External dose reconstruction in tooth enamel of Techa riverside residents

Abstract

This study summarizes the 20-year efforts for dose reconstruction in tooth enamel of the Techa riverside residents exposed to ionizing radiation as a result of radionuclide releases into the river in 1949–1956. It represents the first combined analysis of all the data available on EPR dosimetry with teeth of permanent residents of the Techa riverside territory. Results of electron paramagnetic resonance (EPR) measurements of 302 teeth donated by 173 individuals living permanently in Techa riverside settlements over the period of 1950–1952 were analyzed. These people were residents of villages located at the free-flowing river stream or at the banks of stagnant reservoirs such as ponds or blind river forks. Cumulative absorbed doses measured using EPR are from several sources of exposure, viz., background radiation, internal exposure due to bone-seeking radionuclides (⁸⁹Sr, ⁹⁰Sr/⁹⁰Y), internal exposure due to ¹³⁷Cs/^137mBa incorporated in soft tissues, and anthropogenic external exposure. The purpose of the present study was to evaluate the contribution of different sources of enamel exposure and to deduce external doses to be used for validation of the Techa River Dosimetry System (TRDS). Since various EPR methods were used, harmonization of these methods was critical. Overall, the mean cumulative background dose was found to be 63 ± 47 mGy; cumulative internal doses due to ⁸⁹Sr and ⁹⁰Sr/⁹⁰Y were within the range of 10–110 mGy; cumulative internal doses due to ¹³⁷Cs/^137mBa depend on the distance from the site of releases and varied from 1 mGy up to 90 mGy; mean external doses were maximum for settlements located at the banks of stagnant reservoirs (~500 mGy); in contrast, external doses for settlements located along the free-flowing river stream did not exceed 160 mGy and decreased downstream with increasing distance from the site of release. External enamel doses calculated using the TRDS code and derived from the EPR measurements were found to be in good agreement.

Appendices

Appendix 1: Method-specific performances

Performance parameters as indicators of method-specific dose evaluation quality are defined as follows:

• Critical level (critical value, detection decision) is a threshold below which there is practically no chance to distinguish a radiation-induced signal from a blank response (Eq. 10). The critical level for dose evaluation is called the critical dose (D _c);

  $$ P(\hat{D} > D_{\text{c}} \left| {D = 0) \le 0.05} \right. $$

  (10)

• Detection limit (limit of detection, minimal detectable amount) is a threshold above which a measurement can be definitely interpreted as a response to radiation (Eq. 11). The limit of dose evaluation is indicated as LD;

  $$ P(\hat{D} \le D_{\text{c}} \left| {D = {\text{LD}}) = 0.05} \right. $$

  (11)

• Uncertainty is a non-negative parameter characterizing the dispersion of the quantity values being attributed to a measurand (including both stochastic and residual errors).

Table 8 in Appendix 1 presents the method-specific performance parameters and measurement statistics. The uncertainties in Table 8 in Appendix 1 are given in terms of standard uncertainty (equal to 1σ), where the uncertainty considered both stochastic error of measurement and uncertainty of evaluation of the method-specific bias. The uncertainty of the method-specific bias was considered as a residual systematic error, which was treated similarly as a random one.

Table 8 Method-specific performance parameters and measurement statistics (uncertainties are valid for the dose range from LD to 2 Gydecision)

For the methods indicated by index a, raw data of the calibration experiments were available for a direct evaluation of the method performances using the software “EPR dosimetry performance” (Shishkina et al. 2014a). The software returns the dose dependence of uncertainty as a table. The tabular data were smoothed (with software SigmPlot 12.5) converting the discreet data into a continuous analytical function. The analytical description of the time-dependent performances of the EPR dosimetry in the IMP (indicated by b in Table 8) was evaluated based on the analysis of residual spectral noise depending on time (presented in Ivanov et al. 2011). For the other methods, the performances were assigned based on the analysis of the results of inter-laboratory comparisons (indicated by c) or expert decisions (indicated by d) as it was described in Volchkova et al. (2011). A method, which was found to be similar to the method a or b in terms of repeatability, was assumed to be characterized by the same dose dependence of uncertainty. Therefore, Table 8 presents the analytical expressions obtained for the method-specific uncertainties. The examples of an expert decision and of the use of the intercomparison results are presented below.

Example 1

Expert decision about method-specific uncertainty for EPR measurements by ISS using labial and lingual fractions of the incisor enamel prepared by HMGU (methods 28 and 36, respectively).

Various fractions of the incisor enamel (labial or lingual) as well as the whole enamel of the posterior teeth were prepared by HMGU to be measured by ISS. The uncertainty of the method for measurements of the posterior teeth (method 12) was estimated (Eq. 12) using the software “EPR dosimetry performance” based on the raw data of the calibration experiment.

$$ g\left( D \right) = 29.9 + 25.3 (1 - e^{ - 0.006D} ) $$

(12)

The dose bias, C, of method 12 was evaluated as equal to 52 ± 8 mGy (mean ± SE). The overall uncertainty, U(D), was calculated as a square root of the sum of squares of the method-specific uncertainty, g(D), and the standard error of the dose bias, C, (Eq. 13).

$$ {\text{method }}12: \, U\left( D \right) = 31 + 28 (1 - e^{ - 0.006D} ) $$

(13)

The same sources of uncertainties were assumed for the methods 28 and 36 corresponding to the labial and lingual fractions of the incisor enamel (the biases were evaluated as equal to 430 ± 34 mGy and 280 ± 26 mGy, respectively). In this way, the overall method-specific uncertainties were formulated as follows (Eqs. 14, 15):The same sources of uncertainties were assumed for the methods 28 and 36 corresponding to the labial and lingual fractions of the incisor enamel (the biases were evaluated as equal to 430 ± 34 mGy and 280 ± 26 mGy, respectively). In this way, the overall method-specific uncertainties were formulated as follows (Eqs. 14, 15):

$$ {\text{method }}28: \, U\left( D \right) = 43 + 38 (1 - e^{ - 0.006D} ) $$

(14)

$$ {\text{method }}36: \, U\left( D \right) = 39 + 34 (1 - e^{ - 0.006D} ) $$

(15)

Example 2

The use of the results of intercomparison between IMP, IBP and ICP (methods 1, 15, and 16) to estimate the method-specific uncertainties for methods 15 and 16.

The first intercomparison of the EPR methods involved in the dose reconstruction in the Urals was performed by three EPR laboratories in 1993. The results were reported by Kleschenko et al. (1994). Each participant (IMP, IBP and ICP) prepared a mixture of grains from the enamel of several teeth using their own sample-preparation techniques (methods 1, 15, and 16). Each of the mixtures was divided into portions. All portions were exposed to four doses within the range 0–500 mGy.

No statistically significant systematic differences were found between the methods. The repeatability (standard deviation of three repeated measurements) of the results obtained with methods 15 and 16 was by 5–15% better than that obtained using method 1. Since a limited number of measurements was available (four samples per method), the uncertainty of the method with the worse repeatability (method 1) was extrapolated to the other two methods (to avoid underestimation). Therefore, the uncertainties of the methods 1, 15 and 16 were described by the same Eq. (16), which was obtained by propagation of uncertainty of the 1st method (Ivanov et al. 2011) and uncertainty of the bias, C = 140 ± 20 mGy, for method 1.

$$ {\text{methods }}1,\;15{\text{ and }}16: \, U\left( D \right) = 31 + 6.5 10^{ - 2} D + 10^{ - 5} D^{2} $$

(16)

Appendix 2: Dose coefficients

Dose coefficients (DCs) represent dose rates in tooth enamel per unit of ⁹⁰Sr/⁹⁰Y activity concentration in a source-tissue (enamel, primary/secondary dentin fraction or a part of the root with 3 mm thickness, which is adjacent to the tooth crown). For teeth which had been completed at the time of intakes, the secondary dentin was considered as a source. For teeth with roots under formation, the primary dentin was considered as a source. The dose coefficients are the factors to be multiplied to the ⁹⁰Sr/⁹⁰Y activity concentration in the source. In the routine practice, it is impossible to separate different dentin fractions for measurements of activity concentrations. The measurements of activity concentrations were therefore performed using the whole dentine volume. Assuming a uniform ⁹⁰Sr distribution in one of the tissue fractions, and absence of the radionuclides in the other one (see Supplemental materials), the mass ratio of the whole dentin and its emitting fraction was used as a correction factor to the dose coefficients (Eq. 17).

$$ \dot{D} = A_{i} \times {\text{DC}}_{i}^{/} = \bar{A} \times \frac{{m_{d} }}{{m_{2} }} \times {\text{DC}}_{i}^{/} = \bar{A} \times {\text{DC}}_{i} , $$

(17)

where $\dot{D}$ is the dose rate in the enamel; ${\text{DC}}_{i}^{/}$ is the result of direct Monte Carlo simulations of enamel dose rate per unit of ⁹⁰Sr/⁹⁰Y activity concentration in the dentin fraction i; ${\text{DC}}_{i}$ is the corrected dose coefficient. In the paper and hereinafter, the term “dose coefficient” assumes the corrected dose coefficient.

DCs of the anterior teeth were calculated for three “detectors”, viz. whole enamel, lingual fraction and labial fraction. DCs due to root dentin contamination were evaluated to be 0.09 (mGy/year) per (Bq/g) and 0.006 (mGy/year) per (Bq/g) for anterior and posterior teeth, respectively. Incisor’s DCs are dependent on the height (H) of tooth crown (Table 9) in Appendix 2.

Table 9 Dose coefficients (DC) for anterior teeth and their relative uncertainties in brackets, mGy/year per Bq/g (relative units)

The height of incisors, H, (being affected by attrition) depends on the age of an individual. The population under study includes individuals born in 1913–1943. In this period, people lived during the time of the first and the second world wars, changing of the social system, collectivization and food requisition, punctuated by periods of relative prosperity. Thus, the resulting differences in the diets of people with different years of birth could influence the hardness of enamel and, as a result, the rate of incisor attrition. According to Volchkova and Shishkina (2012), H of the incisors of rural residents in the Urals was found to be dependent on both human age (t) and birth year (T). Eq. (18) expresses this dependence for erupted teeth.

$$ \left\{ {\begin{array}{*{20}l} {H\left( {t,T} \right) = \frac{a\left( T \right)}{{1 + e^{{ - \frac{{t - t_{0} \left( T \right)}}{b\left( T \right)}}} }}} \hfill \\ {a\left( T \right) = c + d \cdot e^{{ - e^{{ - \frac{T - r}{f}}} }} ,} \hfill \\ {b\left( T \right) = g + h \cdot e^{{ - e^{{ - \frac{T - m}{n}}} }} } \hfill \\ {t_{0} \left( b \right) = o \cdot b + w} \hfill \\ \end{array} } \right. $$

(18)

where c, d, r, f, g, h, m, n, o and w are position-specific constants presented in Table 10.

Table 10 Position-specific constants c, d, r, f, g, h, m, n, o, w used in Eq. (18)

DCs of posterior teeth were calculated considering whole enamel as a “detector”. Impact of attrition on tooth height (and, therefore, on DCs) was found to be insignificant and was ignored (Volchkova and Shishkina 2012). However, a change in the volume, V, of secondary dentin, which is growing inwards over the lifetime since the tooth eruption, influences the DC value. Equation (19) presents the age dependence of the secondary dentine volume.

$$ {\text{for}}\;t > T_{er} ;\;\;V\left( t \right) = \frac{0.011k}{{1 + e^{{ - \frac{t - 43}{9.66}}} }}, $$

(19)

where t is a person’s age; T _er is a position-specific age of tooth eruption; and k is a position-specific fitting parameter.

Table 11 in Appendix 2 presents DCs of posterior teeth. The table also includes the parameters (T _er and k) used for calculation of V of the secondary dentin, according to Eq. (19).

Table 11 Dose coefficients (DC) for posterior teeth and their relative uncertainties in brackets, mGy/year per Bq/g (relative units) as well as the parameters used for calculation of the volume V of the secondary dentin (T _er and k) (Eq. 19)