4. How is radiation exposure measured and assessed?
The SCENIHR opinion states:
3.5 Dosimetric aspects
To evaluate the risk contribution from scans performed with security scanners based on technologies using ionising radiation as described in chapter 3.2.1 and 3.2.2, it is necessary to describe the amount (doses) of ionising radiation received by the passengers. To do so it is important to clarify the various terms used.
3.5.1 Dose concepts
When dealing with ionising radiation, the basic concept used to describe the energy deposition caused by the radiation to any kind of material is the quantity ‘absorbed dose’ D. This is defined as the energy E imparted into a small amount of material:
D = dE/dm. where m is the mass of material.
This dose is a pure physical descriptor of energy transfer due to the ionising radiation. The values of the measurements are given in the SI unit Gray (Gy) which is J/kg. This physical parameter is in general not sufficient to describe the biological effects caused by ionising radiation. To take into account this dependence of the biological effects on the radiation type (alpha, beta, gamma, etc.) and energy, a weighting factor for the radiation quality wR (ranging from 1 to 20) has been introduced and an additional dose term has been implemented for radiation protection purposes. This is the quantity ‘equivalent dose’ H and is defined as:
H = wR * D
The SI unit for the equivalent dose is the Sievert (Sv), which is also expressed in J/kg. The security scanners using ionising radiation that are commercially available utilise X-rays with 50 kVp to 220 kVp (with some additional filtering) which have a nominal radiation quality factor wR = 1.
One can distinguish between doses determined for specific persons (personal dose) and doses measured or assessed at specific locations (ambient dose).
22.214.171.124 Organ doses
First of all, in most applications of X-rays on humans, in circumstances of non uniform radiation as for example a chest X-ray, the equivalent dose to each organ might be different. As most epidemiological data refer to studies of external radiation exposures with quite high energies in large homogeneous fields, in these investigations one can assume a uniform dosage to the whole body. The security scanners, at the low energies of the ionising radiation used, will result in different doses to different organs. As the energy imparted decreases, so does the penetration and therefore the differences between the various organ doses is greater. There may even be differences within single organs. One assumes that the risk related to the dose in the same tissue is described by the average energy imparted multiplied by the radiation quality factor in the specific organ. Therefore the organ doses are given by the average of the equivalent dose over the whole organ HT. These averages have to be determined for all organs.
HT,R = wR * DT,R
126.96.36.199 Effective doses
Large epidemiological studies on the risk of ionising radiation, especially the life span study of the atomic bomb survivors of Hiroshima and Nagasaki, have shown that different organs show a different risk of stochastic effects like cancer development caused by ionising radiation (see Epidemiology section 3.6.3.). Based on morbidity and mortality data on the survivors and their basically uniform irradiation, specific risk coefficients have been determined for various organs. Assuming that the sum of the potential risks for all individual organs should represent the total risk of the total body irradiation then resulted in the approach of the effective dose E, where the risk coefficients are transferred to tissue weighting factors wT for organs. By multiplying these risk factors with the corresponding equivalent organ doses and summing up the resulting weighted organ doses, one gets a dose describing a probability of health detriment comparable to a total body dose. Effective dose is defined as:
E = ΣTwT * HT,R
This dose value is not intended for the determination of a risk for an individual but is only an estimate of the average risk in a population even though the risk for an individual may vary due to age at exposure, gender or other risk factors. The tissue weighting factors are listed in various ICRP (International Commission on Radiological Protection) publications. According to the actual determination of the ICRP, the risk factors are as tabulated in Table 1
(from ICRP publication 103 (ICRP 2007)). Dose limits in legislation are expressed in effective doses and equivalent doses.
188.8.131.52 Specific doses
Regarding the use of the various scanners, doses to the skin and some other organs are of specific interest because of the non-homogeneous exposure due to irradiation geometry and the low energy of the radiation involved.
For the determination of organ doses some measurements are required. For risk assessment, equivalent doses are assessed. They are typically evaluated either as H*(10), which describes a personal equivalent dose measured at 10 mm depth of a reference sphere consisting of soft tissue equivalent according to the ICRU. For a close representation of the skin dose, H*(0.07) is used, which represents the equivalent dose value at a depth of 70 µm.
In the backscatter systems with relatively low photon energies (low radiation beam qualities) the organs close to the body surface such as the lens of the eye, the female breast or the testes will receive higher doses than organs deeper in the body. In systems using higher beam qualities (higher tube voltages, harder filtration) the dose distribution within the bodies would be more uniform.
3.5.2 Dose determination
Since it is difficult to measure doses of ionising radiation directly within the body, organ doses are typically evaluated by measuring doses on representative areas and then performing simulations using models of the human body. These simulations typically provide conversion factors to obtain organ doses from the measured entrance dose values. Historically, the first simulations were done on simple geometric mathematical phantoms. This has been the case for some studies already published; other studies provide dose measurements. These measurements are summarized in Table 2.
Voxel phantoms with realistic anatomy were then produced. The new standard reference phantoms representing the standard man and the standard woman were introduced in ICRP publication 110 (ICRP 2009). Some conversion factors have already been determined for these new reference phantoms. Exposure of an average person in the context of a security scanner can be simulated. These simulations certainly do not take into account the variations between different persons. The determination of organ and effective doses for the average person is sufficiently accurate in view of the inherent uncertainty related to the low doses typical for security scanners.
The Monte Carlo simulations were performed on the new ICRP standard voxel phantoms and on the model for pregnant women. For the calculations, some simplifications about the geometry of the scanning process have been made. These should be conservative and of minor importance for the resulting effective and relevant (important organs and those with higher doses compared to other organs) organ doses. The determined values (always for two sided (AP/PA) scans) are summarized in Table 3. The complete table can be found in the appendix.
The radiation dose from a single passenger being scanned is approximately equivalent to natural background radiation [Query – Judy Burns] received within an hour on the ground or during 10 minutes of flight at a typical cruising altitude (30,000-35,000 feet).
It should be noted that an effective dose for the pregnant woman model is not really meaningful but given to allow a certain possibility of comparison. The dose values presented here are in the same range as those presented by most other publications.
Different studies have shown consistent results in terms of measured radiation doses for similar equipment. Furthermore, agreement between measured doses and doses calculated through simulations is good. However, a recent paper (Rez et al. 2011) estimated doses to the skin as high as 2.5 μGy for 50 kVp X-rays and 0.68 μGy for 50 kVp X-rays (effective doses of 0.9 and 0.8 µSv, respectively). Those results were, however, based on an approach different from the other studies (number of quanta needed for achieving image quality properties, with uncertain assumptions on geometry and signal-to-noise ratio likely to heavily influence the results). The Working Group concluded that the mainstream of the empirical studies is more likely to provide accurate dose estimates than the single outlier.
One should state here that it is very difficult to give reliable and meaningful estimates of effective doses for children up to the age of at least 14 years, since the variations in stature and size are even greater than for adults. In addition, there is not yet a new family of reference children. Besides which, the effect of geometric proportions of the child to the scanner and the mode of use of the scanning would result in large variations. It should be reasonable to assume that the effective doses would be in the same order of magnitude as those of the adults.
3.5.3 Special groups
There is some variability in the effective dose from body scanners between individuals in relation to their physical characteristics (body size and gender). Therefore, the calculated dose values only indicate an average of doses due to the use of security scanners. The range of doses to an adult may vary by up to a factor of two. Groups likely to be scanned frequently include frequent flyers, couriers, air crews and airport staff. To assess the maximal plausible dose from security scanners, someone flying each working day of the year with several connecting flights might be scanned three times daily with a total of up to 720 times annually. The cumulative effective dose from a backscatter scanner would thus amount to roughly 300 μSv (assuming a dose of 0.4 μSv per scan, i.e. higher than the typical values estimated). If all scans were performed using transmission technology (assuming an effective dose of 4 μSv per scan) the corresponding annual cumulative effective dose would be one order of magnitude higher, approaching 3,000 μSv or 3 mSv. This would clearly exceed the dose limit for the general public (which is applicable to the passengers, but also to other frequently scanned groups such as airline crews, airport personnel, etc.). The principle of dose limitation would therefore indicate a preference for backscatter technology, unless the capacity to detect objects within the body is deemed crucial. Sensitivity (susceptibility to harmful effects) varies also within the population in relation to age, sex and other factors. Potentially sensitive groups within the population include pregnant women (foetuses) and children. This is addressed in section 3.6.3.