What you’ll be able to do after this lesson
- Explain why translational force on a ferromagnetic object depends on the spatial gradient of the field, not on field strength alone
- Say where in the room the force on an object is greatest, and why that is not the isocentre
- State what the 5-gauss line has historically meant, what the current standard says, and how to talk about both without confusing anyone
- Describe what ferromagnetic detection systems do and do not solve
- Run a practical ferromagnetic check on a person and on a piece of equipment
Why things fly
Two distinct forces act on a ferromagnetic object in the magnet’s field, and they behave differently.
Torque twists an object to align its long axis with the field. It is greatest where the field is strongest — near and inside the bore — and it is what matters most for an elongated implant such as an aneurysm clip or a wire.
Translational force pulls the object towards the magnet. This is the projectile force, and here is the part people get wrong: it does not depend on the field strength alone. It depends on the product of the field and the rate at which the field changes over distance — the spatial field gradient, often written dB/dz and expressed in tesla per metre.
That has a consequence that is genuinely counter-intuitive. At the exact centre of the bore, the field is at its highest and the spatial gradient is close to zero, so translational force is minimal. The force peaks somewhere around a quarter to a half metre inside the bore opening, along the inner bore wall, where the field is still very strong but falling away steeply. That region — not the isocentre — is the most dangerous place in the room for a loose ferromagnetic object.
This is also why acceleration is so brutal. An object released near the bore mouth does not drift in; it accelerates through the steepest part of the gradient. Oxygen cylinders, floor polishers, drip stands, mop buckets, scissors, hair grips and staplers have all done exactly this. A patient or a colleague between the object and the bore is in the path.
The 5-gauss line, and what changed
For decades, MR departments have drawn a line on the floor plan at the 0.5 mT (5 gauss) contour — the point where the fringe field has fallen to five gauss — and treated the area inside it as the region of concern. The 5-gauss figure entered practice as a conservative boundary for interference with implanted electronic devices, and it remains the contour named in a great deal of published national guidance and on a great many department floors.
The standards have since moved. Cardiac implantable electronic devices tested to ISO 14117:2019 must demonstrate immunity at higher field levels, and the ACR Manual (2026) records the change in as many words: “A recent update to the International Electrotechnical Commission (IEC) standard (IEC 60601-2-33:2022) has revised the fringe field limit to 9 G (0.9 mT)” — and the ACR Manual on MR Safety (2026) delimits the MR Environment by the 9 gauss line.
Because the 9-gauss contour lies inside the 5-gauss contour — closer to the magnet — a department already using the 5-gauss line is being more conservative, not less. The ACR says so explicitly: “Because the 9-G line is located within the 5-G line (ie, it is closer to the magnet), facilities following prior recommendations to use the 5-G line as the B0 hazard zone do not require adjustments to maintain the safety of the MR Environment.”
So the 5-gauss line has not been abolished, and a 5-gauss site plan has not been rendered wrong. What changed is the contour that defines the MR Environment in the current edition of the standard.
So how should you talk about it? Say what the boundary is, and say which standard you are quoting. “Our controlled access area is drawn at the 5-gauss line” is a correct and conservative statement. “The current IEC standard defines the MR Environment at 9 gauss” is also correct. What you must not do is assume the two are interchangeable when reading a device’s labelling or a site plan, and you must not assume the line on your floor was drawn to the current standard. Check the current edition and your local policy — and check what your own site plan actually says.
For the purposes of your daily work, the operational rule is unchanged and simple: the controlled area is bounded by a field contour marked on a plan, access to it is restricted, and nothing ferromagnetic crosses into Zone IV.
Ferromagnetic detection: useful, not a substitute
Many departments now install ferromagnetic detection systems (FMDS) — pillar or doorway units that alarm when ferrous material passes. They are a genuine improvement and are recommended as part of a layered approach in current guidance.
They are also frequently misunderstood.
An FMDS detects ferromagnetic mass in motion past its sensors. It does not tell you what the object is, where on the body it is, whether it is implanted or in a pocket, or whether an implant is MR Conditional. It will not reliably alarm on small items, on non-ferrous metals (titanium, most surgical stainless steels, aluminium), or on an object carried through outside its detection field. And it produces false alarms often enough that departments learn to work around it — which is the real risk.
Treat an FMDS as the last net, not the first. The screening conversation is the first net, the pocket check and gowning are the second, and the detector catches what those two missed. A department that leans on the detector instead of the interview has made itself less safe while feeling more so.
Doing the check
On a person. Gowning into department clothing without pockets is the single most effective control, because it removes the entire class of “forgot what was in my pocket”. Where full gowning is not used, the check has to be deliberate and physical: pockets emptied and turned out, then hands over pockets. Watch for the items people never think of as metal — hair grips and clips, hearing aids, dentures with metal clasps, wigs with fasteners, jewellery, body piercings, underwired bras, belt buckles, steel toe-caps, keys on a lanyard, vapes, and the phone that went back into a pocket after the screening. Add clothing with conductive or metallic thread, which current ACR guidance calls out specifically as an emerging burn and heating hazard as well as a projectile one.
On equipment. Every object entering Zone IV needs a positive answer to one question: what is its MR labelling, and does the labelling cover this scanner? If it carries no MR label, it does not enter. The classic offenders are oxygen cylinders, drip stands, infusion pumps, monitoring equipment, wheelchairs, transfer boards, ward beds, sandbags with steel shot, fire extinguishers, cleaning trolleys and floor buffers. Nearly all have MR-conditional equivalents; the failure is almost always that the ward’s own version came along with the patient.
Where a hand-held magnet fits. A small permanent magnet can demonstrate that an item is ferromagnetic. It cannot demonstrate that an item is not, and it tells you nothing about RF heating or device function. It is a teaching aid and a triage tool, not a clearance test — and it must never be used on or near a patient’s implant.
At the console
An inpatient arrives from the ward for an urgent head MRI. She is on the ward bed, with an oxygen cylinder clipped to the frame, a syringe driver running, and a nurse escorting.
Work outwards from the patient. The bed does not enter Zone IV — she transfers to the department’s MR-conditional trolley in Zone III. The oxygen cylinder does not enter; the department’s MR-conditional cylinder or the piped supply takes over. The syringe driver does not enter unless it is a labelled MR-conditional model used within its stated conditions, and the usual solution is an extension line with the pump outside Zone IV — check your local policy, because extension-line practice varies and has its own risks. The escorting nurse is non-MR personnel: she is screened, her pockets are emptied, and she is escorted continuously or she stays in Zone III.
None of this requires you to be certain about anything exotic. It requires you to treat every object as unlabelled until you have read its label.
Second scenario: your FMDS alarms as a colleague walks a patient through, and the colleague says “it does that, it’s oversensitive”. That sentence is the finding. Stop, find the cause, and if the cause cannot be found, the patient does not proceed until a Level 2 colleague has assessed it. An alarm you cannot explain is not a false alarm; it is an unexplained alarm. Log it.
Check yourself
Picture the last trolley transfer you did. Name every object that crossed into Zone IV with the patient, and for each one say how you knew its MR labelling. Then the harder question: in your department, where is the field contour line physically marked, which contour is it, and how would you find out?
This is a private prompt. Nothing you write is stored or assessed.
Key takeaways
- Translational (projectile) force depends on the product of field strength and spatial field gradient, so it peaks near the bore opening, not at isocentre. Torque peaks where the field is strongest.
- The 0.5 mT (5 gauss) contour is the historic and still widely used boundary. IEC 60601-2-33 Edition 4.0 (2022) defines the MR Environment at 0.9 mT (9 gauss), and the ACR Manual on MR Safety (2026) delimits the MR Environment by the 9 gauss line, which lies inside the 5-gauss line. Check the current edition and your local policy, and check which contour your own site plan uses.
- Ferromagnetic detection is a final net, not a screening method. It cannot identify an object, locate it, or assess an implant.
- Pocketless gowning removes an entire class of error more reliably than any check.
- Every object entering Zone IV needs a positive MR label that covers your scanner. Unlabelled means it does not enter.
- An unexplained alarm is not a false alarm. Stop, investigate, escalate, log.
References
- Pedrosa I, et al. American College of Radiology Manual on MR Safety: 2024 Update and Revisions. Radiology, 2025 — fringe-field contour revised from 5 G to 9 G https://pubs.rsna.org/doi/10.1148/radiol.241405
- IEC 60601-2-33:2022 (Edition 4.0). Medical electrical equipment — particular requirements for MR equipment for medical diagnosis
- ISO 14117:2019. Active implantable medical devices — electromagnetic compatibility test protocols for CIEDs
- MHRA. Safety Guidelines for MRI Equipment in Clinical Use, 5th edition (July 2026) https://www.gov.uk/government/publications/safety-guidelines-for-magnetic-resonance-imaging-equipment-in-clinical-use
- Elster AD. Questions and Answers in MRI — spatial gradient magnetic field https://mriquestions.com/most-dangerous-place.html