Spectra and air-kerma strength for encapsulated 192 Ir sourcesJette Borg and David W. O. Rogers

Citation: Medical Physics 26, 2441 (1999); doi: 10.1118/1.598763 View online: http://dx.doi.org/10.1118/1.598763 View Table of Contents: http://scitation.aip.org/content/aapm/journal/medphys/26/11?ver=pdfcov Published by the American Association of Physicists in Medicine Articles you may be interested in Assaying 192 Ir line sources using a standard length well chamber Med. Phys. 29, 2692 (2002); 10.1118/1.1517046 Dosimetry comparison of 192 Ir Sources Med. Phys. 29, 2239 (2002); 10.1118/1.1508378 Erratum: “Fitted dosimetric parameters of high dose-rate 192 Ir sources according to the AAPM TG43 formalism”[Med. Phys. 28(4), 654–660 (2001)] Med. Phys. 28, 1964 (2001); 10.1118/1.1398562 Fitted dosimetric parameters of high dose-rate 192 Ir sources according to the AAPM TG43 formalism Med. Phys. 28, 654 (2001); 10.1118/1.1359438 Dosimetric characteristics of a new high-intensity 192 Ir source for remote afterloading Med. Phys. 24, 2008 (1997); 10.1118/1.598114

Spectra and air-kerma strength for encapsulated 192Ir sourcesJette Borg and David W. O. Rogersa)

Institute for National Measurement Standards, National Research Council,Ottawa, Ontario K1A 0R6, Canada

~Received 4 May 1999; accepted for publication 16 August 1999!

The photon spectra in vacuum around four types of192Ir HDR brachytherapy sources are calculatedusing the Monte Carlo code EGS4 and the most recent spectral information on192Ir decay. Theair-kerma strengths per unit activity are calculated based on the photon fluence around a bare192Irsource and around each of four types of encapsulated sources using recent mass energy-absorptioncoefficients. For the full spectrum the bare vs encapsulated difference is up to 23% due to the largeair-kerma contribution from the unfiltered low-energy photons. For the penetrating part of thephoton spectrum~.11.3 keV!, the air-kerma strength per unit source activity on the transverse axisfor a bare source is 2–15 % higher than for the encapsulated sources due to the attenuation andabsorption in the core and the encapsulating material. The contribution to the air-kerma strengthfrom photons scattered in the capsule and from bremsstrahlung are calculated to increase theair-kerma strength by 2–4% and 0.2–0.3%, respectively. Air-kerma strengths for a variety ofsources agree well with previously reported results for sources from Nucletron International, BestIndustries, Inc., and Alpha-Omega Services, Inc. In addition we present air-kerma strengths for thepresent model of the HDR source from Nucletron International and the source from Varian Asso-ciates, Inc. @S0094-2405~99!00211-4#

Key words: 192Ir, brachytherapy sources, fluence spectra, air-kerma strength, Monte Carlo, EGS4

fh.m

tior-

udur

try-n

rgecar-re-aroim

esdffi

aro

ctralr isport

dha-rcerceew3.6al

of

heonity

edore

as

neat-m

I. INTRODUCTION

The recommended method of specifying source strengthbrachytherapy sources is in terms of air-kerma strengt1,2

However, currently no primary standard exists for air keror exposure from high-dose rate~HDR! 192Ir sources andthese sources are calibrated using an interpolatechnique.3–5 As part of a project to develop an ion-chambebased primary standard for192Ir HDR sources, detailedknowledge on the spectra outside different types of192Irbrachytherapy sources is required. A side result of the stof these spectra is that the air-kerma strength per unit soactivity can be calculated and compared.

EGS4 Monte Carlo calculations of the photon specfrom different types of192Ir seed sources were performed bThomasonet al.6,7 in 1989. Preliminary Monte Carlo calculations done at NRCC in Oct. 1992 to verify the spectra, ain particular to study the generation of photons with enebelow 10 keV, did not confirm the contribution from thlow-energy photons, which were present in the spectraculated by Thomasonet al. Since these could play an impotant role in a primary standard, further studies are requiIn 1994, Buermannet al.8 reported the air-kerma rate constants for the old microSelectron-HDR source and a b192Ir source. For both the old and the new microSelectsource, the VariSource, and two types of seed sources, slar Monte Carlo calculations are performed with the latspectral data for the192Ir nuclide published by Duchemin anCoursol9 and values of the mass energy-absorption coecients in air from Hubbell and Seltzer.10 Furthermore, thecontributions to the air-kerma strength from photons sctered in the core and the encapsulating material and f

2441 Med. Phys. 26 „11…, November 1999 0094-2405/99/26 „11

or

a

n

yce

a

dy

l-

d.

ren

i-t

-

t-m

bremsstrahlung are calculated, and the influence of spebin size on the calculated air kerma is studied. This papea much shortened version of a more extensive internal rewhich is available online.11

II. SOURCES

Four types of 192Ir sources are modelled—the seesources manufactured by Best Industries, Inc., and by AlpOmega Services, Inc., the microSelectron-HDR soumanufactured by Nucletron International, and the VariSoumanufactured by Varian Associates, Inc. The nmicroSelectron-HDR source is assumed to consist of amm long cylinder with diameter 0.65 mm of pure Ir metwith the radioactive192Ir uniformly distributed in it. Aroundthis core is a capsule with outer diameter of 0.9 mm madeAISI 316L steel, and connected to a 2.0 mm long~in themodel! steel cable with diameter 0.7 mm. The old type of tmicroSelectron-HDR source is also modelled for comparisof the calculated air-kerma strength per unit source activto the value reported by Bu¨ermannet al.8 This source, with adesign nearly identical to the source used by the GammaM12i afterloader, has a 3.5 mm long and 0.6 mm diameter cof pure Ir encapsulated in stainless steel.8,12

The 192Ir core of the VariSource is 10.0 mm long withdiameter of 0.34 mm~private communication with StavroPrionas, Varian Associates, Inc.!. The encapsulation is Niti-nol ~55.8% wt Ni and 44.2% wt Ti! with a density of 6.42g cm23. The diameter of the encapsulation is 0.6 mm and oend is covered by 0.1 mm Nitinol and the other end istached to a 2.5 m long Nitinol wire. In the model the 2.5

2441…/2441/4/$15.00

ta

mino

er0%m

flesn

sintresl

oRte

s-a

orchonu

od

,

fi-r

llsenva

e

e.la-theV–

rgy-hisod-ientin-Inc.,s

6%,in

theeVifactrgy

ma. It isdif-

ichre.t intionce

gniths

ewec-ed

ledum

Fors the

2442 J. Borg and D. Rogers: Spectra and air-kerma strength 2442

Nitinol wire is reduced to 1.9 cm, since scatter from the towire is insignificant.

The seed source from Best Industries, Inc., is 3.0 mlong, the core consisting of 30% Ir/70% Pt is 0.1 mmdiameter, and the cladding is 0.2 mm thick consistingstainless steel.2 The seed source from Alpha-Omega Svices, Inc., is also 3.0 mm long, the core consisting of 1Ir/90% Pt is 0.3 mm in diameter, and the cladding is 0.1 mthick consisting of 99.9% Pt.2 The total outer diameter oboth types of seed sources is 0.5 mm. All types of stainsteel in the sources are approximated by the compositioAISI 304 steel.12

III. CALCULATIONS

The photon fluence around the sources is calculated uthe NRCC user-code FLURZ which uses the EGS4 MoCarlo system.13 Each source is modelled as a cylindrical coregion with the particles being emitted from homogeneoudistributed points~using the uniform isotropically radiatingcylinder source routine in FLURZ!. The encapsulation is alscylindrical with the end caps for the microSelectron-HDand VariSource being disks. Also part of the cable connecto the sources is modelled.

In the Monte Carlo calculations K-shell x-ray flourecence and Rayleigh scattering are taken into account inregions. In the calculations with photons starting in the cthe energy cutoff for electron transport is 2.0 MeV, whimeans that all energy transferred to electrons will be depited at the point of the interaction and thus there areradiative losses. Photons are followed till they reach the coff energy of 0.001 MeV.

From the fluence calculated using the EGS4 user-cFLURZ, the air-kerma per initial photon,Kair8 , is calculatedusing the following discrete equation:

Kair8 51.602•10210(Emin

Emax

f8~Ei !Ei S men~Ei !

r D3DE @Gy ~initial photon!21], ~1!

where Ei is the midpoint of each energy binf8(Ei) (MeV21 cm22 photon21) is the differential photonfluence at energy Ei ~MeV! per initial photon,men(Ei)/r (cm2 g21) is the mass energy-absorption coefcient at energyEi , and DE is the bin size. The facto1.602310210 is required to convertKair8 from ~MeV g21! intoGy. This equation ignores any distinction between the cosion air kerma and the total air kerma since radiative losare negligible~,0.1%!. On average one decay will result ithe emission of 2.363 photons with energy in the interfrom a few keV to 885 keV,9 which includes photons directlyfollowing the b decay and also K- and L-shell x-rays. Thair-kerma strength per unit source activity,Sk /A, in~Gy m2 s21 Bq21! is calculated from:

Sk /A5Kair8 ~d!d2~2.36360.3%!, ~2!

or in ~U Bq21!

Sk /A53.63109Kair8 ~d!d2~2.36360.3%!, ~3!

Medical Physics, Vol. 26, No. 11, November 1999

l

f-

sof

nge

y

d

lle

s-ot-

e

i-s

l

whered is the distance to the cylindrical axis of the sourcChoosing the proper bin size in the Monte Carlo calcu

tions is important, since 88% of the photon energies forbare192Ir spectrum happen to be in the upper half of 10 kebins when divided into bins 0–10 keV, 10–20 keV,..., 890900 keV. The kerma is calculated using the mass eneabsorption coefficient at the middle of the energy bins. Tmay cause a binning artifact due to the variation of the pruct of photon energy and mass energy-absorption coefficwithin the bin. The binning artifact was studied for the staless steel encapsulated seed source from Best Industries,for bin sizes of 10, 5, and 2 keV. Scoring in 5 keV bininstead of 10 keV bins increases the air-kerma rate by 0.and scoring in 2 keV bins instead of 5 keV bins resultedan additional increase of 0.04%. However, the smallerbin size, the more memory is required. The bin size of 5 kis chosen as a compromise between smallest binning artand an acceptable amount of memory. Photons with enebelow 210 keV contribute less than 2% of the total air-kerstrength for the stainless steel encapsulated seed sourcethe photons with energy above 210 keV that cause theference in air kerma for different bin sizes.

To estimate the effects of bremsstrahlung from theb de-cay within the source, we do a separate calculation in whthe source is an isotropic source of electrons in the Ir coWe also estimate the bremsstrahlung from electrons semotion by the source photons but these lead to a contributo the air kerma of less than 0.05%. The photon fluenspectrum for bremsstrahlung from theb decay is scored in10 keV bins, which will not result in a significant binninartifact, since theb spectrum is continuous. The electrotransport for the calculation of bremsstrahlung is done wEGS4/PRESTA,14 and a kinetic energy cutoff for electronof 10 keV.

IV. RESULTS AND DISCUSSION

The fluence spectrum per decay calculated for the nmicroSelectron source type is shown in Fig. 1. Similar sptra are calculated for the VariSource and the two sesources and more information is given in the detaireport.11 The spectra are calculated with both air and vacu

FIG. 1. Fluence spectrum for the new type of the microSelectron source.fluence spectra for the other source types see Ref. 11 which also containdata in digital format.

rclc

pef

1

.u

ts

e

th

seesei--

onen-

hercein

ew.6–the%

dueuser-the

o-hellnd

ir-heysely,Fornd

ab-entnstheve an.a

r ari-seede toe

n-oldunit

ure-

the10

to

r-

iedblete.

stivy6

gthicot

d. T

n

y

2443 J. Borg and D. Rogers: Spectra and air-kerma strength 2443

surrounding the sources, and for the encapsulated southere is no significant difference between the spectra calated in air and in vacuum.

In Table I the calculated values of air-kerma strengthsunit source activity are shown as an average of the valuesdistances ranging from 2~5 cm for the VariSource! to 50 cmfrom the cylinder axis of the source. The uncertainty isstandard deviation based on the 5 calculatedSk /A values forthese distances not taking into account their uncertaintiesthese distances the geometry factor, i.e., the factor accoing for the source being a line and not a point, timesd2 is 1.0within 0.4% except for the VariSource, where the distancethe source must be at least 4 cm for the source to be conered a point source.15 The values in Table I also includbremsstrahlung contributions of 0.2% of theSk /A values forthe microSelectron-HDR source, the VariSource andplatinum encapsulated seed source~Alpha-Omega! and of0.3% of the value for the stainless steel encapsulatedsource~Best Industries!. The results of earlier work on thbare source, the old microSelectron source and the twosources are shown as well~these do not include the contrbution from bremsstrahlung!. The lower energies of the spec

TABLE I. Air-kerma strength per unit source activity for different sourceThe data are average values of the air-kerma strength per unit source acbased on fluence spectra at distances ranging from 2 to 50 cm, and thecalculated for lower cutoff energies of the fluence spectra of 11.3 andkeV for comparison with other reported values. The air-kerma strencalculated in this work include the contribution from bremsstrahlung, whwas not taken into account in the earlier results. The value of the texposure-rate constant calculated by Glasgow and Dillman~Ref. 16! is cor-rected by 33.97/33.7, since the energy required to produce an ion pair inair has been re-evaluated since 1979 and their value was for humid airuncertainties are 1 standard deviation statistical uncertainties~note that someare given as absolute values and some as percent! except for the values fromBuermannet al. ~Ref. 8! which include uncertainties due to the interactiocoefficients used in the Monte Carlo calculations~1%!, uncertainty on thesource geometry~0.5%!, and uncertainty on the distribution of the activitwithin the core~0.5%!.

Source

This workSk /A

(1028 U Bq21)

OthersSk /A

(1028 U Bq21)

New microSelectron.11.3 keV 9.7360.01.60 keV 9.7060.01Old microSelectron.11.3 keV 9.7960.02.60 keV 9.7760.02 9.861.5%8

VariSource.11.3 keV 10.2560.02.60 keV 10.2260.02Best Industries.11.3 keV 10.7060.02 10.860.16

.60 keV 10.6860.02Alpha-Omega.11.3 keV 9.9260.02 9.9560.146

.60 keV 9.9260.02Bare source.11.3 keV 11.23 11.2016

11.236

.60 keV 10.88 11.061%8

Medical Physics, Vol. 26, No. 11, November 1999

esu-

ror

Atnt-

oid-

e

ed

ed

tra, i.e., 11.3 keV and 60 keV, for the air-kerma calculatiare used for comparison with earlier works. Photons ofergy less than 11.3 keV are considered nonpenetrating.16

The air-kerma strengths per unit source activity for told microSelectron-HDR source and for the bare souagree well with reported results from other authors withtheir stated uncertainties of 1–1.5%. For the nmicroSelectron-HDR source the air-kerma strength is 00.7% lower than for the old one. For the seed sourcesair-kerma strengths calculated in this work are 0.3–1lower than the results of Thomasonet al. However, our val-ues are expected to be about 1% higher than their resultsto our added contribution from bremsstrahlung and becawe scored in 5 keV bins and not in 10 keV bins. The diffeence is attributed to the differences in spectral data for192Ir nuclide and in the (men /r) values.

The Monte Carlo calculations include K-shell x-ray flurescence but no L-shell x-rays. The energies of the K-sx-rays are 76.1 and 78.4 keV for Ir and Pt, respectively, a7.11 keV for stainless steel~mainly Fe!. A calculation with-out x-ray fluorescence shows that the contribution to akerma strength from K-shell x-rays is about 0.2% for tmicroSelectron-HDR source. In Ir and Pt the L-shell x-rahave energies below 13.4 keV and 13.9 keV, respectivand for the stainless steel the energy is below 0.85 keV.stainless steel the L fluorescence yield is practically 0, abecause of the low energy of the photons they will besorbed within 1 cm of air and not show up in a measuremof air kerma. For L-shell x-rays from Ir and Pt no photocreated in the core will pass through the encapsulation ofsources, and those created in the Pt encapsulation will hahigh probability of undergoing a photoelectric interactioThe contribution from L-shell x-rays to the air-kermstrength is thus likely negligible.

V. CONCLUSIONS

The air-kerma strength per unit activity is calculated fobare192Ir source, the microSelectron-HDR source, the VaSource, and stainless steel and platinum encapsulatedsources at distances ranging from the surface of the sourc50 cm in both vacuum and air. The result for thmicroSelectron-HDR source~old type! is in agreement withthe values from Bu¨ermannet al.8 in 1994. The air-kermastrength per unit activity for the new type of microSelectroHDR source is about 0.6–0.7% lower compared to thetype. For the seed sources the air-kerma strengths persource activity are less than 1% smaller than the exposrate constants calculated by Thomasonet al.6,7 in 1989.However, due to the bremsstrahlung contribution andbinning artifact, i.e., the difference between scoring inkeV bins instead of 5 keV bins, our values were expectedbe about 1% higher than the values of Thomasonet al. Thedifferences in data for the primary192Ir spectrum and massenergy-absorption coefficients for dry air explain this diffeence.

The effect of bin size for scoring the fluence was studto reduce the binning artifact as well as using an acceptaamount of memory. A bin size of 5 keV was found adequa

.ityare0s

hal

ryhe

5%nata

su

ir-rcsehome

-va

ajet

innkog

M

ndc-

ou

ali-

a-

e-sam-

ces,

ce

ityntsds

o-forr-

99,

ec-

i-

detan-

cedrt,

rials.

e

2444 J. Borg and D. Rogers: Spectra and air-kerma strength 2444

For the penetrating part of the photon spectrum~.11.3keV!, the air-kerma strength for the bare source is 2–1higher than for the encapsulated sources due to the attetion and absorption in the core and the encapsulating mrial. The bremsstrahlung contribution to the air-kermstrength is calculated for the four sources and increasekerma strength by 0.2–0.3%. Photons scattered in the socontribute 2–4% of the total air-kerma strength.

The contribution of the low-energy photons to the akerma strength was studied for the encapsulated souEliminating photons with energy less than 60 keV decreathe air-kerma strength by 0.2–0.3% and eliminating the ptons with energy less than 130 keV results in the air-kerstrength being reduced by 1%. The contribution from L-shx-rays to the air-kerma strength is considered negligible.

The major purpose of this work is to provide reliable ‘‘inair’’ spectra as presented in Fig. 1. These spectra are aable in digital format online in Ref. 11.

ACKNOWLEDGMENTS

We are grateful to Jan Seuntjens, Ionizing Radiation Stdards, NRCC, for ideas and discussions through this proand for commenting on this manuscript. We also wishthank Stavros Prionas, Varian Associates, Inc., for providus with information about the VariSource. Finally, we thathe referee for valuable comments on the paper. This wwas supported by the National Institutes of Health throuGrant No. R01 CA66852.

a!Electronic mail: dave.rogers@nrc.ca; Tel:~613! 993-2715; Fax:~613!952-98651AAPM TG–32, Specification of Brachytherapy Source Strength, AAPReport 21, AAPM, New York, 1987.

2R. Nath, L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, aA. S. Meigooni, ‘‘Dosimetry of interstitial brachytherapy sources: Reommendations of the AAPM Radiation Therapy Committee Task GrNo. 43,’’ Med. Phys.22, 209–234~1995!.

Medical Physics, Vol. 26, No. 11, November 1999

ua-e-

airrce

es.s-all

il-

n-ctog

rkh

p

3S. J. Goetsch, F. H. Attix, D. W. Pearson, and B. R. Thomadsen, ‘‘Cbration of 192Ir high-dose-rate afterloading systems,’’ Med. Phys.18,462–467~1991!.

4M. J. Rossiter, T. T. Williams, and G. A. Bass, ‘‘Air kerma rate calibrtion of small sources of60Co, 137Cs, 226Ra, and192Ir, ’’ Phys. Med. Biol.36, 279–284~1991!.

5F. Verhaegen, E. van Dijk, H. Thierens, A. H. L. Aalbers, and J. Suntjens, ‘‘Calibration of low activity192Ir brachytherapy sources in termof reference air-kerma rate with large volume spherical ionisation chbers,’’ Phys. Med. Biol.37, 2071–2082~1992!.

6C. Thomason, Dosimetry of Iridium-192 and Cesium-137 seed sourPh.D. thesis, University of Wisconsin, Madison, 1989.

7C. Thomason, T. R. Mackie, and M. J. Lindstrom, ‘‘Effect of sourencapsulation on the energy spectra of192Ir and 137Cs seed sources,’’Phys. Med. Biol.36, 495–505~1991!.

8L. Buermann, H.-M. Kramer, H. Schrader, and H.-J. Selbach, ‘‘Activdetermination of192Ir solid sources by ionization chamber measuremeusing calculated corrections for self-absorption,’’ Nucl. Instrum. MethoA339, 369–376~1994!.

9B. Duchemin and N. Coursol, Reevalution de l’192Ir, Technical NoteLPRI/93/018, DAMRI, CEA, France, 1993.

10J. H. Hubbell and S. M. Seltzer, Tables of X-ray mass Attenuation Cefficients and Mass Energy-Absorption Coefficients 1 keV to 20 MeVElementsZ51 to 92 and 48 Additional substances of Dosimetric Inteest, Technical Report NISTIR 5632, NIST, Gaithersburg, MD 2081995.

11J. Borg and D. W. O. Rogers, Monte Carlo Calculations of Photon Sptra in Air from 192Ir Sources, NRC Report PIRS-629r~see http://www.irs.inms.nrc.ca/inms/irs/papers/PIRS629r/pirs629r.html!, 1998.

12J. F. Williamson and Z. Li, ‘‘Monte Carlo aided dosimetry of the mcroselectron pulsed and high dose-rate192Ir sources,’’ Med. Phys.22,809–819~1995!.

13W. R. Nelson, H. Hirayama, and D. W. O. Rogers, The EGS4 CoSystem, Report SLAC-265, Stanford Linear Accelerator Center, Sford, California, 1985.

14A. F. Bielajew and D. W. O. Rogers, PRESTA: The Parameter ReduElectron-Step Transport Algorithm for electron Monte Carlo transpoNational Research Council of Canada Report PIRS-0042, 1986.

15R. Wang and R. S. Sloboda, ‘‘Influence of source geometry and mateon the transverse axis dosimetry of192Ir brachytherapy sources,’’ PhysMed. Biol. 43, 37–48~1998!.

16G. P. Glasgow and L. T. Dillman, ‘‘Specificg-ray constant and exposurrate constant of192Ir, ’’ Med. Phys.6, 49–52~1979!.