Considerable time and effort is spent in education about the dangers of MR, and in limiting the risks of accidental injury from projectile metallic objects. It is widely assumed that such projectile injuries are painful. However, this has never been fully tested. Our centre was decommissioning a 3T magnet, as part of a move to a larger premises with new equipment. This provided a rare opportunity to evaluate the injury / pain levels inflicted by various metallic objects, without the usual limitations of concern around damage to the imaging hardware.

PARTCIPANT RECRUITMENT:

The IRB declined to approve the study, so were immediately recruited as volunteers. In total, 20 members of the IRB (past and present) were recruited to this study, without consent. The cohort was split into two groups, each of 10 people. Group B was administered a cocktail of strong analgesics (200 mg Tramadol), sedative (midazolam) and anaesthesia (remifentanil).

MR ACQUISITION: Although the main purpose of this experiment was not to acquire images per se, we did include a standard gradient echo EPI sequence protocol, in an event-related design, to evaluate differential BOLD activations in the somatosensory cortex

PROJECTILE TESTING: A selection of seven common house-hold objects (containing various amounts of iron) was included in the test-battery. These included: (i) a kitchen knife (7 inch blade); (ii) meat-cleaver; (iii) kettle; (iv) steam iron; (v) hedge-strimmer; (vi) lawn-mower; and (vii) a kitchen sink. (see Figure 1). Each object was attached to a winch and manouevred to a distance of 1.25 m from isocentre before being released into the magnet. The sequence of objects was randomized to avoid ordering / expectation bias. Our experimental design included plans to repeat each projectile protocol, at least twice, to 'ensure reproducibility of results'.

PAIN RATING: At exactly 30 seconds after each impact, the volunteers were asked to rate their pain rating on a subjective scale ranging from 0 (‘no pain, I'm in complete comfort’) to 10 (‘cannot possibly take any more pain, please stop the experiment and call for an ambulance’).

PAIN: Figure 2 shows the distribution of subjective pain ratings for the various objects for the 10 participants in Group A. While all items caused considerable pain, the lawn-mower (on account of its mass) was rated most highly (ANOVA, p = 0.034) on the pain rating. 14 of the 20 participants required urgent medical attention that precluded their continued involvement in the study. In Group B, participants did not register a score on the pain rating, which we anticipated as zero pain.

INJURY: Figure 3 shows some examples of the extent of injury inflicted by repeated projectile injury. A physician was able to confirm that these injuries were genuine and not psychogenic.

REPRODUCIBILITY: All participants withdrew from the study and so evaluation of reproducibility was not possible.

SOMATOSENSORY ACTIVATION: In Group A, severe participant motion artefacts meant that that the fMRI results were unreliable. In Group B, despite the reduced participant motion, the presence of large metallic objects affecting the shim meant that image quality was insufficient to acquire a reliable BOLD signal.

Our results are consistent with the theory that releasing a range of large ferrous-containing objects into the magnet with a participant inside can affect image quality considerably (with some impact on the well-being of the participant also being hinted at). While this may support the use of a thorough screening program, we have not yet explored the use of increased sedative and emerging techniques for imaging around metal, which we believe may provide optimal image quality without impacting on workflow. This remains an area for future investigation.

This rare study and lends support to the notion that participant screening is important. Considerable time and effort is involved in ensuring that metallic objects are not brought into the vicinity of the magnet However, this comes at the risk of slowing down the acquisiton of data, and reducing overall workflow.

We formally tested whether this time investment is justified. In a retrospective questionnaire, all participants were of the opinion that it was.

We have shown that strong medication may obviate the need for some of this screening, although advanced methods for imaging around metallic objects would be needed as a complementary measure in the future.

We hope that by performing this study, and sharing the data, other groups will not need to perform similar experiments, and will be more enthused about safety testing.