Chest Wall Thickness measurements and the dosimetric implications for male workers in the Uranium Industry.

Gary H. Kramer^*, Barry M. Hauck^*, Steve A. Allen¹, Tae-Young Lee², Jong-Il Lee², and Si-Young Chang²

^*Human Monitoring Laboratory, Radiation Protection Bureau, 775 Brookfield Road, Ottawa, Ontario, K1A 1C1 (Gary_H_Kramer@hc-sc.gc.ca, www.hc-sc.gc.ca/ncrc/)

¹ Department of Compliance and Licensing, Cameco Corporation, 1 Eldorado Place, Port Hope, Ontario, Canada L1A 3A1

² Health Physics Department, Korea Atomic Energy Research Institute, PO Box 105, Yusong, Taejon, 305-600, Korea.

INTRODUCTION

Chest wall thickness (CWT) is often estimated by biometric equations; however, as Vickers (1) has shown, these equations are site specific. Large errors can be introduced into the CWT estimate, and hence the activity estimate, if a literature equation is simply applied to a worker population without verification. Health Canada’s Human Monitoring Laboratory (HML), which operates the National Calibration Reference Centre for In Vivo Monitoring (2), has measured chest wall thicknesses (CWT) of a representative number of male workers at Canadian uranium refinery, a conversion plant, and a fuel fabrication site to derive the appropriate biometric equation. Estimates of the adipose content of the chest wall was made on a selected group as the adipose mass fraction (AMF) can have a significant effect on the transmission of low energy photons (3). The HML has also measured chest wall thicknesses (CWT) of a representative number of male workers at the Korea Atomic Energy Research Institute (KAERI) to derive the appropriate biometric equation.

The CWT data has been used to estimate the sensitivity of lung counting for natural uranium for a phoswich and germanium based lung counting system. The dosimetric implications have been put into the perspective that will result from the imminent change to the Canadian Regulations. This change will result in a dose limit to 100 mSv in a five-year period with a maximum of 50 mSv in any one year. An average of 20 mSv per year has been assumed in this paper.

METHODS AND MATERIALS

Ultrasound Measurements: The HML’s portable ultrasound unit (Aloka SSD-500 echo camera) used a 5 MHz linear array to make the measurements. Each subject was asked to strip to the waist. A clear template and a black marker were used to mark 11 measurement positions on the subject’s chest above each lung. These positions represented areas of the chest that would be measured by both germanium and phoswich detectors. The ultrasound probe was covered with Aquasonic 100 gel prior to application to the subject’s chest. The gel was reapplied to the probe whenever signal loss prevented a clear being obtained on the ultrasound unit.

An estimate of the AMF was also made from selected ultrasound images taken at the conversion facility and at the refinery by measuring the thicknesses of the fat and the muscle layers in the imaged chest wall using the instruments internal calipers. Estimates of the AMF were also made at each measurement point on the subjects measured at the fuel fabrication site.

Canada: Eighty five male subjects at a uranium conversion facility were measured in a sitting geometry simulating the position of a germanium lung counter. Thirty-five male subjects were measured at a uranium refinery and eleven male subjects were measured at a fuel fabrication plant. The latter two groups were measured in a sitting and a supine position. The supine position simulates the counting position in the phoswich based lung counting system used by the company.

Korea: A clear template and a black marker were used to mark 14 measurement positions on the subject’s chest above each lung. These positions represented areas of the chest that would be measured by KAERI’s germanium detectors. One hundred and twenty one male subjects were measured in a supine geometry simulating the counting position of a germanium lung counter.

COUNTER CALIBRATIONS

Lung Counter: The lung counting system at KAERI, supplied by Canberrra, consists of two ACTII units. Each unit contains two Ge detectors (50 mm diameter, 20 mm thick, 0.5 mm Be window) cooled by one Dewar. Each two-detector unit is fitted with a graded shield consisting of 1.0 cm lead and 0.32 cm of copper to reduce background. The units are mounted in a Model 2275 dual purpose lung and whole body counter chamber, constructed of 10 cm thick low background steel, lined with stainless steel. Spectra acquired with the Ge detectors are stored and analysed using ABACOS-PC software.

Lung Counter Calibration: The lung counter had been previously calibrated by KAERI staff using the LLNL torso phantom with and without the B-series overlay plates. These plates simulate 50% adipose and 50% muscle so that when they are added to the chest plate, which is 0% adipose, the following adipose contents can be simulated: 0%, 15%, 21%, 26% and 30%. The KAERI lung counting efficiency has been fitted to adipose mass fraction, CWT, and photon energy as described elsewhere. The LLNL torso phantom containing 701 mg of natural uranium homogeneously distributed throughout the lung substitute material was used to estimate the counting efficiency of both the phoswich and the Ge lung counting systems. The phantom was counted with and without the C-series overlay plates to simulate 100% muscle and thereby give counting efficiency as a function of muscle-equivalent-chest wall thickness directly.

Phoswich Lung Counter: The phoswich lung counter consists of a low background monitoring chamber, a phoswich detector assembly and multichannel analyzer, all enclosed in a 12.2 meter air conditioned transport trailer. The counting chamber is constructed of 10.2 cm thick selected low background steel with a 3 mm lead liner. The interior dimensions are: 213 cm long, 132 cm high and 76 cm deep. Subjects are counted in a supine position using detectors positioned both above and below the chest cavity.

The lung counter consists of four phoswich detectors. Each detector consists of a NaI(Tl) crystal (12.5 cm diameter; 1.25 cm thick) optically coupled to a CsI(Na) crystal (12.5 cm diameter; 5.1 cm thick optically coupled to a photomultiplier tube. The phoswich detectors are connected through a multiplexer to a microcomputer based multichannel analyzer. The multichannel analyzer functions are controlled by the microcomputer, which also performs the data analysis and storage. Data were analyzed for the front two-detector arrays only.

Germanium lung counter: The HML’s germanium lung counting system consists of four large area germanium detectors supplied by EG&G Ortec. Each detector, which is cooled by a 17 liter Dewar, is 70 mm in diameter and 30 mm thick. The beryllium entrance window is 0.5 mm thick. The detectors are housed in a counting chamber that is constructed of 20 cm thick low background steel. The shield’s interior is covered with a lead liner that is approximately 0.6 cm thick. Spectra acquired from the individual Ge detectors were summed and analyzed as an array.

Background Measurements: The background count used for the MDA estimate of the Ge lung counter was obtained from the spectrum of an uncontaminated person who was measured for 1800 seconds in the HML’s lung counter. The counts in the appropriate spectral region of interest were used with the appropriate calibration factor for that energy. Units of MDA will depend on the units of the calibration factor and the counting time. For natural uranium, the HML uses both Bq and milligrams. The following formula is based on the work of Currie (5), with Brodsky’s (6) modification:

Where: N = background counts in the region of interest, E = counting efficiency (cps Bq^-1 or cps mg^-1), T = counting time (sec).

Similarly, the background count used for the MDA estimate of the phoswich lung counter was obtained by measuring an uncontaminated person for 1800 seconds who was of comparable size to the average worker. It is recognised that background, especially in a phoswich lung counter, varies with the individual’s size; however, to build in a size dependent background is beyond the scope of this work. It is suggested that readers applying this methodology measure a range of physical types to characterise their counting systems.

Korea: The lung counting system at KAERI, supplied by Canberrra, consists of two ACTII units. Each unit contains two Ge detectors (50 mm diameter, 20 mm thick, 0.5 mm Be window) cooled by one Dewar. Each two-detector unit is fitted with a graded shield consisting of 1.0 cm lead and 0.32 cm of copper to reduce background. The units are mounted in a Model 2275 dual purpose lung and whole body counter chamber, constructed of 10 cm thick low background steel, lined with stainless steel. Spectra acquired with the Ge detectors are stored and analysed using ABACOS-PC software.

The lung counter had been previously calibrated by KAERI staff using the LLNL torso phantom with and without the B-series overlay plates. These plates simulate 50% adipose and 50% muscle so that when they are added to the chest plate, which is 0% adipose, the following adipose contents can be simulated: 0%, 15%, 21%, 26% and 30%. The KAERI lung counting efficiency has been fitted to adipose mass fraction, CWT, and photon energy as described elsewhere. The background counts used for the MDA estimate of the Ge lung counter were obtained from the spectra of five uncontaminated persons who were measured for 1800 seconds in the KAERI’s lung counter. The counts in the appropriate spectral region of interest were used with the appropriate calibration factor for that energy.

RESULTS AND DISCUSSION

Ultrasound Measurements: Comparison of the data collected in the germanium detector counting regions with that of the phoswich region showed no difference in the mean chest wall thickness: a t-test showed that the null hypothesis (no difference between means) was accepted. Results of the t-test for each site were 0.0014, 4.1 x10^-5, and 0.027. Each site was then compared using ANOVA. The F value was 0.68 which is less than the value of F(2,133) indicating that the null hypothesis is accepted. The Canadian data has, therefore, been kept as a single set. The eleven measurement points on both left and right lungs were averaged to give an average chest wall thickness above each lung. The twenty two points were also combined and averaged to give an average chest wall thickness for a lung counting detector array - this data are reported here.

The grand average of a male supine subject was 3.73 cm. The supine posture used in the facility’s lung counter increased the average CWT by about 0.3 cm to 0.5 cm. This observation agrees with earlier work (7) and shows that there will be a decrease in counting efficiency compared to a seated geometry; consequently, there will be an increase in the minimum detectable activity (MDA).

The AMF of the chest wall of conversion plant and refinery workers lies in the range 5% to 40% with a mean of 22%. The fuel fabrication workers had a mean AMF of 21 " 6%. The range of the mean AMF was 17% to 28%, but individual measurement points on the workers’ chest were found to vary from 4% to 42%. This data seems comparable to the literature data - see Table 2. Comparison of these data with other literature data (1,7,8) indicate that the uranium refinery, conversion plant, and fuel fabrication workers have larger CWT values. Unfortunately this can not be statistically tested as there are insufficient literature data to make the comparison.

The Korean site specific data for the average, standard deviation of a single measurement, standard deviation of the mean, the median, the maximum, and the minimum of Age, Height, Weight, and CWT were collected for 121 KAERI workers. The grand average for male supine subjects was 2.70 cm.

Biometric Equation: Weight, height and age data that were used to derive an empirical equation to predict CWT. The largest CWT of 8.29 cm (Canada) has been excluded from this analysis as the height and weight data were not available for this individual. Linear regression was performed on the data using the function published elsewhere (8) and shown below:

Where: CWT is measured chest wall thickness (cm), Wt is subject weight (kg), Ht is subject height (m), and a, b, and c are the coefficients of regression.

The standard error of the Canadian data indicates that 68% of the predicted CWT values will be within 0.5 cm, 95% within 1 cm, and 99% within 1.5 cm. The standard error of the Korean data indicates that 68% of the predicted CWT values will be within 0.28 cm, 95% within 0.56 cm, and 99% within 0.84 cm. These findings support the rationale that the biometric equation be only used for routine counting and that ultrasound measurements be used to determine an accurate CWT in case of a positive lung burden being discovered.

Implications for MDA - Germanium counting: Although the higher energy (185 keV) photons of ²³⁵U are less attenuated than the 63 keV photons from ²³⁴ Th, the combination of the low natural abundance (0.712%), branching ratio (54%), and decreased counting efficiency as a function of CWT interact to slightly increase the MDA, relative to ²³⁴ Th for the Canadian data and the opposite for the Korean data..

Canada: Given that the range of CWT values is 1.9 cm to about 8 cm, the achievable range of MDA’s is about 8 mg to 57 mg of natural uranium using photons emitted from ²³⁵U (although a slight improvement can be obtained by measuring the photons emitted from ²³⁴Th, they cannot be used for reasons given below). The grand average CWT is 3.7 cm so the corresponding MDA will be about 14 mg. Doubling the counting time to 60 minutes will reduce the MDA to about 10 mg of natural uranium.

Korea: Given that the range of CWT values is 1.9 cm to about 4.1 cm, the achievable range of MDA’s is about 6.6 mg to 13.2 mg of natural uranium using photons emitted from ²³⁵U (although a slight improvement can be obtained by measuring the photons emitted from ²³⁴Th, they cannot be used for reasons given below). The grand average CWT is 2.7 cm so the corresponding MDA will be about 8.5 mg. Doubling the counting time to 60 minutes will reduce the MDA to about 6 mg of natural uranium.

Implications for MDA - Phoswich counting: Certain assumptions have been made in obtaining these data: count data were obtained for fixed regions of interest, and background correction is performed by matched subject counting. The data show that the phoswich detectors used in this study have a lower MDA for natural uranium if it is estimated from the 63 keV photopeak emitted from the daughter product ²³⁴ Th compared to the MDA values estimated from the 185 keV photons emitted by ²³⁵U; however, this assumes that the daughter (24 day half-life) is in equilibrium with the parent ²³⁸U. If the equilibrium is not established, then natural uranium must be estimated from the ²³⁵U photopeak and the accuracy of this estimate depends on the mass composition of natural uranium. An uncertainty will be introduced into the activity estimate if any enrichment or depletion has occurred during refinement.

Lung counting of uranium refinery, conversion plant, and fuel fabrication workers for natural uranium assumes that daughter equilibrium has not been established. Lung burdens are, therefore, determined from the 185 keV photons emitted by ²³⁵U. The two-detector-phoswich array is more sensitive than the germanium detector array. For example, at a muscle-equivalent-chest-wall-thickness of 4 cm the phoswich array’s MDA (²³⁵U) is approximately 12 mg whereas the germanium array’s MDA (²³⁵U) is about 15 mg.

Dosimetric implications: Table 1 shows the expected lung burden at various times following an intake of either 428 mg or 128 mg (the Annual Limit on Intake for Type M and S, i.e., that amount of activity that will give a committed effective dose of 20 mSv). The values were calculated from data published elsewhere (9). Type F was not considered as it will not be retained in the lung and should be part of a urinalysis bioassay program.

Four intake scenarios were investigated: one day before a lung count, seven days before a lung count, six months before a lung count, and one that occurred just after the last lung count, assuming an annual frequency.

Canada: Consider a hypothetical individual who has a CWT equal to the average value (3.7 cm, AMF 30%) has an intake in each time scenario. The hypothetical individual is measured using a phoswich array lung counter and the activity estimate is based on ²³⁵U. Type M intake: it will be readily detected on days 1 and 7; however, as the MDA for a 30 minute count time is 10.8 mg, days 180 and 360 will be missed. If the counting time is increased to 60 minutes the MDA would be expected to drop to 7.7 mg, and day 180 will be detected, but not day 360. Type S intake: days 1 and 7 may be detected if the counting time is 60 minutes, but days 180 and 365 will be missed as they are well below the MDA.

The conclusions are not substantially different if the counting system is a germanium lung counter; however, there are some modifying factors. The CWT of a seated subject is less than that of a supine subject: the average decrease in the CWT for the workforce was 0.3 cm. This decrease in the CWT will lower the MDA slightly but not sufficient to make any substantial difference in the level of detection and the conclusions presented below. The hypothetical individual is now measured using a germanium array lung counter and the activity estimate is based on ²³⁵U. Type M: days 1, and 7 would be detected, but not days 180 and 360 - there is no change in this finding even if the counting time is increased to 60 minutes. Type S: days 1, 7, 180 and 365 will be missed as they are below the MDA. Similarly, there is no change if the counting time is increased to 60 minutes.

For countries using earlier ICRP recommendations and a dose limit of 50 mSv/y the situation is somewhat different. The values in Table 1 would be multiplied by 2.5 and then compared against the MDA values in tables 6 and 7. Under these circumstances the hypothetical individual would have a Type M or S intake detected by a phoswich lung counter or a Ge lung counter for days 1, 7, 180 and 360 if the counting time was 60 minutes.

Korea: Consider that a hypothetical individual, who has a CWT equal to the average value (2.7 cm) and an AMF of 20%, has an intake in each time scenario. The hypothetical individual is measured using the KAERI lung counting Ge detector array and the activity estimate is based on ²³⁵U. Type M intake: it will be readily detected on days 1, 7 and 180; however, as the MDA for a 30 minute count time is 8.5 mg, day 360 will be missed. If the counting time is increased to 60 minutes the MDA would be expected to drop to 6.0 mg which is still too high to detect the intake on day 360. Type S intake: no intake will be detected if the counting time is 30 minutes. Days 1 and 7 will be detected if the counting time is 60 minutes, but days 180 and 365 will be missed as they are below the MDA.

For countries using earlier ICRP recommendations and a dose limit of 50 mSv/y the situation is somewhat different. The hypothetical individual would have a Type M or S intake detected for days 1, 7, 180 and 360 if the counting time was 30 minutes.

CONCLUSIONS

The average chest wall thickness of the seated persons measured at the uranium conversion plant, refinery, and at the fuel fabrication facility was 3.7 cm. Persons measured in a seated geometry had a thinner chest wall thickness than persons measured in a supine geometry - the average decrease was 0.3 cm. It follows that a seated geometry will give a slightly lower MDA (or decision level) than a supine geometry.

Achievable MDA’s (30 minute counting time) with a two-phoswich-detector array lie in the range of 6.7 mg to 19.1 mg of natural uranium based on the ²³⁵U emissions over a range of CWT of 1.6 cm to 6.0 cm. The average achievable MDA is about 11 mg which can be reduced to about 8 mg by doubling the counting time. Similarly, MDA’s (30 minute counting time) obtainable with a germanium lung counting system will lie in the range of 7 mg to 30 mg of natural uranium based on the ²³⁵U emissions over a range of CWT of 1.6 cm to 6.0 cm. The average achievable MDA will be about 14 mg which can be reduced to about 10 mg by doubling the counting time.

The average chest wall thickness of the supine persons measured at KAERI was 2.7 cm. Achievable MDA’s (30 minute counting time) with the KAERI lung counting Ge detector array lie in the range of about 6.6 mg to 13.2 mg of natural uranium based on the ²³⁵U emissions over a range of CWT of 1.9 cm to 4.1 cm. The average achievable MDA is about 8.5 mg which can be reduced to about 6 mg by doubling the counting time.

Unfortunately, all of these MDA values are close to, or above, the predicted amounts of natural uranium that will remain in the lung after an intake equivalent to the Annual Limit on Intake that corresponds to 20 mSv.

REFERENCES

1 Vickers, L.R. The gender specific chest wall thickness prediction equations for routine measurements of ²³⁹Pu and ²⁴¹Am within the lungs using HPGe detectors. Health Phys. 70(3), 346-357 (1996).

2 Kramer, G.H.; Limson Zamora, M. The Canadian National Calibration Reference Centre for Bioassay and In-Vivo Monitoring: A Program Summary. Health Phys. 67(2), 192-196 (1994).

3 Kruchten, D.A; Anderson, A.L. Improved ultrasonic measurement techniques applied to assay of Pu and other transuranics in lung. Health Phys. 59(1), 117-123 (1990).

4 Griffith R.V., Dean P.N., Anderson A.L., Fisher J.C. Tissue equivalent torso phantom for intercalibration of in vivo transuranic nuclide counting facilities. In: Advances in Radiation Monitoring, Proceedings of an International Atomic Energy Agency Conference. IAEA-SM-229/56, 4493-504. (IAEA, Vienna) (1978).

5 Currie; L.A. Limits for qualitative detection and quantitative determination: Application to radiochemistry. Analytical Chemistry 40, 586-593 (1968).

6 Brodsky, A. Accuracy and detection limits for bioassay measurements in radiation protection. Report NUREG-1156 (Washington D.C., US Nuclear Regulatory Commission) (1986).

7 Palmer, H.E.; Rieksts, G.A; Jefferies, S.J.; Gunston, K.J. Improved counting efficiencies for measuring ²³⁹Pu in the lung in the sitting position. Health Phys. 57(5), 747-752 (1989).

8 Summerling, T.J.; Quant, S.P. Measurements of the human anterior chest wall by ultrasound and estimates of chest wall thickness for use in determination of transuranic nuclides in the lung. Rad. Prot. Dosim. 3(4), 203-210 (1982).

9 International Commission on Radiological Protection. Individual Monitoring for internal exposure of workers. ICRP publication No. 78 (Oxford, Pergamon Press) (1997).

Table 1: Amount of natural uranium retained (mg) in the lung following an acute intake of an amount equal to one Annual Limit on Intake (ICRP 1997).

	Amount retained in the lungs (mg)
Day since intake	M: (428 mg)	S: (128 mg)
1	24.8	8.2
7	23.1	7.8
180	9.4	4.9
360	5.1	4.1