Abstract
As most readers will already know, the 2007 Recommendations of the International Commission on Radiological Protection (Publication 103; ICRP, 2007) set out the foundations of the current system of radiological protection. Now 13 years old, most international guidance and many national regulations relating to radiological protection are based on Publication 103.
An important new feature in the 2007 Recommendations was a change in the way that doses from internal and external sources of ionising radiation were calculated. Previously, relatively simple mathematical models of the human body were used to calculate how energy from exposure to radiation is deposited in the various organs and tissues. With Publication 103 (ICRP, 2007), more sophisticated reference computational phantoms based on medical tomographic images replaced the simpler models.
Priority was given to developing reference computational phantoms for the adult male and female, as these are needed to calculate doses related to occupational exposure. These two phantoms, presented in Publication 110 (ICRP, 2009), are based on digitised medical image data of real people whose body height and mass were close to the reference data. The result was voxel (three-dimensional pixel) phantoms consistent with the reference anatomical and physiological parameters for males and females laid out in Publication 89 (ICRP, 2002).
The enormity of the effort and specialised skills needed to develop these phantoms may not be immediately obvious. Adjustment of the medical imaging data to match the reference parameters and a suitable posture was painstaking, as was the segmentation of the phantom into individual organs and tissues.
These reference phantoms were used to develop the dose coefficients for occupational exposure published in the Occupational Intakes of Radionuclides series of publications (ICRP, 2015, 2016, 2017, 2019, 2020b).
The current publication completes the reference computational phantom family, so that doses to non-adults can also be calculated. It is a large family: in addition to the two adult phantoms are 10 computational phantoms representing the Reference Male and Female at birth, 1 year, 5 years, 10 years, and 15 years of age. Like the adult phantoms, the paediatric phantoms are consistent with Publication 89 (ICRP, 2002).
Voxel phantoms were state-of-the-art in 2009, and stretched the capabilities of the computer hardware and Monte Carlo transport simulation codes available at that time. However, voxel phantoms have several important limitations. Voxels of any practicable size are unable to reproduce fine structures, some of which are important for radiological protection purposes. This includes, for example, the lens of the eye, the skin, and microstructures in bones and blood vessels. Calculation of absorbed dose to these tissues had to rely on specialised partial-body phantoms.
Like the adult phantoms, the reference paediatric phantoms also began with medical imaging data of real individuals, meticulously modified to align with reference parameters. However, the digital data were used to model the organs and tissues using mesh surfaces, capturing and adding as much fine detail as possible, based on anatomical knowledge. The results were voxelised at a very high resolution, considering the age-dependent total skin thickness, to preserve much of this detail. Today’s computer hardware and codes, enormously more powerful than those a decade ago, make it practical to use these high-resolution voxel phantoms to calculate dose coefficients. In fact, they have already been used to calculate dose coefficients for exposures to external environmental sources, to be released in the next ICRP publication (ICRP, 2020a). In addition, the Public Intake of Radionuclides series of publications, also using these high-resolution phantoms, will start to be released shortly.
This is not the end of the story. Soon it will be possible to take the next step: calculating doses using whole-body mesh phantoms directly. We are already preparing for this future. An ICRP publication on mesh-type computational phantoms is now in press, with the intention that these will be used as ICRP reference phantoms in the future. This will make it possible to represent the finest important structures, some just microns thick, in Monte Carlo simulations of particle transport and energy deposition.
Although mesh-type phantoms are not expected to have a significant impact on the numerical values of dose coefficients for everyday radiological protection purposes, other characteristics of these phantoms are attractive. Resizing mesh-type phantoms to match an individual more closely will be relatively straightforward, and the phantoms can be deformed to change posture. This could have a significant impact, for example, on the ability to assess absorbed doses to individual patients who rarely match the size and shape of the reference phantoms, and to reconstruct absorbed doses in accidental exposures where postures may also be an important consideration.
The paediatric reference computational phantoms in this publication mark an important shift from purely voxel-based phantoms to a hybrid format developed using anatomical data from medical scans and knowledge of fine anatomical structures, resulting in voxel phantoms of arbitrarily high resolution.