MRI Safety: Why the Magnet Is Always On and What That Means for Implants

The magnet never turns off, and that is the whole point

The MRI magnet is always on. A clinical scanner is built around a superconducting magnet that holds its field in persistent-current mode around the clock, through nights, weekends, and building power failures, so any ferromagnetic object near it is a constant projectile risk. Because the field cannot simply be switched off before each patient, MRI safety is built from controlled space and documented device conditions rather than a switch flipped for each scan. The American College of Radiology four-zone model screens every person and object before it reaches the field, and MR Conditional labeling states the exact conditions under which an implant can be scanned safely.

Why "always on" is a design feature

Superconducting coils carry current with essentially no electrical resistance, so once the magnet is energized the current keeps circulating and the field remains steady without ongoing power input. Engineers design that field to be remarkably stable, drifting by only a tiny fraction over hours. Turning it off is not a matter of unplugging the machine. The field can be ramped down deliberately, or dropped abruptly in an emergency quench that releases the coils' stored energy, but a quench is disruptive, costly, and reserved for genuine emergencies. The practical result is that the bore and the space around it remain hazardous to ferromagnetic objects at every hour of every day, which means safety cannot rely on anyone remembering to power something down.

The projectile effect, and why distance is everything

The always-present field produces what MRI physicists call the projectile or missile effect. Ferromagnetic objects such as oxygen cylinders, steel wheelchairs, IV poles, and loose tools are pulled toward the magnet, and the pull intensifies steeply as they approach the bore because the force depends on how fast the field strength changes across space, not on the field alone. An item that feels only a gentle tug several steps away can be pulled from a hand or a cart near the opening and accelerate into the scanner, where a patient is lying. Documented injuries and deaths from ferromagnetic projectiles are the reason this is treated as a first-order hazard rather than a theoretical one. The ACR guidance from Kanal and colleagues, and the College's updated MR safety manual, treat ferromagnetic screening of people and equipment as a central defense precisely because the field cannot be switched off to make the room briefly safe.

The four zones: physics translated into a floor plan

The ACR four-zone model, formalized in the ACR Guidance Document on MR Safe Practices and carried into the 2024 update of the ACR Manual on MR Safety, converts this hazard into architecture. Zone I is freely accessible public space where the field poses no risk. Zone II is the interface where patients are received and screened, completing written questionnaires and interviews about implants, prior surgeries, and metal exposure. Zone III is access-restricted by physical barriers such as locked or badge-controlled doors, open only to screened patients and trained MR personnel. Zone IV is the magnet room itself, where the static field is strongest and always present. The zones escalate deliberately, so that by the time any person or object reaches the field, it has passed through layers of screening designed to catch ferromagnetic hazards and unverified implants.

MR Conditional is a contract, not a blanket clearance

Implants are where the always-on field meets modern device engineering, and this is where labeling matters. Under ASTM standard F2503, which the U.S. Food and Drug Administration recognizes in its nonbinding guidance on testing and labeling devices for the MR environment, every item falls into one of three categories. MR Safe means nonconducting, nonmetallic, and nonmagnetic, posing no known hazard in any MR environment. MR Unsafe means known to be hazardous in all MR environments, such as ferromagnetic scissors. MR Conditional is the nuanced middle: the device has been shown to pose no known hazard only under stated conditions.

Those conditions are specific and enforceable. A conditional label may specify a maximum field strength (commonly 1.5 or 3 tesla), a limit on the spatial gradient, a cap on the radiofrequency energy deposited (specific absorption rate), and sometimes restrictions on which body region may be scanned or how the device must be programmed first. Read that way, MR Conditional is closer to a contract than a clearance. A pacemaker, neurostimulator, cochlear implant, or aneurysm clip may be perfectly scannable in one configuration and hazardous in another, which is why the label and its exact parameters, not a general reassurance, govern whether and how the scan proceeds.

The logic that ties it together

Seen together, the pieces form a coherent chain of reasoning. The static field is always on, so ferromagnetic objects are always a projectile risk, so space is zoned and everyone is screened before crossing into the field. The radiofrequency and gradient fields can heat tissue and induce currents in conductive implants, so device labeling encodes the exact conditions under which those effects stay within safe limits. The 2024 ACR Manual refines who counts as trained MR personnel and how conditional devices are handled, but the underlying logic has held for two decades. None of it is arbitrary; each requirement answers a measurable physical risk. This is educational information about how the system works and is not medical advice about any individual, implant, or scan.