Appl. Radiat. Isot. Vol. 43, No. 6, pp. 795-800, 1992 Int. J. Radiat. Appl. Instrum. Part A
synthetic ribs, sternum, cartilage and clavicles, configured as described by the International Commission on Radiological Protection (ICRP). The phantom was based on a cadaver which was representative of a population of radiation workers at the Lawrence Livermore National Laboratories and at Los Alamos National Laboratories. The use of human skeletons would have created unacceptable interphantom variability, therefore synthetic bone was developed to standardize phantom construction. The phantom was moulded about a mandrel with a cavity that may be filled with tissue-equivalent organs. Three sets of chest plates can be used with the phantom to simulate various tissue types and tissue compositions. Each set of plates is composed of either 87% adipose and 13% muscle equivalent, or 50% adipose and 50% muscle equivalent, or 100% lean muscle equivalent tissue substitutes. Within each set there are four plates of progressively greater thickness, so that the phantom may represent a wide variety o f subject dimensions. The lungs were modelled after cadaver lungs inflated to a median respiratory state and were made of polyurethane foam with a calcium carbonate filler for radio-equivalence. A solid block of inert, muscle-equivalent, polyurethane is positioned beneath the heart, and between the right and left lungs to simulate the mediastinum and provide recesses for lymph nodes. The heart, also muscle-equivalent polyurethane, completes the upper torso organ assembly, such that with lymph node capsules in place there are minimal air spaces. The muscle-equivalent liver, which is encased in an inert liver envelope, fills the lower torso area of the phantom while ensuring the correct anatomical position of the liver with respect to other organs. The total weight of the phantom is 31.9 kg (with all organs placed in the cavity) and the chest circumference is 101 cm. The weights and fat to muscle ratios of the phantom and the various chest plates are given in Table 1.
Pergamon Press Ltd 1992. Printed in Great Britain 0~83-2889/92 $5.00 + 0.00
Evaluation of the Lawrence Livermore National Laboratory (LLNL) Torso Phantom by Bone Densitometry and x-Ray G A R Y H. K R A M E R ~ and COLIN E. WEBBER 2. t t4uman Monitoring Laboratory, BRMD, 775 Brookfield Road, Ottawa, Ontario, Canada K1A 1CI and 2E)epartment of Nuclear Medicine, Chedoke-McMaster Hospitals, Hamilton, Ontario, Canada L8N 3Z5 (Received 2 4 O c t o b e r 1991)
"I he recent Workshop on Standard Phantoms recommended that the LLNL torso phantom be adopted as a calibration slandard for the quantitation of in vivo radioactivity. This paantom was designed for the calibration of systems for the detection of x-rays of less than 20 keV. The anthropomorpaic characteristics and tissue substitute composition of the paantom were assessed with techniques using photons of h g h e r energy. Dual photon absorptometry at 42 and 1)0 keV showed that the phantom was representative of in v vo tissue composition. Chest radiography showed that the phantom was representative of a human even though the s~omach, GI tract and the scapulae were not present and air gaps were observed at organ boundaries.
Introduction
Apparatus and Experimental Procedure
I he Workshop on Standard Phantoms for in vivo Radioactivity Measurement was held at the National Institute of Standards and Technology (NIST), Gaithersburg, MD, on 13-15 February 1990. The workshop was organized by !",IST and the Human Monitoring Laboratory (HML) fi~llowing positive reaction resulting from the proposal published in the April 1989 Health Physics Newsletter. The proceedings of the workshop are presently in preparation aad will soon be available (Kramer et al., 1990). The x~orkshop recommended that the LLNL torso phantom be azlopted as the standard design for calibration of lung counting systems with perhaps some minor modifications to be made at a later date. Since the original design criterion fiw the phantom was specifically to provide accurate calibration of systems for the detection of 13-20 keV photons, this recommendation prompted us to investigate the phantom from a different perspective. It was subjected to traditional chest radiography and to bone densitometry to obtain information about the gross morphology and tissue equivalence at other photon energies. The LLNL torso phantom has been described (Gritfith e' al., 1978) and its use is exemplified elsewhere (Northcutt e a l . , 1988; Palmer et al., 1987). It is based on a 1.77 m tall, 76kg male and is manufactured from tissue-equivalent material corresponding to 100% lean muscle. The torso extends to the mid-abdomen with a stub neck, stub arms,
Chest radiography was performed using hospital equipment in the Department of Radiology, Chedoke-McMaster Hospital. A standard anteroposterior chest x-ray was obtained at 140kvp. The films were reviewed by a staff radiologist. Bone Densitometry was performed with a Norland 2600 Dichromatic Bone Densitometer.t This instrument measures the transmission of photons of energy 42 and 100 keV emitted from a ~53Gd source. The instrument emits a pencil beam of photons that is scanned in a rectilinear pattern throughout the tissue region of interest. If the assumption is made that the region scanned consists of only two radiologically distinct materials then at every point of measurement the mass per unit area of each material can be obtained. This is the most common mode of operation of the instrument when it is used to examine bone mineral density in specific regions of the skeleton which may be subject to loss of bone mineral due to osteoporosis. When dual photon measurements are made throughout the body, mass per unit area can be summed to yield a total body mass. Initially the two materials considered are bone mineral and soft tissue and the total mass of each component is measured. In addition, at all measurement sites where no bone mineral is present, a second dual photon analysis can be performed in which the two materials are considered to be lean tissues and fat. If it is assumed that the average fat fraction measured at all soft tissue sites applies also to the soft tissue measured at sites which included bone mineral, then it is possible to derive total lean body mass and total body fat mass. Such measurements can
*Author for correspondence. tNorland Corporation, Norland Drive, Fort Atkinson, WI 53538, U.S.A. 795
Technical Note
796
Table 1. Weight and composition of the LLNL phantom, the chest plate and the A4, B4 and C4 overlays Sample LLNL composite Chest Plate (CP) CP + A4 CP + B4 C P + C4
Weight (kg)
Fat: muscle
31.9 3.1 8.3 8.2 8.5
0:100 0:100 35 : 65 20: 80 0:100
be used to examine the effects of various clinical conditions and of the effect of exercise on gross body composition. For bone mineral density measurements the system is calibrated by m e a n s of daily measurements of a standard containing a known mass of mineral. For body composition measurements the densitometer relies on an internal calibration which is based upon the relative attenuation at the two photon energies (the Rst value) for water and for lard. Such a calibration can lead to small errors since water and lard are not ideal substitutes for lean muscle and fat but are good robust standards for daily clinical operation. The L L N L p h a n t o m was measured radiographically and densitometrically with a filled chest cavity (i.e. all organ inserts present) but with no overlay plates. The tissue composition measurements were made with the chest plate alone and with the chest plate covered with the thickest overlay plate from each o f the three available sets: A, 87% adipose 13% muscle; B, 50% adipose 50% muscle; C, 100% muscle. This configuration was chosen to maximise the tissue composition change but to keep the thickness of material to a minimum. The spine scan was performed with the chest cavity empty.
Results The chest radiograph is shown in Fig. 1. Figure 2 shows a normal dual photon scan of the lumbar spine and of the whole body obtained from an adult male. For normal adult males, the bone mineral density for the L2 to L4 region of the lumbar spine is typically between 0.90 and 1.45 g cm-2. For the whole body scan shown in Fig. 2, it would be possible to obtain total bone mineral mass, lean body mass and fat mass. Figure 3 shows a total body scan of the L L N L phantom. Tissue composition measurements were made for the phantom, the chest plate and the chest plate with various overlays. These results are given in Table 2 and are compared in Fig. 4 with the fat to muscle ratio specified by the manufacturer. The slope of the regression line shown in Fig. 4 is 1.145, the intercept is - 0 . 0 6 and the correlation coefficient is 0.997. The sum of mineral, lean and fat tissue mass is equal to the total mass of the phantom. Figure 5 compares the dual photon measured weight with that obtained from regular clinical scales. The slope of the regression equation is 1.006, the intercept is - 0 . 7 kg and the correlation coefficient is 0.999. Figure 6 shows a high resolution dual photon scan of the upper region of the chest phantom. The detailed structure o f the ribs can be seen. Finally, Fig. 7 shows a spine scan of the phantom. It can be seen that the spine is not segregated into individual vertebral bodies. Nevertheless in those regions of the spine which include the ribs and those Table 2. Tissue composition by dual photon absorptiometry TBM LM FM SUM Sample (kg) (kg) (kg) (kg) Torso CP CP + A4 CP + B4 CP + C4
0.42 0.026 0.025 0.024 0.028
32.25 1.72 5.08 6.56 8.71
- 1.45 -0.12 2.88 1.24 -0.52
31.22 1.63 7.99 7.82 8.22
immediately distal to the ribs, the bone mineral density is within the normal adult male range. Further distally where the spine broadens, abnormally high values up to 1.8 g cm-2 were observed.
Discussion Originally the tissue equivalence o f the L L N L p h a n t o m was ensured by the selection of appropriate substitutes based on elemental photon interaction data (White, 1977). Tissue equivalence for low energy x-rays has been verified (Newton et al., 1978) and the p h a n t o m has been used for interlaboratory comparisons (Campbell et al., 1981). However, if the p h a n t o m is to be used for the calibration o f other systems at different energies such as uranium lung counting for example, then tissue equivalence must be confirmed at that energy and the morphological appearance of the phantoms should be appropriate for the application o f concern. The major discrepancies between the p h a n t o m radiograph and that expected from a normal adult male were that the stomach, the gastrointestinal tract and the scapulae were not present. N o airways or blood vessels were apparent and the presence of air gaps around the organs was obvious. No subcutaneous fat was visible. On a h u m a n radiograph the posterior ribs would be more accentuated than the anterior ribs because of differences in the a m o u n t of cartilage and because of size and shape differences. Bone densitometry measurements gave results which were within acceptable ranges for lumbar spine measurement in patients at C h e d o k e - M c M a s t e r Hospitals. The structure of the p h a n t o m ' s spine is not anthropomorphically correct and the area density in the distal region of the spine is greater than would be expected in adult males. However, this should not present a problem provided the p h a n t o m is used for front mounted detectors. Dual photon absorptiometry provides accurate and precise measurements of total body mineral mass and total body soft tissue mass. This is confirmed by the close agreement observed between dual photon measured weight and the weight measured on regular clinical scales. To separate the soft tissue mass into lean and fat components the densitometer relies on an internal calibration which relates the average ratio of the attenuation coefficients measured at the two photon energies for all pixels containing no bone mineral. This value is converted to a fat to lean ratio by comparison with a calibration curve based on ratio measurements for mixtures of water and lard as tissue substitutes. Since these materials are not ideal substitutes for muscle and fat, a bias is introduced. This is the origin of the negative fat fractions and fat masses found in the chest plate and the chest plate overlay measurements. Negative fat values have been observed previously with this equipment in lean subjects such as long distance runners and in body composition measurements in dogs (Kolkman et al., 1990). The results show that on first analysis, the p h a n t o m is representative of the expected tissue types for the energy range evaluated and that the manufacturer's claimed tissue composition values are correct.
Conclusions The data obtained during this study substantiates the recommendation of the Workshop on Standard P h a n t o m s to adopt the L L N L torso p h a n t o m as a standard design for calibration of lung counting systems. Providing that the p h a n t o m is used for frontal detector calibrations the degree of anthropomorphicity is high for lung counting. Failings have been identified in the GI tract and the back of the p h a n t o m (spine and shoulder blades). This study has shown that dual photon absorptiometry is a useful tool that can be applied to p h a n t o m verification. Its use should seriously be considered for inclusion in a quality assurance program to maintain a high degree of confidence in the L L N L phantom.
Fig. 1. x-Ray of the LLNL torso phantom.
Fig. 2. Typical clinical dual photon scans of a lumbar spine (left) and the whole body (right). 797
Fig. 3. Torso scan of the LLNL phantom.
30
0,80 - ¢
0.60
O]
~)
//
20
0.40 L~
0.20 o,oo
--
r"
o 0
0.00
0,20
0.40
0.60
0.80
Phantom Fat/Muscle
Fig, 4. Comparison o f dual p h o t o n fat: muscle ratios with
the manufacturer's specifications.
10
20
30
40
Scale Weight (kg)
Fig. 5. Comparison of dual photon measured weight and scale weights for the phantom, the chest plate and the overlays.
798
Fig. 6. High resolution scan of the upper torso of the L L N L phantom.
Fig. 7. Spine scan of the L L N L phantom.
799
800
Technical Note
Acknowledgement--The authors thank Dr Craig Cobblentz of the Department of Radiology, Chedoke McMaster Hospital, Hamilton, Ontario for his review of the chest radiographs of the LLNL phantom. References
Campbell G. W., Anderson A. L., Fry F. A., Newton D. and Ramsden D. (1981) Calibration of phoswich detectors for assessment of plutonium in lungs: the methods of four laboratories compared. Health Phys. 40, 405. Griffith R. V., Dean P. N., Anderson A. L. and Fisher J. C. (1978) 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, 493 504. IAEA, Vienna. Kolkman P., Belbeck L. W. and Webber C. E. (1991) Bone mineral measurements and Tc-99m IDP retention in the dog. Calc. Tiss. Int. 48, 120.
Kramer G. H., Inn K. G. W. and Burns L. (1990) Proceedings" of the Workshop on Standard Phantoms for In-Vivo Radioactivity Measurements. Human Monitoring Laboratory Technical Document. Ottawa: Environmental Radiation Hazards Division; HMLTD-90-4. Newton D., Fry F. A., Taylor B. T., Eagle M. C. and Sharma R. C. (1978) Interlaboratory comparison of techniques for measuring lung burdens of low energy photon emitters. Health Phys. 35, 751. Northcutt A. R., Binney S. E. and Palmer H. E. (1988) In-vivo counting of ~41Am in human lungs and tracheobronchial lymph nodes. Health Phys. 54, 73. Palmer H. E., Rieksts G. A., Brim C. P. and Rhoads M. C. (1987) Hanford Whole Body Counting Manual. Pacific Northwest Laboratories, Washington; PNL-6198 UC-41. White D. R. (1977) The formulation of tissue substitute materials using basic interaction data. Phys. Med. Biol. 22, 889.
synthetic ribs, sternum, cartilage and clavicles, configured as described by the International Commission on Radiological Protection (ICRP). The phantom was based on a cadaver which was representative of a population of radiation workers at the Lawrence Livermore National Laboratories and at Los Alamos National Laboratories. The use of human skeletons would have created unacceptable interphantom variability, therefore synthetic bone was developed to standardize phantom construction. The phantom was moulded about a mandrel with a cavity that may be filled with tissue-equivalent organs. Three sets of chest plates can be used with the phantom to simulate various tissue types and tissue compositions. Each set of plates is composed of either 87% adipose and 13% muscle equivalent, or 50% adipose and 50% muscle equivalent, or 100% lean muscle equivalent tissue substitutes. Within each set there are four plates of progressively greater thickness, so that the phantom may represent a wide variety o f subject dimensions. The lungs were modelled after cadaver lungs inflated to a median respiratory state and were made of polyurethane foam with a calcium carbonate filler for radio-equivalence. A solid block of inert, muscle-equivalent, polyurethane is positioned beneath the heart, and between the right and left lungs to simulate the mediastinum and provide recesses for lymph nodes. The heart, also muscle-equivalent polyurethane, completes the upper torso organ assembly, such that with lymph node capsules in place there are minimal air spaces. The muscle-equivalent liver, which is encased in an inert liver envelope, fills the lower torso area of the phantom while ensuring the correct anatomical position of the liver with respect to other organs. The total weight of the phantom is 31.9 kg (with all organs placed in the cavity) and the chest circumference is 101 cm. The weights and fat to muscle ratios of the phantom and the various chest plates are given in Table 1.
Pergamon Press Ltd 1992. Printed in Great Britain 0~83-2889/92 $5.00 + 0.00
Evaluation of the Lawrence Livermore National Laboratory (LLNL) Torso Phantom by Bone Densitometry and x-Ray G A R Y H. K R A M E R ~ and COLIN E. WEBBER 2. t t4uman Monitoring Laboratory, BRMD, 775 Brookfield Road, Ottawa, Ontario, Canada K1A 1CI and 2E)epartment of Nuclear Medicine, Chedoke-McMaster Hospitals, Hamilton, Ontario, Canada L8N 3Z5 (Received 2 4 O c t o b e r 1991)
"I he recent Workshop on Standard Phantoms recommended that the LLNL torso phantom be adopted as a calibration slandard for the quantitation of in vivo radioactivity. This paantom was designed for the calibration of systems for the detection of x-rays of less than 20 keV. The anthropomorpaic characteristics and tissue substitute composition of the paantom were assessed with techniques using photons of h g h e r energy. Dual photon absorptometry at 42 and 1)0 keV showed that the phantom was representative of in v vo tissue composition. Chest radiography showed that the phantom was representative of a human even though the s~omach, GI tract and the scapulae were not present and air gaps were observed at organ boundaries.
Introduction
Apparatus and Experimental Procedure
I he Workshop on Standard Phantoms for in vivo Radioactivity Measurement was held at the National Institute of Standards and Technology (NIST), Gaithersburg, MD, on 13-15 February 1990. The workshop was organized by !",IST and the Human Monitoring Laboratory (HML) fi~llowing positive reaction resulting from the proposal published in the April 1989 Health Physics Newsletter. The proceedings of the workshop are presently in preparation aad will soon be available (Kramer et al., 1990). The x~orkshop recommended that the LLNL torso phantom be azlopted as the standard design for calibration of lung counting systems with perhaps some minor modifications to be made at a later date. Since the original design criterion fiw the phantom was specifically to provide accurate calibration of systems for the detection of 13-20 keV photons, this recommendation prompted us to investigate the phantom from a different perspective. It was subjected to traditional chest radiography and to bone densitometry to obtain information about the gross morphology and tissue equivalence at other photon energies. The LLNL torso phantom has been described (Gritfith e' al., 1978) and its use is exemplified elsewhere (Northcutt e a l . , 1988; Palmer et al., 1987). It is based on a 1.77 m tall, 76kg male and is manufactured from tissue-equivalent material corresponding to 100% lean muscle. The torso extends to the mid-abdomen with a stub neck, stub arms,
Chest radiography was performed using hospital equipment in the Department of Radiology, Chedoke-McMaster Hospital. A standard anteroposterior chest x-ray was obtained at 140kvp. The films were reviewed by a staff radiologist. Bone Densitometry was performed with a Norland 2600 Dichromatic Bone Densitometer.t This instrument measures the transmission of photons of energy 42 and 100 keV emitted from a ~53Gd source. The instrument emits a pencil beam of photons that is scanned in a rectilinear pattern throughout the tissue region of interest. If the assumption is made that the region scanned consists of only two radiologically distinct materials then at every point of measurement the mass per unit area of each material can be obtained. This is the most common mode of operation of the instrument when it is used to examine bone mineral density in specific regions of the skeleton which may be subject to loss of bone mineral due to osteoporosis. When dual photon measurements are made throughout the body, mass per unit area can be summed to yield a total body mass. Initially the two materials considered are bone mineral and soft tissue and the total mass of each component is measured. In addition, at all measurement sites where no bone mineral is present, a second dual photon analysis can be performed in which the two materials are considered to be lean tissues and fat. If it is assumed that the average fat fraction measured at all soft tissue sites applies also to the soft tissue measured at sites which included bone mineral, then it is possible to derive total lean body mass and total body fat mass. Such measurements can
*Author for correspondence. tNorland Corporation, Norland Drive, Fort Atkinson, WI 53538, U.S.A. 795
Technical Note
796
Table 1. Weight and composition of the LLNL phantom, the chest plate and the A4, B4 and C4 overlays Sample LLNL composite Chest Plate (CP) CP + A4 CP + B4 C P + C4
Weight (kg)
Fat: muscle
31.9 3.1 8.3 8.2 8.5
0:100 0:100 35 : 65 20: 80 0:100
be used to examine the effects of various clinical conditions and of the effect of exercise on gross body composition. For bone mineral density measurements the system is calibrated by m e a n s of daily measurements of a standard containing a known mass of mineral. For body composition measurements the densitometer relies on an internal calibration which is based upon the relative attenuation at the two photon energies (the Rst value) for water and for lard. Such a calibration can lead to small errors since water and lard are not ideal substitutes for lean muscle and fat but are good robust standards for daily clinical operation. The L L N L p h a n t o m was measured radiographically and densitometrically with a filled chest cavity (i.e. all organ inserts present) but with no overlay plates. The tissue composition measurements were made with the chest plate alone and with the chest plate covered with the thickest overlay plate from each o f the three available sets: A, 87% adipose 13% muscle; B, 50% adipose 50% muscle; C, 100% muscle. This configuration was chosen to maximise the tissue composition change but to keep the thickness of material to a minimum. The spine scan was performed with the chest cavity empty.
Results The chest radiograph is shown in Fig. 1. Figure 2 shows a normal dual photon scan of the lumbar spine and of the whole body obtained from an adult male. For normal adult males, the bone mineral density for the L2 to L4 region of the lumbar spine is typically between 0.90 and 1.45 g cm-2. For the whole body scan shown in Fig. 2, it would be possible to obtain total bone mineral mass, lean body mass and fat mass. Figure 3 shows a total body scan of the L L N L phantom. Tissue composition measurements were made for the phantom, the chest plate and the chest plate with various overlays. These results are given in Table 2 and are compared in Fig. 4 with the fat to muscle ratio specified by the manufacturer. The slope of the regression line shown in Fig. 4 is 1.145, the intercept is - 0 . 0 6 and the correlation coefficient is 0.997. The sum of mineral, lean and fat tissue mass is equal to the total mass of the phantom. Figure 5 compares the dual photon measured weight with that obtained from regular clinical scales. The slope of the regression equation is 1.006, the intercept is - 0 . 7 kg and the correlation coefficient is 0.999. Figure 6 shows a high resolution dual photon scan of the upper region of the chest phantom. The detailed structure o f the ribs can be seen. Finally, Fig. 7 shows a spine scan of the phantom. It can be seen that the spine is not segregated into individual vertebral bodies. Nevertheless in those regions of the spine which include the ribs and those Table 2. Tissue composition by dual photon absorptiometry TBM LM FM SUM Sample (kg) (kg) (kg) (kg) Torso CP CP + A4 CP + B4 CP + C4
0.42 0.026 0.025 0.024 0.028
32.25 1.72 5.08 6.56 8.71
- 1.45 -0.12 2.88 1.24 -0.52
31.22 1.63 7.99 7.82 8.22
immediately distal to the ribs, the bone mineral density is within the normal adult male range. Further distally where the spine broadens, abnormally high values up to 1.8 g cm-2 were observed.
Discussion Originally the tissue equivalence o f the L L N L p h a n t o m was ensured by the selection of appropriate substitutes based on elemental photon interaction data (White, 1977). Tissue equivalence for low energy x-rays has been verified (Newton et al., 1978) and the p h a n t o m has been used for interlaboratory comparisons (Campbell et al., 1981). However, if the p h a n t o m is to be used for the calibration o f other systems at different energies such as uranium lung counting for example, then tissue equivalence must be confirmed at that energy and the morphological appearance of the phantoms should be appropriate for the application o f concern. The major discrepancies between the p h a n t o m radiograph and that expected from a normal adult male were that the stomach, the gastrointestinal tract and the scapulae were not present. N o airways or blood vessels were apparent and the presence of air gaps around the organs was obvious. No subcutaneous fat was visible. On a h u m a n radiograph the posterior ribs would be more accentuated than the anterior ribs because of differences in the a m o u n t of cartilage and because of size and shape differences. Bone densitometry measurements gave results which were within acceptable ranges for lumbar spine measurement in patients at C h e d o k e - M c M a s t e r Hospitals. The structure of the p h a n t o m ' s spine is not anthropomorphically correct and the area density in the distal region of the spine is greater than would be expected in adult males. However, this should not present a problem provided the p h a n t o m is used for front mounted detectors. Dual photon absorptiometry provides accurate and precise measurements of total body mineral mass and total body soft tissue mass. This is confirmed by the close agreement observed between dual photon measured weight and the weight measured on regular clinical scales. To separate the soft tissue mass into lean and fat components the densitometer relies on an internal calibration which relates the average ratio of the attenuation coefficients measured at the two photon energies for all pixels containing no bone mineral. This value is converted to a fat to lean ratio by comparison with a calibration curve based on ratio measurements for mixtures of water and lard as tissue substitutes. Since these materials are not ideal substitutes for muscle and fat, a bias is introduced. This is the origin of the negative fat fractions and fat masses found in the chest plate and the chest plate overlay measurements. Negative fat values have been observed previously with this equipment in lean subjects such as long distance runners and in body composition measurements in dogs (Kolkman et al., 1990). The results show that on first analysis, the p h a n t o m is representative of the expected tissue types for the energy range evaluated and that the manufacturer's claimed tissue composition values are correct.
Conclusions The data obtained during this study substantiates the recommendation of the Workshop on Standard P h a n t o m s to adopt the L L N L torso p h a n t o m as a standard design for calibration of lung counting systems. Providing that the p h a n t o m is used for frontal detector calibrations the degree of anthropomorphicity is high for lung counting. Failings have been identified in the GI tract and the back of the p h a n t o m (spine and shoulder blades). This study has shown that dual photon absorptiometry is a useful tool that can be applied to p h a n t o m verification. Its use should seriously be considered for inclusion in a quality assurance program to maintain a high degree of confidence in the L L N L phantom.
Fig. 1. x-Ray of the LLNL torso phantom.
Fig. 2. Typical clinical dual photon scans of a lumbar spine (left) and the whole body (right). 797
Fig. 3. Torso scan of the LLNL phantom.
30
0,80 - ¢
0.60
O]
~)
//
20
0.40 L~
0.20 o,oo
--
r"
o 0
0.00
0,20
0.40
0.60
0.80
Phantom Fat/Muscle
Fig, 4. Comparison o f dual p h o t o n fat: muscle ratios with
the manufacturer's specifications.
10
20
30
40
Scale Weight (kg)
Fig. 5. Comparison of dual photon measured weight and scale weights for the phantom, the chest plate and the overlays.
798
Fig. 6. High resolution scan of the upper torso of the L L N L phantom.
Fig. 7. Spine scan of the L L N L phantom.
799
800
Technical Note
Acknowledgement--The authors thank Dr Craig Cobblentz of the Department of Radiology, Chedoke McMaster Hospital, Hamilton, Ontario for his review of the chest radiographs of the LLNL phantom. References
Campbell G. W., Anderson A. L., Fry F. A., Newton D. and Ramsden D. (1981) Calibration of phoswich detectors for assessment of plutonium in lungs: the methods of four laboratories compared. Health Phys. 40, 405. Griffith R. V., Dean P. N., Anderson A. L. and Fisher J. C. (1978) 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, 493 504. IAEA, Vienna. Kolkman P., Belbeck L. W. and Webber C. E. (1991) Bone mineral measurements and Tc-99m IDP retention in the dog. Calc. Tiss. Int. 48, 120.
Kramer G. H., Inn K. G. W. and Burns L. (1990) Proceedings" of the Workshop on Standard Phantoms for In-Vivo Radioactivity Measurements. Human Monitoring Laboratory Technical Document. Ottawa: Environmental Radiation Hazards Division; HMLTD-90-4. Newton D., Fry F. A., Taylor B. T., Eagle M. C. and Sharma R. C. (1978) Interlaboratory comparison of techniques for measuring lung burdens of low energy photon emitters. Health Phys. 35, 751. Northcutt A. R., Binney S. E. and Palmer H. E. (1988) In-vivo counting of ~41Am in human lungs and tracheobronchial lymph nodes. Health Phys. 54, 73. Palmer H. E., Rieksts G. A., Brim C. P. and Rhoads M. C. (1987) Hanford Whole Body Counting Manual. Pacific Northwest Laboratories, Washington; PNL-6198 UC-41. White D. R. (1977) The formulation of tissue substitute materials using basic interaction data. Phys. Med. Biol. 22, 889.
