Radiation Measurements: PDF for review
COMPLEX EXPERIMENTAL RESEARCH ON INTERNAL TOOTH
DOSIMETRY FOR TECHA RIVER REGION: VALIDATION OF
MODEL FOR 90SR IN THE TEETH FORMED BY TIME OF INTAKES
08 Jun 05
Name Fig No FormatUse Description
daily release during this period was 1.6 1014 Bq with the following contribution of different radionuclides to the total
COMPLEX EXPERIMENTAL RESEARCH ON INTERNAL TOOTH DOSIMETRY FOR THE TECHA RIVER REGION: A
MODEL FOR 90SR ACCUMULATION IN HUMAN TEETH FORMED BY TIME OF INTAKES
D.D. Tikunov*1, A.I. Ivannikov*, E.A. Shishkina†, D.V. Petin*, N.B. Borysheva*, S. Orlenko*, M. Nalapko*, V.A.
Shved†,V.G. Skvortsov*, V.F. Stepanenko*
* Medical Radiological Research Center, Korolyov str., 4, Obninsk 249036, Russia
†Urals Research Center for Radiation Medicine, Medgorodok, Chelyabinsk 454076, Russia
Abstract – Samples of calcified tooth tissues (enamel, root and crown dentine) collected from the Techa riverside
population exposed to radiation caused by radioactive releases from the nuclear weapon plant in South Ural were
investigated. Accumulated absorbed dose in the samples was measured using the EPR-spectroscopy method. Beta
activity of the samples containing radioactive 90Sr was measured by the method of low background anti-coincidence
thin scintillating detection. High correlation between absorbed dose and beta activity was observed for enamel and
root dentin but not for crown dentin. Otherwise, poor correlation was observed between absorbed doses as well as
between beta activities for different tooth tissues of the same tooth. The results of dose measurement by EPR-
spectroscopy are analysed with the use of Monte Carlo simulation of dose formation due to 90Sr incorporated in
tooth tissues taking into account biological elimination of 90Sr. Influence of 90Sr distribution in the tooth body on
absorbed dose is discussed.
Keywords: internal dosimetry, external dosimetry, EPR/ESR, tooth enamel.
Vast territories of South Ural region were contaminated as a result of radioactive wastes released by the
Mayak Production Association into the Techa River in 1949-1956. During this period 76⋅106m3of liquid wastes
with total activity of about 1017 Bq were discharged into the river. The radioecological conditions and dosimetric
investigations in this region are described elsewhere (Marey, 1959; Stepanenko et al., 1990; Degteva et al., 1994;
Vorobiova et al., 1999; Tolstykh et al., 2003).
About 98% of total activity was released in the period from March 1950 to October 1951. The average
1Corresponding author. Tel/Fax: (095) 956-14-40, e-mail: firstname.lastname@example.org
1 of 27
Thursday , December 08, 2005
90Sr (Ignatiev et al., 1999; Shishkina et al., 2001). Good agreement was found for doses modelled using Monte-
activity: 89Sr – 8.8%, 90Sr – 11.6%, 137Cs – 12.2%, 95Zr-95Nb – 13.6%, 103,106Ru – 25.9%, and rare-earth elements –
26.8% (Vorobiova et al., 1999). Exposure of the population (about 30,000 persons in the Techa riverside) was
caused mainly by long-lived isotopes of 90Sr (half-life 29 years) and 137Cs (half-life 30 years). Internal irradiation
was resulted primarily from bone seeking beta-emitting 90Sr. External irradiation was caused primarily by 137Cs
(Degteva et al., 1994). External doses estimated using the Techa River Dosimetry System 2000 (TRDS-2000)
(Degteva et al, 2000) expected to vary in the range from 0 to 3 Gy depending on residence history of individuals.
The maximal doses of external exposure were estimated for residents of the Upper Techa region who lived in
immediate proximity to the shoreline.
Estimated external doses could be validated by the method of EPR-dosimetry. This method gives
possibility to determine a dose accumulated in calcified tooth tissues (IAEA 2002). For the population of the Techa
riverside, a dose accumulated in tooth tissues is a result of both external and internal exposure. Therefore, in order to
obtain external dose from EPR dosimetry data, the total accumulated dose detected by EPR should be corrected on
internal dose contribution taking into account the presence of 90Sr in the tooth tissues.
Several pilot studies of the teeth collected from population of the Techa riverside were aimed to find out
the relationship between 90Sr content in calcified tissues and cumulative dose (Tolstykh et al., 2000; Romanyukha et
al., 1996, 2001; Wieser et al., 1996; Göksu et al., 2002). All these studies demonstrate significant influence of 90Sr
on cumulative dose in enamel that cannot be ignored in the framework of the Techa River dosimetry. In Tolstykh et
al. (2000) internal dose formation was investigated for enamel of teeth that crown formation had been completed by
the time of radioactive release in assumptions that the amount of 90Sr incorporated in the enamel is negligible and
the main source of internal dose should be the crown dentin; 90Sr was assumed to be uniformly distributed in the
whole dentin body (root and crown dentine). For Metlino residents (Upper Techa) the results of individual
measurements using EPR were found to be consistent with the doses calculated according to TRDS-2000 approach
(Degteva et al., in press). In Göksu et al. (2000) accumulated dose in tooth tissues was compared with 90Sr content in
the same tissues. High correlation was observed for absorbed dose in enamel, root and crown dentin with beta dose
rate from incorporated 90Sr. No correlation was observed for beta dose rate in enamel and dentine as well as between
absorbed dose in these tissues of the same tooth. For the population of the Techa region it was found that maximal
doses in tooth tissues measured by EPR are observed for tissues, whose period of formation coincided with maximal
releases of radioactivity (Romanyukha et al., 2000a; Romanyukha et al., 2001; Tolstykh et al., 2003). The process of
dose formation in dentine and enamel was analysed for dog's teeth after a short period of 52 hours after injection of
Carlo method and experimental doses measured by EPR-spectroscopy.
2 of 27
Thursday , December 08, 2005
as described in Skvortsov et al. (1995). A dosimetric radiation-induced signal was obtained by subtraction of a
The objectives of the present work are the following: to investigate 90Sr distribution and absorbed dose
after 50 years of intakes in various calcified tooth tissues which were in the process of maturation by the time of
radioactive releases into the Techa River; to analyse the effect of 90Sr incorporated in tooth tissues on the absorbed
dose determined by EPR-spectroscopy. Teeth obtained from the residents of the Middle and Low Techa regions
were used for this study.
MATHERIALS AND METHODS
Sample collection and preparation
Teeth obtained from the Techa riverside donors were selected specifically for this study. Teeth were
extracted according to medical indications in the frame of creation of URCRM tooth sample bank. Part of samples
was collected from control uncontaminated localities of the South Ural region. Totally, 8 teeth of exposed people
(22 samples of tooth tissues) and 2 teeth from controls (6 samples of tooth tissues) were studied. For these teeth the
calcified tissues were at different formation stages of at the time of radioactive releases (1950-1951).
Crown dentin and enamel samples were prepared by sawing off the surface enamel layer of the tooth by a
diamond disk saw. Root dentin was prepared from the root sawed off from the crown. Dental cement was removed
from the root by dental drills. In order to eliminate effects of mechanically induced radicals due to mechanical
treatment, the sawing was performed at low linear speed under water-cooling. Dentine and enamel fragments were
crashed by nippers to chips of 0.5 – 2.0 mm in size. After mechanical treatment, the samples were etched 2-3 min in
20% acetic acid at room temperature, then rinsed up by distilled water, ethanol and dried.
EPR spectra measurement and spectra processing
EPR spectra measurement was performed at room temperature with the use of ESP-300E (Bruker,
Germany) X-band spectrometer equipped with high sensitivity TM110 mode cavity ER-4108-TMH. The following
registration parameters were taken: microwave power 10 mW; modulation 0.3 mT at 100 kHz; sweep width 8 mT;
sweep time 165 s; receiver time constant 163 ms; number of accumulations 16. Fourth line of an Mn2+ containing
sample permanently mounted near the bottom of the cavity was used as a reference amplitude signal. Aliquots of
100 - 130 mg sample mass were used for measurements.
Spectra processing was performed with the use of the basic software of the built-in spectrometer computer
simulated native signal, which was obtained from a standard spectrum recorded for enamel of milky teeth. Shift,
3 of 27
Thursday , December 08, 2005
Dose determination in dentin
amplitude and width of the standard spectrum were changed using control knobs of the computer to fit a low-field
part of the recorded spectrum where the native signal is positioned. The same standard spectrum was used for
simulation of the native signal both in enamel and dentin. The intensity of dosimetric signal was determined as
amplitude of its maximum. This methodical approach was tested at 2-nd International Intercomparison on EPR tooth
dosimetry (Wieser et al., 2000), and an accuracy of dose determination in enamel was characterized by mean
standard deviation of experimental doses from nominal doses as 32 mGy in dose range up to 800 mGy.
Irradiation of dentin and enamel samples was performed by 60Co gamma-ray source. Samples at irradiation
were placed between two 4-mm polymethylmethacrylate plates. At such conditions of irradiation, dose absorbed in
enamel and dentin is very close to dose absorbed in soft biological tissue (Wieser et al., 2000; Ivannikov et al.,
2004). Therefore, dose values obtained with the use of tissue equivalent dosimeter (with standard uncertainty of 3%)
were interpreted as doses absorbed in the samples.
Dose determination in enamel
Absorbed dose in enamel was determined from the amplitude of radiation-induced EPR signal normalized
by sample mass and marker signal amplitude with the use of universal calibration coefficient obtained at irradiation
of pooled enamel by 60Co gamma-rays. Standard variation of individual enamel radiation sensitivity (the ratio of
EPR signal to dose absorbed in enamel) was assumed to be 10% (Tikunov et al., 1996; Wieser et al., 2001). This
variation gives appropriate relative contribution to the uncertainty of dose determination in enamel at the use of
universal calibration coefficient, which was taken into account at error estimation. For enamel grain size used,
contribution to the uncertainty of radiation-induced signal amplitude due to sample anisotropy was accepted as not
exceeding 5% (Iwasaki et al., 1995). The uncertainty of dose determination in enamel (Ue, in Gy) was estimated
using an expression analogous to that derived in Ivannikov et al. (2002), which comprises constant, variable
depending on sample quality and dose dependent contributions:
here: Ue1 = 40 mGy – constant contribution to the uncertainty resulting from an error in spectra processing caused
by variation in the native signal line shape of individual samples, anisotropy of the native signal and an error in the
intercept of calibration line; De–experimental dose; ke1 = 0.12 – relative contribution resulting from the variation of
individual radiation sensitivity (10%), sample anisotropy (5%) and uncertainty of the slope of calibration line (3%).
Since variation of radiation sensitivity of dentin was expected to be significant, dose determination was
performed with the use of additive dose method. Samples were irradiated in dose 5 Gy, 10 Gy and 20 Gy. After
4 of 27
Thursday , December 08, 2005
An uncertainty of the specific activity determination (UN, in Bq g-1) was determined using the equation:
additional irradiation, the samples were aged for about 10 days before being measured to eliminate transient signals.
Dose in dentin was determined from an abscissa intersection of the regression line of the radiation-induced EPR
signal amplitude versus the irradiation dose. An uncertainty of dose determination in dentin was estimated taking
into account the uncertainties of regression analysis and uncertainty of reference dosimetry.
Detection of 90Sr concentration
Low background beta-counting measurement
Beta activity was measured by a nuclear information processor SNIP 204G (Silena, Italy) supplied with a
low-background beta counting set. It consists of the following components: plastic scintillation beta detector S2P;
plastic scintillation anti-coincidence detector S4P for the detection of coincidence gamma radiation; background led
shield with sample holder drawer; two analogue electronic units HPA/B, each accommodating the signal
amplification circuitry and the high voltage power supplier for the photo-multiplier.
The effectiveness of registration of the beta counter was determined using a set of 90Sr/90Y reference
sources with an overall activity ranged from 3.9 up to 82 Bq. The effectiveness of registration was found to be
0.350 ± 0.012 counts s-1 Bq-1.
The background count rate without sample was estimated as followed. The background measurement was
repeated 130 times by 10 min. The mean value of count rate was found to be 6.4 ± 0.1 min-1. For the average value
obtained after six time 10-min measurements, standard deviation of background count rate was estimated as 0.4
min-1, which corresponds to 0.02 Bq taking into account estimated value of effectiveness of registration. This value
was subtracted from measuring activity of the samples. Contribution of natural radionuclides and radionuclides from
global fallouts to the measured beta activity of tooth tissues was estimated using tooth samples of donors resident in
the control localities.
Determination of the total beta activity in tooth tissues
For activity measurement, tooth tissue samples were additionally powdered to have grain size of about
0.1 mm. The measurements were performed six times by 10 min and an average was taken.
In order to estimate the contribution of sample geometry and sample mass into the error of activity
determination, the activity measurements were performed for different aliquots of one of the dentin samples. For the
aliquots with the masses of 0.11, 0.15, 0.21, 0.25 and 0.30 g and experimental activities ranged from 11 to 27 Bq,
the count rate versus the sample mass was found to be linear within 3%.
5 of 27
Thursday , December 08, 2005
parameters of teeth typical for Ural population were used in the refined tooth model. The thickness of the lateral
here: UA1 = 0.02 Bq – constant contribution due to uncertainty of background count rate; UA2 – uncertainty of an
average of six measurements; A –measured activity (in Bq); kA= 0.05 – coefficient, contribution to which is given
by uncertainties of effectiveness of registration at calibration (3.5%) and linearity of mass response (3%); km=
0.01 g – uncertainty of sample mass; m –sample mass (in g).
Electron transport simulation
Absorbed doses in enamel and dentin were calculated by Monte Carlo simulation of electron transport
using the MCNP-4C code (Briemeister, 2000). The coupled electron-photon transport physics in the MCNP-4C code
take into account in a rather accurate way the diffusion and slowing down of all radiations in the electron-photon
cascade established in the media. The most important theory for the algorithms in MCNP is the Goudsmit-
Saunderson theory. A continuous-slowing-down model is used for electron transport that includes positrons, x-rays,
and bremsstrahlung, but does not include external or self-induced fields. The geometry of MCNP treats an arbitrary
three-dimensional configuration of user-defined materials in geometric cells bounded by first- and second-degree
The default mode of detailed physics for electron transport was used with the 1.0 keV energy cut-off. The
energy deposited in the probe cells was calculated using the pulse height tally, which records the energy deposited in
a cell by each source particle and its secondary particles. The calculated deposited energy was then normalized by
the cell mass to give the absorbed dose.
The initial approach to the problem of tooth-geometry description involved an extreme simplification of
tooth shape. Fig. 1a demonstrates simple geometric approximations for posterior teeth. The dimensions of the
dentin cylinder were associated with representative dimensions of crown dentin in various posterior tooth positions.
In this case a cylindrical shell (the lateral enamel) is an annual ring. The dentin and the lateral enamel form a single
whole cylinder. The cylinder head (masticatory enamel) and its base (root dentin) are prisms.
A more accurate model was constructed for the first bottom molar of permanent teeth. This geometric
model is presented in Fig. 1b. The shape of the crown, crown dentin, pulp, and the fragment of root dentin adjacent
to the cervical junction can be described as a rectangle with rounded corners. Actual values of the average geometric
enamel layer increases from the cervical junction to the height of the crown. For the description of the lateral enamel
shape, the intersection of a truncated cone and a truncated pyramid turned small base down is used. The shape of
6 of 27
Thursday , December 08, 2005
the beta count rate detected for the tissues of control teeth. Only 2 control teeth (6 samples of tooth tissues) were
the masticatory enamel arises from the intersection of a parallelepiped and a prism with an elliptical base. A
parabolic equation describes the upper surface of the masticatory enamel.
The absorbed dose was scored in the dentin and enamel separately. The different source tissues (enamel,
crown dentin including its primary and secondary fractions, root dentin) were examined as dose forming factors.
Uniform, isotropic sources of electron emission from 90Sr/90Y decay were assumed in the each specific source
volume. The density of each tooth tissue was considered as a constant. Element composition and density were
taken for moist weight according to Tolstykh et al. (2000). All results are based on a sample of 10M histories of
emitted beta particles.
RESULTS AND DISSCUSSION
Experimental estimation of absorbed dose and 90Sr concentration in tooth tissues
Results of absorbed dose and total beta activity determination in tooth tissues are presented in Table 1.
Personal code and sample numeration are given according to URCRM register. Samples 89 – 250 were collected in
the Techa region, samples 276 and 277 in control territories.
It should be noted that in contrast to tooth enamel, dosimetric properties of dentin are poorly investigated.
Based on the measurements performed in this work, it was found that radiation sensitivity of different dentin
samples varies with standard deviation of 35%. Average dose response of dentin is approximately five times lower
than that of enamel. Similar ratio was found for dog's teeth (Ignatiev et al., 1999). No significant difference was
observed for the average dose response of crown and root dentin.
The cumulative absorbed dose for Techa River residents detected by EPR represents the superposition of
three main types of exposure, viz.: background, external and internal irradiation. Contribution of the natural
background radiation to absorbed dose in enamel was assumed to be 1.0 mGy y-1 (Ivannikov et al., 1997) and it was
subtracted from cumulative dose in tooth enamel. It should be noted that cumulative doses detected in enamel of
Ural control teeth (276 and 277) do not contradict to this estimation. The contribution of background radiation to
absorbed dose in dentin was not investigated before. Therefore, for this study as a preliminary result the average
dose in dentine of control teeth detected by EPR have been used. The background dose level in the dentin of
investigated control samples was estimated as 0.77±0.15 Gy. This rough estimation was used at dose determination
The results of beta counting were converted into 90Sr concentration in the tooth tissues taken into account
tested for determination of radionuclide contamination. Average total activity of all investigated tooth tissues from
7 of 27
Thursday , December 08, 2005
teeth was found in the value interval from 0.09 up to 9.3 Bq g-1. From three teeth, which related to regions 4 and 5 in
unexposed donors was estimated as 1.3±0.4 Bq g-1. This value was subtracted from the total beta activity measured
for exposed donors to determine 90Sr/90Y specific activity (N). Since the 90Sr/90Y is assumed to be in secular
equilibrium, the concentration of 90Sr is equal to 0.4985 N.
It should be noted, that the background doses cannot be estimated individually and the values used
represent the expected values that are more probable for the investigated samples. Therefore, after subtraction of
natural background dose, the experimental estimations and uncertainties of cumulative external and internal dose (as
far as its current 90Sr content) for the Techa River residents can be considered as mathematical expectation and
corresponding confidential bounds (Table 2).
Dependencies of accumulated dose on 90Sr concentration in tooth tissues together with linear regression
parameters are presented in Fig. 2. High correlations are observed for enamel and root dentin. The crown dentin
demonstrates a weak conformity between accumulated dose and 90Sr content. Five samples of crown dentin that
have the variation of 90Sr concentration over the range from 3 to 22 Bq g-1, demonstrate dose independence on the
90Sr activity because theirs cumulative doses are practically equal (about 2 Gy).
Dependence of 90Sr concentration on the time of intake
The both concentration of 90Sr in the tooth tissues and the radionuclide distribution in the tooth body should
strongly depend on mineralization status of tooth at the time of radioactive releases. The maturation processes in
root dentin, crown dentin and enamel are very different (Tolstykh et al., 2000; Romanyukha et al., 2001; Göksu et
al., 2002). In the case of coincidence of the periods of 90Sr intake and tissue mineralization, the concentration of
90Sr in enamel is high. Table 3 shows the years of birth for the Techa River residents for whom the highest
concentrations of 90Sr (and, accordingly, the highest cumulative doses) in tooth tissues at various positions should be
expected (based on the data of tooth mineralization presented in Logan and Kronfeld (1933); Lewis and Garn
(1960); Novik (1971) and Kolesov (1991)). It should be noted that individual variations in tooth-development times
are significant, and these calculations represent average values.
Fig. 3 demonstrates the concentration of 90Sr in the investigated samples depending on the mineralization
status. The results obtained generally correspond to the conception of the maximal radionuclide accumulation at the
time of tissue calcification at high rate. Although the reliable detected levels of 90Sr concentration were found in 4
out of 7 samples of enamel completed by 1950-1951. As seen in the Fig. 3, those were the teeth with crown dentin at
different stages of formation (regions 2 and 3 in Fig. 3a). The estimated 90Sr concentration in the enamel for such
Fig. 3a (formation of root dentin and beginning of secondary dentin formation), two enamel samples have
8 of 27
Thursday , December 08, 2005
insignificant content of 90Sr and one demonstrates a concentration level not exceeding 0.5 Bq g-1(expected value is
0.2 Bq g-1).
Two teeth with different positions belonging to the same person (IDP 66951) demonstrate practically equal
enamel doses and 90Sr concentrations. However, for the same teeth the high differences between both crown and
root dentins were found. Both of these teeth can be related to the period of completion of primary crown dentin
mineralization and beginning of root dentin formation in terms of their mineralization status in 1950-1951 (see
Fig. 3). Earlier (Shishkina et al., 2001) the large individual variations in the tooth burden were described in a dog
experiment for a short period after intake. However, even 50 years after 90Sr intake we can observe a large
difference in radionuclide concentration between different teeth of the same person.
Preliminary estimate of internal dose fraction in enamel (Tolstykh et al., 2000) had been performed
assuming an absence of 90Sr in the enamel of teeth with crown matured by the time of radioactive releases.
However, the detection of 90Sr in the enamel of teeth which crown had been completed by 1950-1951 indicates that
enamel self-irradiation can take place even for such teeth and therefore it cannot be ignored in the internal enamel
Comparing the 90Sr concentration in the different parts of dentin (crown and root zones) it is obvious that
the radionuclide is distributed non-uniformly in the whole dentine body. Moreover, the concentrations in the crown
and root dentin are not only unequal but also miscorrelated, because at the time of 90Sr intakes the dentin of different
teeth was at different stages of maturity (depending on donor’ s age, tooth position and individual variations in the
rate of tissue mineralization). This agrees with the results of Göksu et al. (2002), who have no found correlation
between cumulative dose and activity in different tooth tissues. Therefore, obtained results allow verifying other
approximation, which was used in (Tolstykh et al., 2000) as a first rough approximation for mature teeth - uniform
radionuclide distribution in the whole dentin body. For teeth that were at the stage of mineralization at the time of
intakes, the assumption of equal 90Sr concentration in the crown and root dentin cannot be applied.
Verification of dose formation model in teeth
In order to analyse the dependencies of the cumulative dose on the specific activity, dose formation in tooth
tissues due to incorporated 90Sr should be considered. Dose formation may be described by a following equation
analogous to that presented by Shved and Shishkina (2000b):
Dose formation in the tissue-target (j) is caused by radiation from tissue-sources (i). The enamel and crown dentin is
assumed as target tissues (detectors). Enamel, crown and root dentin are assumed to be a source of irradiation (n=3).
9 of 27
Thursday , December 08, 2005
of 90Sr from both enamel and dentin. Moreover, as was noted by Tolstykh et al (2000; 2002), tooth enamel
The dose accumulation period, T, for the Techa River donors is assumed as a period from beginning of 90Sr intakes
to the year of tooth examination. The value of
is the dose rate coefficient (DRC) in the target tissue (j), due to
exposure by one of the possible source tissue (i) with specific activity of 90Sr/90Y 1 Bq g-1. Ni(t) is the specific
activity of 90Sr/90Y in the source tissue (i).
Model approximations for description of radionuclide concentration in the source tissues
The specific activity of the source tissue is a function of time that depends upon metabolism and
radioactive decay. The specific activity of 90Sr in tooth tissues of Techa River residents is, in general, described by
here: t is the time since 1950 y; R(t) is retention function; ε(t) is excretion function.
Because the time of intake was significantly less than the time of exposure (the experimental examination
of teeth were performed in 2000) the intake can be assumed as a singular event. Correspondingly, the retention
function can be approximated by a constant value:
)( , 0NtRt
here: N0is initial specific activity of 90Sr/90Y after the intake.
In the Tolstykh et al (2000) the excretion functions for dentin were assumed as an exponential decay:
here: λ = 0.0238 is the constant of radioactive decay of 90Sr corresponding to its half-life period of 29 y and ξ is a
90Sr excretion rate. The excretion rate was estimated on the base of in vivo long-term monitoring of surface-beta
activity on front teeth (e.g., tooth beta counting, TBC) for Techa River residents (Tolstykh et al., 2000). The
measurements of teeth at high level of 90Sr accumulation in enamel (for age cohorts born in 1949-1951 which have
maximal TBC levels) were taken into account for excretion rate estimation. It was demonstrated that beta count rate
decreased with time and the rate of that decrease was higher than the radioactive decay by 1.9% per year. The TBC
method registers the surface-beta activity of anterior teeth. The thickness of enamel (0.3−0.5 mm) is shorter than the
average path length (2.1 mm) in enamel of beta particles emitted by 90Y (daughter of 90Sr). Therefore, mostly beta
particles originating in enamel, but also partially in the dentin layer (located near the junction of the enamel and
dentin), are registered by this method. Thus, the value of 1.9% per year can reflect the rate of biological elimination
10 of 27
Thursday , December 08, 2005
undergoes attrition, and age decreasing of beta count (1.9% per year) can reflect the loss in amount of emitting stuff.
Therefore, the approach of biological elimination rate should be verified.
Model approximation for calculation of DRC
The DRC depends on tooth position (size and shape variation in denture). The initial approach to the
problem of tooth-geometry description involved an extreme simplification of tooth shape. Such a very simplified
model of teeth (Seltzer et al., 2001, Shved and Shishkina, 2000b) did not take into account the complicated
morphology of dental tissues and tissue alteration during the human life. For example, the part of dentin formed
after eruption is often called secondary dentin. The extension of secondary dentin results in gradual narrowing of the
pulp chamber. According to Samusev et al. (2002), this process reaches its maximum speed in 13-19 years of
individual. During this period the pulp cavity decreases twice in comparison with its initial size. Further on, the
process of secondary dentin formation continues at smaller intensity until pulp is alive. Therefore, the assumption of
permanent dentin volume during the period of dose accumulation should be considered as preliminary
approximation. The age dependence of tissue volume can result in age dependence of DRC. The next
approximation was in the uniform radionuclide distribution in the tissues. Special investigations of 90Sr distribution
in dental tissues of Techa River residents were performed in 1960−1970 with the use of autoradiography (Ivanov,
1968; Saurov et al., 1972) and later using photostimulable phosphor imaging (Romanyukha et al., 2002). Their
results demonstrate two different patterns of 90Sr deposition: (1) for persons who were adult in the period of intake,
90Sr is concentrated mainly on the surface of the pulp channel and in the root and (2) for persons with tooth
formation during the period of intake, 90Sr is concentrated mainly in the enamel and dentin layer near the junction of
enamel and dentin.
Accumulated dose in enamel due to incorporated 90Sr in tooth tissues
The thickness of tooth enamel can, on average, vary from 0.1 mm (near the tooth neck) to 2 mm
(masticatory enamel). The mean path length of 90Sr/90Y beta particles in the enamel is 0.6 mm and maximal path
length is 2.3 mm (Seltzer et al., 2001). Because of it, non-uniformity of 90Sr distribution inside the enamel has
practically no effect on the dose accumulation.
The predictions of DRC for enamel self-exposure computed based on the described tooth models of 1st low
molar were compared. The statistical computation uncertainty was not exceeding 0.2%. The computational
uncertainty due to the uncertainty in the description of chemical composition and in the discretization of the source
energy distribution does not exceed 3%. Variation of enamel DRC due to average enamel thickness for both models
11 of 27
Thursday , December 08, 2005
root dentin as well as by external exposure. It was surprising that for the second group the individual variation in
was 6%. According to the simple model the average value of DRC for enamel self-irradiation is 0.063±0.004 nGy g
s-1 Bq-1 and the average value estimated by refined model is 0.056±0.004 nGy g s-1 Bq-1. Therefore, the results
obtained were not in contradiction one for the other, and the cylindrical simplification can be used for estimation of
dose accumulation for enamel self-exposure with 13% of uncertainty. The results of the computation are presented
in Table 4.
The assumption of biological elimination of 90Sr from enamel can be tested using samples for which
enamel self-exposure should be the main factor of dose accumulation. Only one sample (IDT 200 with very large
90Sr concentration in enamel) is suitable for such test. High 90Sr concentration in the enamel of this tooth should
result in a high internal dose in the enamel in such a way that the external dose fraction and uncertainty of
background dose estimation are significantly smaller than the detected value of the total accumulated dose.
According to Eqs 3-6, the computation of internal enamel dose based on measured current 90Sr contents due to self-
irradiation was performed for this tooth (Eq. 7). In this case, the current concentration of 90Sr/90Y is N = 2.006 NSr.
The accumulated dose during T = 50 y computed under an assumption of biological elimination (ξ=1.9) is
equal to 12±2 Gy. If the radionuclide concentration decreased only due to radioactive decay (ξ=0), the enamel dose
should be equal to 6.7±1.2 Gy. According to EPR measurement, the dose was found as 7.9±0.8 Gy. It should be
taken into account that the computation was made only for enamel self-irradiation. Some additional influence of
dentin should be present in the total internal enamel dose. The comparison of computations and experimental data
demonstrates that the approximation of biological elimination for enamel results in overestimation of internal dose,
whereas the underestimation was expected. Therefore, the assumption about the absence of biological elimination of
90Sr from enamel hydroxyapatite seems more correct.
Figure 4 demonstrates the comparison of EPR detected dose and dose due to self-exposure for the enamel
of investigated teeth computed under assumption of ξ=0. Teeth were divided on two groups. The first group
comprised the teeth with crowns matured by 1950, so accumulation of radionuclides should be mostly in the
secondary dentin. The second group was the teeth that have been formed in the time of 90Sr intakes, so radionuclides
can be located inside the primary dentin. Regression line that was plotted for the second group of teeth is practically
parallel to the line of data overlapping. The difference between computed enamel self-exposure and experimental
results amounted to 1.0 ± 0.2 Gy. This value can be conditioned by additional exposure due to 90Sr in the crown and
90Sr concentration in the dentin (from 3 Bq g-1 to 44 Bq g-1) produced approximately the same variation of
12 of 27
Thursday , December 08, 2005
process of dentin formation can result in full or partial obliteration of pulp chamber (zone 3 in Fig 1b). The new
accumulated dose in tooth enamel. Such effect can be possible only due to non-uniform radionuclide distribution in
the dentin body.
The variations in 90Sr localization in the crown dentin can significantly influence on dose formation in
enamel. The enamel DRC of dentin irradiation computed at different assumptions of source localization for 1stlow
molar is shown in Table 5. These computations assumed the uniform distribution of 90Sr in the primary dentin (zone
1 in Fig. 1b), secondary dentin (zone 2 in Fig. 1b) and whole dentin (zone 1 + zone 2). Based on Table 5, it has been
shown that variations in 90Sr distribution can result in the difference of enamel DRC by about an order of magnitude.
Therefore the enamel dose due to dentin irradiation cannot be calculated based only on the knowledge about the
average radionuclide contents. For the teeth, which have been formed in the time of intake, the approximation of
uniformity does not work.
For Techa River residents the variations in radionuclide distribution is more expectable for teeth with
crowns in the process of formation in 1950-1951. Most of investigated teeth exactly correspond to this condition.
Therefore, it seems reasonable that the external dose validation were done only using EPR measurements of teeth
that had been completed toward the 1950 (Degteva et al., in press).
The influence of root dentin on the enamel dose formation was estimated as 0.0002 nGy g Bq-1 s-1,
independent on tooth position.
Accumulated dose in crown dentin due to incorporated 90Sr in tooth tissues
The necessity of additional information about 90Sr distribution is vital for crown dentin too. According to
Table 5, the dose computed under the assumptions of uniform contamination of primary and secondary dentins can
be differing by about factor 3. Non-uniform 90Sr distribution inside the primary and secondary dentin can result in
yet more difference. This fact can only partially explain the discrepancy between the results obtained for crown
dentin by EPR measurements and beta count for teeth with IDTs 204, 116, 92 and 89. As can be seen from Table 2,
these samples have similar EPR doses (about 2 Gy), but their average specific activity of 90Sr varies in the range
from 3 to 21.8 Bq g-1 (the difference is about factor 7). The other source of the found discrepancy can be the
permanent extension of dentin volume. In contrast to bone, wherein osteoblasts transform into osteocytes that
become embedded within the mineralised matrix where they implement the bone resorption, the continuous
secretion of dentin matrix proteins is associated with the progressive lengthening of the odontoblast cell process and
retraction of the odontoblasts toward the dental pulp. (Papagerakis et al., 2002). At elderly age the continuous
formed hydroxyapatite is, so to speak, a dilution of exposed crystals. Correspondingly, the parameter
in Eq. 3
13 of 27
Thursday , December 08, 2005
Review Copy Download full-text
demonstrates its applicability for estimation of enamel self-exposure. However, the dentin as a source of irradiation
should depend on time after intake. Therefore, the formation of EPR centres should depend on time due to
radionuclide disintegration, biological elimination and hydroxyapatite extension. Thus, the obtained results
demonstrate that the dose accumulation in the crown dentin strongly depends on tooth physiology and mineral
exchange that is especially critical for teeth under formation at the time of radionuclide intakes. Another source of
discrepancy can be due to sample loss, which is unavoidable in the process of sample preparation. This fact can be
reflected in a certain disproportion between the detected dose and activity (due to non-uniformity of radionuclide
The dentin DRC due to enamel irradiation is equal to 0.009 nGy g Bq-1s-1. Root dentine adjacent to tooth
neck can induce the dose in the dentine that on the average was equal to 0.005 nGy g Bq-1s-1. However, the main
dose forming factor in the dentin is self-exposure.
Determination of 90Sr concentration in the tooth tissues allows detection of 90Sr in the enamel samples
completed by the time of radionuclide intakes. Therefore, in spite of the non-organic basis of tooth enamel (98% of
hydroxyapatite) the slow metabolism process (mostly due to ion exchange with saliva) results in the incorporation of
radionuclides in the tissue of completed teeth. Therefore the enamel self-exposure can not be ignored as a factor of
internal dose accumulation in the tooth enamel.
High individual variation between 90Sr tissue concentrations in the different teeth of the same donor was
found. Therefore, the precise estimation of internal dose in the tooth tissues required the individual measurements of
90Sr concentrations for each investigated tooth.
Teeth that were been formed in the time of radioactive releases into the Techa River demonstrate no
correlation between 90Sr contents in their different tissues as well as between doses accumulated in them. Therefore,
the assumption of a uniform (or proportional) tissue distribution of 90Sr inside the tooth cannot be applied to age
groups under study.
A preliminary test of one tooth does not allow making assumptions about the biological elimination of
radionuclides in the enamel. This fact casts doubt on the assumption of biological elimination in the enamel.
However, the final conclusions about enamel excretion rate can be made after a number of analogical tests are
Verification of simple tooth model describing the dose formation in the enamel and crown dentin
should be described by a more complicated geometry. Moreover, to enable internal dose estimation for teeth with
14 of 27
Thursday , December 08, 2005