- Email: [email protected]
Gamma-ray backscatter for body composition measurement
~
Pergamon PIh
Appl. Radiat. Isot. Vol. 49, No. 5/6, pp. 555 557, 1998 ,~" 1998 Elsevier Science Ltd. All rights reserved Printed in Great Britain S0969-8043(97)00186-3 0969-8043/98 $19.00 + 0.00
Gamma-ray Backscatter for Body Composition Measurement H. M. M O R G A N * ,
J. T. S H A K E S H A F T
a n d S. C. L I L L I C R A P
Medical Physics Department, Royal United Hospital, Combe Park, Bath BA1 3NG, U.K. The purpose of this study was to examine the potential of using backscatter information to assess regional body composition at selected sites. Two measurement techniques are examined: the measurement of the ratio of coherent to Compton scatter, and the measurement of the Compton scatter profile. Two possible applications are considered: the measurement of trabecular bone mineral density, and the measurement of the average fat/muscle ratio in a tissue volume. The results presented indicate that the analysis of coherent and Compton backscattered y-ray spectra from an -~4'Amsource has the potential for measuring both trabecular bone mineral density and average fat/muscle ratio in a tissue volume, with a low absorbed dose to the subject. © 1998 Elsevier Science Ltd. All rights reserved
Introduction The purpose of this study was to examine the potential of using backscatter information to assess regional body composition at selected sites. Two measurement techniques are examined: the measurement of the ratio of coherent to C o m p t o n scatter, and the measurement of the Compton scatter profile. Most of the previous work on coherent/Compton scattering (Ling et al., 1982; Puumalainen et al., 1982; Karellas et al., 1983) has employed smaller scattering angles. The C o m p t o n scatterer profile is sensitive to the chemical composition of the scatterer as a result of Doppler broadening due to the momentum distribution of the target electrons (Williams, 1977). Multiple scatter also influences the profile (MacKenzie, 1990; Tartari et al., 1992). Two possible applications are considered: the measurement of trabecular bone mineral density, and the measurement of the average fat/muscle ratio in a tissue volume. The results presented indicate that the analysis of coherent and Compton backscattered 7-ray spectra from an 24~Am source has the potential for measuring both trabecular bone mineral density and average fat/muscle ratio in a tissue volume, with a low absorbed dose to the subject.
ExperimentalArrangement The experimental arrangement for backscatter measurement is shown in Fig. 1. The radiation source is a collimated 7.4 GBq 24~Am point source. A 32 mm diameter 10 mm depth hyperpure germanium detector recorded the scattered photon spectrum. A *To whom all correspondence should be addressed.
backscatter measurement geometry has clear advantages in a clinical setting as it allows for a simple measurement set up and positioning relative to the patient. Various phantom materials were used to simulate different tissues. Aqueous solutions of K2HPO4 in the concentration range 0-90 g per 100 ml were used to simulate the density range of trabecular bone. The solutions were contained in thin-walled polythene bottles, 38 mm in diameter and 60 mm in height. Ethanol and water were used, respectively, as fat and muscle substitutes, and these were contained in a large polythene tank.
Bone MineralMeasurement Results To explore the potential of the technique for measurement of trabecular bone density at different depths in soft tissue, measurements were made with the bone phantom bottles placed first in air and then at different depths (35-60 mm to centre of bottle) in a water tank. Figure 2 shows how the measured coherent/Compton scatter ratio R varies with the concentration of K2HPO 4 for the bone phantom in air and at different depths in water. Measured Compton profiles show differences of up to 70 eV in F W H M values with different K_,HPO4 concentrations for the above phantom in air. However, the differences reduce to below a significant level when the bone phantom is measured in the water tank. Conclusions Over the range of expected trabecular bone mineral density the coherent/Compton scattering ratio R is a sensitive indicator of bone mineral density for peripheral bone locations. The technique is less 555
556
H.M. Morgan et al. 6
Fat/muscle phantom or water tank Bone
,
5
I~l
7
I
I
'
I
,
.
6
steelcollimator ~ HPGedetector~ L
30 degrees
~
P
b
J
.
.
_
.
'
/4 8
sliield
I Ain-241 source Pb collimator
Fig. 1. The experimental arrangement used for gamma-ray backscattering measurements. The Am-241 source has a 5 mm diameter lead collimator for all measurements. The hyperpure germanium detector was used with a 5 mm diameter collimator for the studies with bone mineral investigations, and a 10 mm diameter collimator for the studies of fat/muscle ratios.
sensitive for bone lying at depths beyond 50 m m due to the overlying scatter c o m p o n e n t s in the measured spectrum. Fat/Muscle
.
Ratio
Results
In order to explore whether the technique could distinguish clearly between muscle a n d fat, p h a n t o m s of water, ethanol (both in a polythene tank) and epoxy resin blocks were used as tissue substitutes. Figure 3 shows the dependence of the coherent scatter and C o m p t o n scatter on p h a n t o m thickness. The m a j o r c o n t r i b u t i o n to b o t h spectra is from material lying within 50 m m of the surface. Thus the technique would be useful for investigating the composition of tissue within 50 m m of the body surface, since 9 0 % of the counts arise from this region. The measured
i
0
0
~ 20
40
' ; 60 80 Depth, mm
~ 100
0 120
Fig. 3. Variation of backscatter with depth. Slabs of epoxy-resin phantom (300 mm x 300 mm) were used to increase the depth. The front surface of the phantom remained at a constant distance from the detector, 50 mm in front of the source~detector focus. Counts in the Compton region are denoted by ( 0 ) , and in the coherent region by (O).
values of the c o h e r e n t / C o m p t o n scatter ratio for depths more t h a n 50 m m represent an approximate average of the ratios in the first 50 mm. W a t e r (H: 11%, O: 89%) has a similar elemental composition to muscle (C: 14%, H: 10%, O: 71%, N: 3.4%) and ethanol - CzHsOH (C: 52%, H: 13%, O: 35%) has a similar elemental composition to fat (C: 60%, H: 11%, O: 2 8 % ) ( I C R U , 1989). F o r depths greater t h a n 50 m m the measured c o h e r e n t / C o m p t o n ratio R for water is 10.5 x 10 -4 and for ethanol 5.7 x 10 4, a difference of 85%. Differences were also detected in the C o m p t o n profiles. F o r 30 min acquisitions, the F W H M of the C o m p t o n profile for water is 2.02 _+ 0.005 keV, and for ethanol is 1.92 _+ 0.005 keV, a difference of 0.10 _+ 0.007 keV. Conclusions
0.012
Both the c o h e r e n t / C o m p t o n scatter ratio a n d the C o m p t o n profile shape are sensitive to the composition differences between ethanol and water, and might be explored as a means o f m o n i t o r i n g the p r o p o r t i o n of fat in a tissue volume.
0,010 o
0.008 o e~
0.006 Summary
0.004 o
0.002 0.000
,
0
'2'0 Concentration
40' K 2 H P O 4,
,
i
6'0 80 I00 g per 100 ml water
Fig. 2. Variation ofcoherent/Compton ratio with solutions of K2HPO4 in water of different concentrations, calculated values (©), in air (0), and at two depths in a water tank, 35mm(x)and 60mm (+).
The results presented indicate that the analysis of coherent a n d C o m p t o n backscattered y-ray spectra from a n 24'Am source has the potential for measuring b o t h trabecular bone mineral density a n d average fat/muscle ratio in a tissue volume, with a low a b s o r b e d dose to the subject. This work has been performed using tissue substitute materials. The next step is to construct a multi-source instrument suitable for clinical use. It will be tested first using animal tissues.
Gamma-ray backscatter for body composition Acknowledgement--J. T. Shakeshaft acknowledges support from the Wellcome Trust.
References ICRU (1989) Tissue substitutes in radiation dosimetry and measurement. Report 44, ICRU, Bethesda, MD. Karellas, A., Leichter, I., Cravan, J. D. and Greenfield. M. A. (1983) Characterization of tissue via coherent to Compton scattering ratio: sensitivity considerations. Medical Physics 10, 605. Ling, S.-S., Rustgi, S., Karellas, A., Craven, J. D., Whiting, J. S. and Greenfield, M. A. (1982) The measurement of trabecular bone mineral density using coherent and
557
Compton scattered photons in vitro. Medical Physics 9, 208. MacKenzie, I. K. (1990) An axially symmetric gamma ray backscatter system for Dumond spectroscopy. Nuclear Instruments and Methods A299, 377. Puumalainen, P., Uimarihuhta, A. and Olkkonen, H. (1982) A coherent/Compton scattering method employing an X-ray tube for measurement of trabecular bone mineral content. Physics in Medicine and Biology 27, 425. Tartari, A., Casnati, E., Felsteiner, J., Baraldi, C. and Singh, B. (1992) Feasibility of in vivo tissue characterisation by Compton scattering profile measurements. Nuclear Instruments and Methods B71, 209. Williams, B. G. (Ed.) (1977) Compton Scattering. McGrawHill, London.
Pergamon PIh
Appl. Radiat. Isot. Vol. 49, No. 5/6, pp. 555 557, 1998 ,~" 1998 Elsevier Science Ltd. All rights reserved Printed in Great Britain S0969-8043(97)00186-3 0969-8043/98 $19.00 + 0.00
Gamma-ray Backscatter for Body Composition Measurement H. M. M O R G A N * ,
J. T. S H A K E S H A F T
a n d S. C. L I L L I C R A P
Medical Physics Department, Royal United Hospital, Combe Park, Bath BA1 3NG, U.K. The purpose of this study was to examine the potential of using backscatter information to assess regional body composition at selected sites. Two measurement techniques are examined: the measurement of the ratio of coherent to Compton scatter, and the measurement of the Compton scatter profile. Two possible applications are considered: the measurement of trabecular bone mineral density, and the measurement of the average fat/muscle ratio in a tissue volume. The results presented indicate that the analysis of coherent and Compton backscattered y-ray spectra from an -~4'Amsource has the potential for measuring both trabecular bone mineral density and average fat/muscle ratio in a tissue volume, with a low absorbed dose to the subject. © 1998 Elsevier Science Ltd. All rights reserved
Introduction The purpose of this study was to examine the potential of using backscatter information to assess regional body composition at selected sites. Two measurement techniques are examined: the measurement of the ratio of coherent to C o m p t o n scatter, and the measurement of the Compton scatter profile. Most of the previous work on coherent/Compton scattering (Ling et al., 1982; Puumalainen et al., 1982; Karellas et al., 1983) has employed smaller scattering angles. The C o m p t o n scatterer profile is sensitive to the chemical composition of the scatterer as a result of Doppler broadening due to the momentum distribution of the target electrons (Williams, 1977). Multiple scatter also influences the profile (MacKenzie, 1990; Tartari et al., 1992). Two possible applications are considered: the measurement of trabecular bone mineral density, and the measurement of the average fat/muscle ratio in a tissue volume. The results presented indicate that the analysis of coherent and Compton backscattered 7-ray spectra from an 24~Am source has the potential for measuring both trabecular bone mineral density and average fat/muscle ratio in a tissue volume, with a low absorbed dose to the subject.
ExperimentalArrangement The experimental arrangement for backscatter measurement is shown in Fig. 1. The radiation source is a collimated 7.4 GBq 24~Am point source. A 32 mm diameter 10 mm depth hyperpure germanium detector recorded the scattered photon spectrum. A *To whom all correspondence should be addressed.
backscatter measurement geometry has clear advantages in a clinical setting as it allows for a simple measurement set up and positioning relative to the patient. Various phantom materials were used to simulate different tissues. Aqueous solutions of K2HPO4 in the concentration range 0-90 g per 100 ml were used to simulate the density range of trabecular bone. The solutions were contained in thin-walled polythene bottles, 38 mm in diameter and 60 mm in height. Ethanol and water were used, respectively, as fat and muscle substitutes, and these were contained in a large polythene tank.
Bone MineralMeasurement Results To explore the potential of the technique for measurement of trabecular bone density at different depths in soft tissue, measurements were made with the bone phantom bottles placed first in air and then at different depths (35-60 mm to centre of bottle) in a water tank. Figure 2 shows how the measured coherent/Compton scatter ratio R varies with the concentration of K2HPO 4 for the bone phantom in air and at different depths in water. Measured Compton profiles show differences of up to 70 eV in F W H M values with different K_,HPO4 concentrations for the above phantom in air. However, the differences reduce to below a significant level when the bone phantom is measured in the water tank. Conclusions Over the range of expected trabecular bone mineral density the coherent/Compton scattering ratio R is a sensitive indicator of bone mineral density for peripheral bone locations. The technique is less 555
556
H.M. Morgan et al. 6
Fat/muscle phantom or water tank Bone
,
5
I~l
7
I
I
'
I
,
.
6
steelcollimator ~ HPGedetector~ L
30 degrees
~
P
b
J
.
.
_
.
'
/4 8
sliield
I Ain-241 source Pb collimator
Fig. 1. The experimental arrangement used for gamma-ray backscattering measurements. The Am-241 source has a 5 mm diameter lead collimator for all measurements. The hyperpure germanium detector was used with a 5 mm diameter collimator for the studies with bone mineral investigations, and a 10 mm diameter collimator for the studies of fat/muscle ratios.
sensitive for bone lying at depths beyond 50 m m due to the overlying scatter c o m p o n e n t s in the measured spectrum. Fat/Muscle
.
Ratio
Results
In order to explore whether the technique could distinguish clearly between muscle a n d fat, p h a n t o m s of water, ethanol (both in a polythene tank) and epoxy resin blocks were used as tissue substitutes. Figure 3 shows the dependence of the coherent scatter and C o m p t o n scatter on p h a n t o m thickness. The m a j o r c o n t r i b u t i o n to b o t h spectra is from material lying within 50 m m of the surface. Thus the technique would be useful for investigating the composition of tissue within 50 m m of the body surface, since 9 0 % of the counts arise from this region. The measured
i
0
0
~ 20
40
' ; 60 80 Depth, mm
~ 100
0 120
Fig. 3. Variation of backscatter with depth. Slabs of epoxy-resin phantom (300 mm x 300 mm) were used to increase the depth. The front surface of the phantom remained at a constant distance from the detector, 50 mm in front of the source~detector focus. Counts in the Compton region are denoted by ( 0 ) , and in the coherent region by (O).
values of the c o h e r e n t / C o m p t o n scatter ratio for depths more t h a n 50 m m represent an approximate average of the ratios in the first 50 mm. W a t e r (H: 11%, O: 89%) has a similar elemental composition to muscle (C: 14%, H: 10%, O: 71%, N: 3.4%) and ethanol - CzHsOH (C: 52%, H: 13%, O: 35%) has a similar elemental composition to fat (C: 60%, H: 11%, O: 2 8 % ) ( I C R U , 1989). F o r depths greater t h a n 50 m m the measured c o h e r e n t / C o m p t o n ratio R for water is 10.5 x 10 -4 and for ethanol 5.7 x 10 4, a difference of 85%. Differences were also detected in the C o m p t o n profiles. F o r 30 min acquisitions, the F W H M of the C o m p t o n profile for water is 2.02 _+ 0.005 keV, and for ethanol is 1.92 _+ 0.005 keV, a difference of 0.10 _+ 0.007 keV. Conclusions
0.012
Both the c o h e r e n t / C o m p t o n scatter ratio a n d the C o m p t o n profile shape are sensitive to the composition differences between ethanol and water, and might be explored as a means o f m o n i t o r i n g the p r o p o r t i o n of fat in a tissue volume.
0,010 o
0.008 o e~
0.006 Summary
0.004 o
0.002 0.000
,
0
'2'0 Concentration
40' K 2 H P O 4,
,
i
6'0 80 I00 g per 100 ml water
Fig. 2. Variation ofcoherent/Compton ratio with solutions of K2HPO4 in water of different concentrations, calculated values (©), in air (0), and at two depths in a water tank, 35mm(x)and 60mm (+).
The results presented indicate that the analysis of coherent a n d C o m p t o n backscattered y-ray spectra from a n 24'Am source has the potential for measuring b o t h trabecular bone mineral density a n d average fat/muscle ratio in a tissue volume, with a low a b s o r b e d dose to the subject. This work has been performed using tissue substitute materials. The next step is to construct a multi-source instrument suitable for clinical use. It will be tested first using animal tissues.
Gamma-ray backscatter for body composition Acknowledgement--J. T. Shakeshaft acknowledges support from the Wellcome Trust.
References ICRU (1989) Tissue substitutes in radiation dosimetry and measurement. Report 44, ICRU, Bethesda, MD. Karellas, A., Leichter, I., Cravan, J. D. and Greenfield. M. A. (1983) Characterization of tissue via coherent to Compton scattering ratio: sensitivity considerations. Medical Physics 10, 605. Ling, S.-S., Rustgi, S., Karellas, A., Craven, J. D., Whiting, J. S. and Greenfield, M. A. (1982) The measurement of trabecular bone mineral density using coherent and
557
Compton scattered photons in vitro. Medical Physics 9, 208. MacKenzie, I. K. (1990) An axially symmetric gamma ray backscatter system for Dumond spectroscopy. Nuclear Instruments and Methods A299, 377. Puumalainen, P., Uimarihuhta, A. and Olkkonen, H. (1982) A coherent/Compton scattering method employing an X-ray tube for measurement of trabecular bone mineral content. Physics in Medicine and Biology 27, 425. Tartari, A., Casnati, E., Felsteiner, J., Baraldi, C. and Singh, B. (1992) Feasibility of in vivo tissue characterisation by Compton scattering profile measurements. Nuclear Instruments and Methods B71, 209. Williams, B. G. (Ed.) (1977) Compton Scattering. McGrawHill, London.