Export citation and abstract BibTeX RIS
Although there is routine, day-to-day, work in radiological protection, the discipline has a way of generating a wide variety of issues and problems that require the development of new or modified techniques for their solution. It is one of the purposes of this journal to provide a means whereby novel solutions to specific problems are brought to the attention of the radiological protection community, where they may have applicability beyond their original context. Therefore, in our annual review of papers under consideration for the Bernard Wheatley award, both the novelty and applicability of the reported work are important considerations. In our short-list for the 2018 award, we celebrate both the diversity of the issues that must be addressed in radiological protection and the ingenuity of our community in meeting the challenges that those issues pose.
In the management of radioactively contaminated sites, it is important to have an objective measure of the degree to which remediation activities improve the radiological situation across the site, or in selected areas. Chizhov et al (2018) address this issue at the Andreeva Bay site that was used for the refuelling of nuclear submarine reactors and the temporary storage of spent nuclear fuel. This complex, 2.3 ha site has been subject to remediation activities since 2002. To quantify the radiation situation across the site, Chizhov et al (2018) use maps of ambient dose equivalent rate (ADER) based on detailed surveys of the site since 2002, with kriging of the results used to give spatially continuous estimates of ADER. These estimates are then integrated spatially to give time-series of integrated ADER values that can be examined to determine the degree to which they are affected by key remediation activities. In addition, various mathematical techniques were used to examine both the spatial and temporal distributions of ADER values, with a view to optimising future monitoring of the site to ensure that time trends in integrated ADER values are adequately estimated. This paper is notable for the use of sophisticated statistical analyses adapted to the results obtained from existing detailed site surveys, to identify future requirements for monitoring during the on-going remediation process. The authors have an excellent record of well-thought-through studies at Andreeva Bay, having been recipients of the first-ever Bernard Wheatley Award (Chizhov et al 2014; see also Thorne 2015).
Of course, sophisticated statistical analyses typically require extensive datasets as input. Thus, techniques for rapidly producing detailed maps of environmental contamination or ambient dose rates are of importance, particularly in post-accident situations. Furthermore, there are advantages in developing systems that can be used by individual members of the public to evaluate their own situation, while also providing inputs to a central facility for analysis and interpretation. In this context, we welcomed the work of Martin et al (2018) who have developed a small, portable, CsI(Tl)-based gamma spectrometer. However, the key aspect of their system is the secured, Bluetooth low-energy data transmission between the detector and the operator's wireless-enabled smart phone operating a specially developed iOS App showing map and dose rate information, as well as RF data transmission of summary data back to a remote base station, with full gamma-spectroscopy results stored in the device. As the base station can receive data in close to real time from any number of devices, the potential exists to deploy substantial numbers of devices simultaneously in a survey, updating the surveying strategy continuously as results are obtained and interpreted. To illustrate the approach, three devices were deployed to undertake a radiation survey of the Geevor Tin Mine in Cornwall, with real-time feedback to each operator to ensure a high degree of ground coverage and a focus on areas of specific interest. Over 18 man-hours, the three units mapped a total of 70 km of survey length.
Clever mathematical techniques are also used by Liu et al (2018). The issue that is addressed is the calculation of radiation dose rates around irregularly shaped objects. This might not seem to be a novel issue. However, Liu et al (2018) have in mind the use of such dose rate estimates in real-time, virtual-reality-based simulations of cutting and dismantling tasks in nuclear facility decommissioning. Thus, they require a highly efficient computational technique that provides acceptably accurate estimates of dose rates at various locations in the vicinity of irregularly shaped objects. In their approach, the initial geometry is designed using three-dimensional, computer-aided design tools and a hybrid model of the object composed of cuboids and a point-cloud is generated automatically. That model is then converted to a weighted point-cloud model appropriate to the detector position and dose rates are calculated at the detector position using the point-kernel method. The validity of the approach was demonstrated by simulating a variety of basic geometries and showing that the adopted approach gave acceptable accuracy relative to MCNP and ran substantially faster than existing methods. A particularly nice feature of the paper is the use of figures to illustrate the mathematical techniques adopted. These include both sequences of geometrical configurations and flow charts showing the procedural steps.
Software also features in Joana et al (2018), but in this case a simple procedural programme not a complex mathematical analysis. This paper describes how the SEVRRA software package, which is being used to perform risk analyses relating to radiation therapy in approximately 100 facilities in Brazil, was used to evaluate risks at 20 facilities that collaborated in the study. The risk analysis facilitated by using SEVRRA distinguishes initiating events (observed events that could lead to an accident sequence giving rise to high risks) and missing barriers (where a barrier is a technical or management feature that could help to reduce a high risk to a medium or low risk). Barriers were further characterised by an importance index (a measure of the number of types of high-risk event affected by this barrier) and an impacted facilities index (the percentage of facilities that show potentially a high-risk level due to the absence of the barrier under consideration). The barrier effectiveness was calculated as the product of these indices. Thus, a high barrier effectiveness indicated that the barrier is protective against several types of initiating event in a wide variety of facilities. Use of the two indices and the derived quantity barrier effectiveness, allowed priorities to be identified in terms of the key barriers that need to be put in place to reduce risks in future. This highlighted weekly in vivo dosimetry to detect errors in the dose delivery process and an annual external audit for the control of reference dose rate as the top two barrier measures that needed to be implemented. The paper is of importance in illustrating that simple, well-defined risk-assessment procedures have an important role to play in ensuring radiological safety.
Various of our other short-listed papers also address radiological protection in medicine. Al-Affan et al (2018) showed how adding lead sheets of thickness 1–4 mm to the walls and floor of a treatment room maze could reduce the dose rate at the maze entrance by up to 80%.
This study made good use of the FLUKA Monte Carlo code to investigate the feasibility and likely effectiveness of the approach followed up by an experimental investigation using leadlined plywood boards with a thickness of 1.3 mm of lead. The observed reduction in dose rate was 50%, which was consistent with the Monte Carlo calculations, as complete coverage of the walls and floor could not be obtained. As the authors note, the technique could be a useful optimisation tool when modifying treatment rooms with short mazes and may be cost effective when modifying rooms with limited space. It might also be used to replace heavy maze doors or extended mazes when using mobile radiation sources in NDT. This paper by Al-Affan et al (2018) shows that simple adaptations of facilities can sometimes be very useful and demonstrates a nice balance of theoretical calculations and practical investigations.
At the other end of the scale of investigations, Balonov et al (2018) evaluated doses from medical examinations in Russia over the period 2009–2015. Two approaches were used, one was based on a major data collection by the authors from hospitals in St Petersburg and 17 other Russian regions and the other was based on collective dose estimates from the Russian federal data bank. As in other countries, the frequency of CT examinations is increasing most rapidly. For the various types of examinations, doses are similar to those observed in other countries, but there is a great deal of variation, with doses from similar radiography examinations varying by up to two orders of magnitude and doses from similar CT examinations differing by up to one order of magnitude. This paper merits attention not just because it represents a substantial source of information on medical exposures, but because it is informing the work of the authors in developing national guidelines on the justification of x-ray and nuclear medicine examinations and on Diagnostic Reference Levels (DRLs) for radiography. The variation in patient doses for similar examinations emphasises the importance of such national guidelines as a tool in optimisation.
Dose optimisation in medical imaging is the topic of the review by Samei et al (2018). This recognises a general desire to reduce effective doses from imaging modalities such as CT examinations to below 1 mSv but emphasises that this must not be achieved through a reduction in image quality that grossly increases the clinical risk to the patient. This leads into a discussion of the need to optimise the risk to the patient by determining the level of exposure that minimises the overall risk (corresponding to a level of exposure at which the radiation risk from increasing the exposure by a small increment is balanced by the decrease in clinical risk arising from the improvement in image quality that is achieved). Such optimisation can be targeted at the individual patient or at a population of patients. Optimisation over a population involves consideration of variability and diversity and this issue is considered before going on to addressing the practical matter of the choice of suitable metrics for characterising both radiation and clinical risks, bearing in mind that the two need to be expressed in comparable terms such that they can be traded-off against each other. A programme of gradual and continuing improvement is recommended in this thoughtful review that does not underestimate the practical difficulties in achieving optimisation across this sector.
Boos et al (2018) also address the topic of DRLs, but at the detailed technical level of adjusting the DRLs in CT examinations to allow for patients of different size. A key finding of this study was that there were strong linear correlations between the size-specific dose estimate (SSDE) for a patient and their water-equivalent diameter (Dw). Such correlations were developed for chest, abdominopelvic and upper abdominal CT examinations. By fitting the data, not only could a best fit of SSDE against Dw be obtained, but so also could 25th and 75th percentiles of SSDE as a function of Dw. As either the 50th or the 75th percentile is commonly used for DRL calculation, this shows that the specification of size-specific DRL values is feasible. As expected, in order to obtain adequate image quality, the SSDE and hence the derived DRL increases with an increase in Dw. Thus, reflecting on the comments of Samei et al (2018), there is more work to be done to confirm that use of a linear relationship between DRL and Dw achieves the optimum sum of radiological and clinical risk across the range of body sizes of relevance. Nevertheless, the comprehensive and carefully analysed data set presented by Boos et al (2018) makes an important contribution to this evaluation by clarifying that Dw is a useful quantity in determining both the radiation dose that is delivered and the image quality that is achieved.
In previous years, issues relating to internal dosimetry have featured strongly in our short list. This year, they are less well represented. However, Guiu-Souto et al (2018) address the important issue of estimating internal doses following the administration of I-131 to thyroid cancer patients in whom the metabolism of iodine may differ substantially from that in the standard, healthy individual assumed in the models adopted by the ICRP. To this end, Guiu-Souto et al (2018) adopt a simplified multi-compartmental model for iodine administered orally and taken up from the stomach into body fluids. As this model is to represent patients with thyroid cancer, it includes remaining or residual thyroid tissue, rather than the intact thyroid. The authors give starting values for all the parameters of the model, but then adjust those parameters to optimise its performance for the individual patient. An interesting aspect of the work is that optimisation was undertaken against the I-131 contents of one, two or three compartments (whole body, remaining thyroid tissue and blood, the first two estimated from external counting and the third from samples). This demonstrated that optimisation using data from three compartments substantially reduced uncertainties in model parameterisation and led to closely constrained dose estimates for all the tissues and organs considered. This is a nice example of a well-planned and straightforward study that yielded results of direct clinical applicability.
In a very different context, Hodgson (2018) reports on natural concentrations of Po-210 in urine. The need for this study arose from the potential exposure of a large number of people to Po-210 at the time of the poisoning of Mr Alexander Litvinenko, when there was a need to determine whether individuals were exhibiting enhanced concentrations of Po-210 in their urine. To address this, data were obtained from an extensive review of the literature and personal communications with authors of some of the more recent papers. Overall, 1385 individual measurements were identified, of which 566 were excluded as potentially contaminated by the Litvinenko incident, provided by individuals exposed to higher levels of Po-210 in their diet, or arising from potentially occupationally exposed persons. The remaining 819 measurements were analysed as a cumulative distribution function and distinctions between smokers and non-smokers were clearly demonstrated. This paper sets a standard for the compilation and analysis of published data on contaminant concentrations in urine. It would be useful if similar analyses were undertaken for a wide range of chemical contaminants.
Underpinning all of radiation protection is the epidemiological evidence of radiation risks in exposed human populations. However, the interpretation of epidemiological data can be difficult because relationships between radiation exposure and adverse effects can be modified by the influence of confounding factors, such as smoking. Kudo et al (2018) investigated this confounding effect in their study of male nuclear workers in Japan. This analysis related to 71 733 male workers, with lifestyle questionnaires being issued to different samples of these workers and used to identify potential confounding factors (smoking status, alcohol consumption and work with hazardous substances). A detailed statistical analysis of the data showed that adjustment for smoking reduced the excess relative risk from all causes from 0.97 to 0.45 Sv−1. For lung cancer, the reduction was from 1.94 to 0.94 Sv−1. However, it is important to recognise that these reductions in risk were not statistically significant. Also, as discussed by Akida (2018), it is debatable whether the questionnaire responses are fully quantitative and similar reductions in excess relative risk can be estimated using a simple categorisation scheme of current, former and never smokers. Although the limited statistical power of this study meant that significant results could not be obtained, the methodological approach used was rigorous and can be commended as a model for other such studies. As Akida (2018) notes, further studies on this cohort are required, e.g. to determine whether there is a reason for the large, but statistically not significant, excess relative risk for liver cancer that was observed. Such studies have been initiated by Kudo et al at the Institute of Radiation Epidemiology, Japan and results are awaited with interest.
Finally, the winner of the Bernard Wheatley Award for 2018 (Mille et al 2018) makes its own contribution to radiation epidemiology. As the authors point out, recent advances in cancer treatment have resulted in substantial improvements in survival rates. However, this enhanced survival gives more potential for adverse effects of the treatment, such as secondary primary cancers and cardiovascular disease, to be expressed. Both effects have been associated with radiotherapy. To investigate potential dose-response relationships, it is necessary to determine the doses received by the target tissues for these effects during treatment. However, the tissues of interest typically lie wholly or partly outside the primary beam, so conventional treatment planning algorithms may not provide adequate estimates of the relevant doses. In view of this, Mille et al (2018) systematically compared four different dose calculation algorithms, so that dosimetrists and epidemiologists could better understand the advantages and limitations of the various approaches at their disposal. In general, the algorithms included in commercial treatment planning systems were fast and acceptable in-field or near-field but were not acceptable out-of-field. For out-of-field studies, two Monte Carlo radiation transport codes (EGSnrc and XVMC) gave good accuracy, but EGSnrc was judged to be too slow to use for large epidemiological cohorts. Thus, the faster XVMC code was preferred. This paper took the award based on the importance of the issue being addressed, the comprehensiveness of the analysis undertaken, the clarity with which the work was explained, and the understandability and relevance of the figures.
Thus, again, we celebrate a year in which we received interesting and innovative papers on an extraordinarily wide range of topics from all over the world. Radiological protection may be a mature discipline, but ongoing developments in technology and software continually throw up new issues to address and the tools with which to tackle them. As always, we look forward to reporting these exciting new developments in the pages of this journal.