signals in an open environment, and reports that it improves signal-to-noise ratio by up to ten times versus room-temperature detection. On top of that sits a stack of AI: linear and non-linear active noise cancellation, plus reconstruction tuned specifically for low-field data.

If that works - and "if it works" is doing real load-bearing labor in that sentence, because the company is still pre-clinical - the downstream consequences are large. Drop the field strength enough and two of the most expensive parts of an MRI simply disappear. You don't need the shielded room. You don't need the liquid helium. What's left is a machine that could plausibly be compact, affordable, and mobile, and that could scan people whom today's machines turn away.

Chipiron's own framing is careful: it is "rethinking MRI access, not replacing old machines."

That's a smart position, and not only for marketing reasons. The high-field MRI is genuinely excellent at what it does; nobody is going to swap a hospital's flagship scanner for an ultra-low-field box. The opportunity is the enormous space the flagship never reaches - underserved clinics that can't afford or house a conventional machine, mobile and remote settings, and patients who today get turned away. A pacemaker or certain implants can disqualify someone from a strong-field MRI; a machine running a thousand times weaker changes that calculation.

The company has also picked a first clinical target rather than trying to boil the ocean, which is the correct instinct for hard medical hardware. Its initial focus is breast imaging, with a breast-dedicated sequence and reconstruction designed to reduce the number of acquisition repetitions. Chipiron's stated mission leans hard into that: it talks about ending breast cancer through scalable MRI. Prove the technology on one well-defined, high-stakes use case, then expand.

Chipiron was started in 2020 by Evan Kervella, an engineer who serves as CEO, and Dimitri Labat, a physicist specialized in superconducting quantum materials who serves as CSO. They met that year through Entrepreneur First at Station F, the Paris startup campus - a program built precisely to pair complementary technical founders and push them at an ambitious problem. It's hard to imagine a cleaner match to the problem: the whole company rests on marrying a quantum sensor to a practical medical device, which is to say, on the CSO's physics and the CEO's engineering meeting in the middle.

The team has grown to roughly 52 people, organized around three stated values that are refreshingly free of jargon: passion, courage, and determination. For a company whose core claim is "we can do the seemingly impossible thing with the weak magnet," courage is at least an honest thing to put on the wall.

Kervella's careful phrasing - "the clinical impact we aim to achieve" - is worth noticing. It is the language of a company that knows the hard part is still ahead. The money validates the approach; the clinic will validate the product. Labat has framed the same milestone in the CSO's register: the coming images, he has said, "will form the basis for the product that will profoundly transform medical imaging practice on a global scale." Big words, appropriately hedged behind "will form the basis for."