Ferromagnetic Detection for MRI Safety

An MRI magnet is always on. Not during a scan — always. Outside working hours, during maintenance, in a power cut. That single fact is the origin of the projectile hazard, and the reason ferromagnetic detection systems exist at the door of every well-run MRI suite.

The hazard being managed

A superconducting magnet produces a strong static field that extends beyond the bore as a fringe field. Any ferromagnetic object entering that field experiences an attractive force that rises steeply as it gets closer — and the force scales with the spatial gradient of the field, not merely its strength.

The practical consequence is that an object can feel almost nothing at the doorway and become uncontrollable a metre later. Once it goes, it accelerates into the bore. Oxygen cylinders, IV poles, wheelchairs, floor polishers, tools, scissors and gas cylinders have all caused documented incidents. Where a patient is in the bore, the object travels toward them.

Two features make this hazard distinctive:

  • It is continuous. There is no "off" state to rely on, so the control must be present whenever the room can be entered
  • It defeats intuition. Staff reasonably judge risk by what they can feel, and the field gives almost no warning before the force becomes irresistible

Where FMD fits: the zoning model

MRI suites are organised into access zones, with Zone IV being the magnet room itself and Zone III the restricted region immediately controlling access to it. Screening procedures — questionnaires, checklists, changing patients into gowns, staff training — are the primary control.

Ferromagnetic detection is positioned at the Zone III to Zone IV boundary, as the last technical check before entry. Professional guidance, including the ACR Manual on MR Safety, describes such systems as a supplement to a proper screening programme, not a replacement for one. That distinction matters: FMD catches what procedure missed, and procedures do get missed under time pressure, during emergencies, and by staff from other departments who enter rarely.

The failure mode FMD addresses is not deliberate rule-breaking. It is the cleaner with a steel-handled mop, the porter pushing an unfamiliar trolley, the resuscitation team arriving at speed, or the technologist who has done the check correctly ten thousand times and is distracted on the ten-thousand-and-first.

How FMD differs from a conventional metal detector

This is the central technical point, and it is frequently misunderstood. An airport-style metal detector is the wrong tool here.

Conventional metal detector Ferromagnetic detector
Principle Active — transmits a field, senses eddy currents Passive — senses distortion of the ambient field
Detects All conductive metals Only ferromagnetic material
Relevance to hazard Poor — flags harmless aluminium, titanium, coins Direct — flags what the magnet can attract
Emits a field Yes No
Nuisance alarms High Substantially lower

The nuisance alarm rate is not a comfort issue — it is a safety issue. A detector that alarms on belt buckles and pocket change trains staff to ignore it, and an alarm that is routinely ignored provides no protection at all. Detecting only what is actually dangerous is what keeps the alarm meaningful.

The passive principle

A ferromagnetic object distorts the local magnetic field around it. As it moves past a sensitive magnetometer, that distortion appears as a characteristic time-varying signal. The system is listening, not illuminating — which is why it emits nothing and cannot interfere with imaging.

Because detection depends on movement through the sensing region, a stationary object sitting beside a sensor is invisible to it. That is an inherent property of the method rather than a defect, but it must be understood when siting and interpreting a system.

Engineering considerations

Building or integrating an FMD system involves a set of problems that are largely about discrimination and environment:

  • Sensor arrays : typically multiple magnetometers arranged vertically so detection is consistent from floor to head height, and so approximate height of the object can be indicated
  • Signal discrimination : distinguishing a genuine ferromagnetic mass in motion from ambient noise: lift motors, passing trolleys in the corridor, door mechanisms, building steel, vehicles outside
  • Threshold and sensitivity design : sensitive enough to catch a pair of scissors in a pocket, stable enough not to alarm on an approved non-ferrous wheelchair
  • Operation within the fringe field : the electronics sit near a strong static field, and both the enclosure and the design must account for it
  • Alarm design : clear, unambiguous, immediately understood under stress. Visual and audible, with indication of where on the body the detection occurred
  • Access integration : interfacing with door interlocks, access control and logging, where the installation calls for it
  • Calibration and self-test : staff must be able to confirm the system is working, and it must fail visibly rather than silently

A detector that fails silently is worse than none, because it replaces caution with false confidence. Designing for visible failure is a core requirement, not a refinement.

Limitations worth stating plainly

  • It detects moving ferromagnetic material, not static objects
  • It does not reliably characterise implants inside the body — implant screening remains a documentation and questioning process
  • It cannot identify what the object is, only that ferromagnetic mass is present
  • Poor siting — near moving building metal or a busy corridor — degrades performance considerably
  • It does not replace screening, training, zoning or signage

Why it is worth having anyway

Every documented projectile incident happened in a facility that already had screening procedures. Procedures are necessary and they are not sufficient, because they depend on people performing correctly every single time, including during the emergencies when performing correctly is hardest.

Ferromagnetic detection adds a control that does not tire, does not get distracted and does not make an exception because the situation is urgent. For a hazard that is always present and gives almost no warning, that is a reasonable thing to want at the door.

Working on MRI safety instrumentation?

We have developed ferromagnetic sensors for MRI safety alongside MRI couch control and GUI interfacing.

Talk to our team