An MRI scanner a district hospital can actually buy
editorialDevices & Hardware · partly solved · hardest part: manufacturing, regulatory, capital
Most of the world has no access to MRI. Low-field designs are physically simpler and already exist as open prototypes, but nobody has carried one through manufacture and clearance.
This is editorial, not a record. Registry entries are cited facts about things that exist; an opportunity is our argument about something that does not. The links to existing work are checked against the catalog — the judgment around them is ours, and you should disagree with it where you know better.
High-field MRI is sold as a capital asset with multi-year service contracts, and the installed base is concentrated in wealthy health systems. The cost is not mainly the physics — it is the superconducting magnet, the cryogenics, the siting requirements, and a service model that assumes a hospital with a facilities department. Low-field permanent-magnet designs remove most of that, which is why they are the obvious open target: the barrier stops being fundamental science and becomes engineering and paperwork.
Open prototypes exist and work. What does not exist is a reproducible build: a frozen bill of materials, a magnet that can be sourced and shimmed without a research group, a quality management system, and one regulatory clearance in one jurisdiction that others can follow. The gap is not a better scanner — it is the boring 80% after the prototype.
Populations with no imaging access at all, where the alternative is not a worse scan but no scan. Also every research group currently rebuilding the same console.
Open-source low-field MRI scanner using a 50 mT permanent magnet, portable at roughly 150 kg, for head and extremity imaging.
Open-source MRI console built on low-cost Zynq FPGA hardware, providing spectrometer and acquisition control.
Graphical and command-line client for the MaRCoS open magnetic resonance control system.
Embedded server component of the MaRCoS open magnetic resonance control system, running on the scanner hardware.
Open-source three-axis measurement system with sub-millimetre precision, used to map magnetic fields for MRI construction.
Toolbox for computational magnetic resonance imaging, providing calibration and iterative reconstruction methods.
Streaming framework for medical image reconstruction, used widely for MRI reconstruction pipelines.
- www.opensourceimaging.org/project/osii-one/retrieved 2026-08-05 · via manual
OSI2 ONE documents a 50 mT permanent-magnet scanner at roughly 150 kg under CERN-OHL-W, establishing that an open low-field design exists at prototype stage.
Last revised 2026-08-05.