Introduction Defence Medical Services, supporting CBRN Casualty Care, requested that Dstl consider the use of a crude radiological dispersal device (‘dirty bomb’) in a mass casualty scenario, and the effect of a radioactive fragment embedded in a patient on a diagnostic medical radiograph, as well as the impact on the patient and surgical team. Initial work showed an unpredicted white ‘bloom’ artefact, opposite to the expected darkening; subsequently this trial empirically tested all Defence mobile X-ray systems (DRagon, DART, and DR-Go).

Method Measurements were taken with DART equipment; a 600 GBq sealed iridium-192 source at varying heights to simulate different activities; and a caesium iodide flat panel detector. The width of the resulting artefact was plotted against source height, to determine a threshold dose rate at which the bloom was indistinguishable. From this, minimum detectable activities (MDAs) were calculated for several fragment depths.

Results The MDAs, without the scatter reduction grid (MDAs with it are higher), for radioactive fragments 0.5 cm and 24.5 cm from the detector are 0.79 ± 0.06 GBq and 1900 ± 140 GBq, respectively. Other effects are discussed such as inconclusive results on changing exposure factors, shadows and an unaffected central region at high dose rates. Anatomically representative images were also taken with a test phantom for future clinical training use.

Conclusion Even for fragment activities below the MDA, surgical teams are likely to approach occupational dose limits within an hour, and in some cases, minutes. Patient images could appear normal, with the injury site still receiving several hundred Grays, having severe impacts on patient outcomes and clinical decision making, such as amputation. This work has contributed to surgical and CBRN medical training, planning recommendations in NHS Emergency Preparedness procedures, and will support further work to develop UK and international clinical guidelines, refine dose thresholds, and investigate other imaging modalities.