Design and evaluation of a HDR skin applicator with flattening filterD. Granero, J. Pérez-Calatayud, and J. GimenoRadiotherapy Department, La Fe University Hospital, E46009 Valencia, Spain

F. Ballestera� and E. CasalDepartment of Atomic, Molecular and Nuclear Physics, University of Valencia and IFIC-CSIC,E46100 Burjassot, Spain

V. CrispínFIVO, Fundación Instituto Valenciano de Oncología, E46009 Valencia, Spain

R. van der LaarseNucletron BV, 3900 AX Veenendaal, The Netherlands

�Received 19 April 2007; revised 8 November 2007; accepted for publication 17 November 2007;published 14 January 2008�

The purposes of this study are: �i� to design field flattening filters for the Leipzig applicators of 2and 3 cm of inner diameter with the source traveling parallel to the applicator contact surface,which are accessories of the microSelectron-HDR afterloader �Nucletron, Veenendaal, The Nether-lands�. These filters, made of tungsten, aim to flatten the heterogeneous dose distribution obtainedwith the Leipzig applicators. �ii� To estimate the dose rate distributions for these Leipzig+filterapplicators by means of the Monte Carlo �MC� method. �iii� To experimentally verify these distri-butions for prototypes of these new applicators, and �iv� to obtain the correspondence factors tomeasure the output of the applicators by the user using an insert into a well chamber. The MCGEANT4 code has been used to design the filters and to obtain the dose rate distributions in liquidwater for the two Leipzig+filter applicators. In order to validate this specific application and toguarantee that realistic source-applicator geometry has been considered, an experimental verifica-tion procedure was implemented in this study, in accordance with the updated recommendations ofthe American Association of Physicists in Medicine Task Group No. 43 U1 Report. Thermolumi-nescent dosimeters, radiochromic film, and a pin-point ionization chamber in a plastic �polymeth-ylmethacrylate �PMMA�� phantom were used to verify the MC results for the two applicators of amicroSelectron-HDR afterloader with the mHDR-v2 source. To verify the output of the Leipzig+filter applicators, correspondence factors were deduced for the well chambers HDR100-plus�Standard Imaging, Inc., Middleton, WI� and TM33004 �PTW, Freiburg, Germany� using a specificinsert for both applicators. The doses measured in the PMMA phantom agree within experimentaluncertainties with the dose obtained by the MC calculations. Percentage depth dose and off-axisprofiles were obtained normalized at a depth of 3 mm along the central applicator axis in acylindrical 20�20 cm water phantom. A table of output factors, normalized to 1 U of source airkerma strength at this depth, is presented. Correspondence factors were obtained for the two wellchambers considered. The matrix data obtained in the MC simulation with a grid separation of 0.5mm has been used to build a data set in a convenient format to model these distributions for routineuse with a brachytherapy treatment planning system. © 2008 American Association of Physicistsin Medicine. �DOI: 10.1118/1.2825622�

Key words: skin applicator, HDR, dosimetry, Monte Carlo, brachytherapy

I. INTRODUCTION The applicators are provided with a protective plastic cap,

Leipzig applicators, which are accessories for themicroSelectron-HDR system �Nucletron, Veenendaal, TheNetherlands� are used for treatment for small and shallowsuperficial malignancies in radiation oncology.1–5 These cup-shaped applicators are an alternative to superficial/orthovoltage x-ray treatment units. They limit the irradiationto the required area using tungsten shielding, allowing thetreatment of skin tumors, the oral cavity, vaginal cuff, etc. Aset of six applicators is commercially available with innerdiameters of 1, 2, and 3 cm where the source has either aparallel or perpendicular orientation with respect to the treat-

ment surface.

495 Med. Phys. 35 „2…, February 2008 0094-2405/2008/35

in contact with the surface of the skin, to reduce the surfacedose due to electron contamination in the walls. Lesionsshould be less than 25 mm in diameter, with a smooth sur-face, and located in a region where contact with the entireapplicator surface is possible.6 Due to the scatter in the ap-plicator wall and the air cavity between source and skin, thedose rate distributions are different from those of a free HDRsource in an unbounded water phantom assumed by the treat-ment planning system for dosimetry calculations.

The Leipzig applicators have been recently studied by Niuet al.4 and by Pérez-Calatayud et al.5 Pérez-Calatayud et al.

obtained the absolute dose rate distributions in liquid water

495„2…/495/9/$23.00 © 2008 Am. Assoc. Phys. Med.

496 Granero et al.: HDR skin applicators with flattening filter 496

for the whole set of applicators using the mHDR-v1 �classic�and mHDR-v2 source models7–9 in the microSelectron-HDRafterloaders. Moreover, Pérez-Calatayud et al. developed amethod, based on well chamber measurements with a spe-cific insert, which allows the output verification of theH-type Leipzig applicators on a routine basis.10,11

The Monte Carlo �MC� study of the Leipzig applicators�see Fig. 4 in Ref. 5� shows that the dose at a given depth ina plane which contains the symmetry axis of the applicatordecreases strongly with increasing distance to the symmetryaxis, resulting in a very heterogeneous dose distribution in aplane parallel to the skin surface, which has a large impacton clinical dosimetry usefulness. In this modality, theplanned target volume has a uniform thickness, so the pre-scription at a depth on axis does not guarantee that the entirelayer receives the prescribed dose. Moreover, with theLeipzig applicators a significant penumbra exists, reducingthe useful irradiation size of the applicators.

An option to improve the dose distribution could be theaddition of a flattening filter to equalize the dose laterally.This idea has been previously used for others surface appli-cators: a flattening filter applied to a skin brachytherapy ap-plicator using the Ir-192 HDR brachytherapy source was pre-sented by Kron et al.12 and a method for calculating theoptimal filter shape by Jeraj et al.13

The purpose of the present study is to design filters thatwhen added to the cup of the Leipzig applicators will pro-vide a uniform dose distribution in a plane parallel to thecontact surface at a given depth along the cup axis. Once thedesign of the filters has been completed, the absolute doserate distributions in liquid water for the whole set of Leipzigapplicators with filters will be obtained. This is done bymeans of an experimentally validated MC code. The doserate distributions obtained are used to generate a data set in aconvenient format to allow the modeling of these distribu-tions in a treatment planning system. Moreover, a simplemethod is given to allow the user to verify the output of theirLeipzig+filter applicators. It is based on a measurement in awell chamber with a specific insert10 from which follows acorrespondence factor. Such correspondence factors �CF� arepresented for two widely used well chambers.

II. METHODS AND MATERIALS

II.A. Design of the flattening filters

There are six Leipzig applicators with inner diameters of1, 2, and 3 cm. Three of them use a source moving parallel tothe treatment surface �H1, H2, and H3� while the remainingthree use a source which moves perpendicularly to the treat-ment surface �V1, V2, and V3�. Each one is provided with aprotective plastic cap 1 mm thick which is in contact with thesurface of the skin. Currently, the available source modelsthat can be used with our modified Leipzig applicators arethe mHDR-v1 “classic” and the mHDR-v2 sources.

The applicators for which filters have been designed arethe H2 and H3, taking into account the improvement in thedose distribution introduced by the filters with respect to

those of the Leipzig applicators and the frequency of use in

Medical Physics, Vol. 35, No. 2, February 2008

clinical practice. In the following these applicators with filterwill be named as Valencia applicators �VH2 and VH3�.Moreover, in the VH2 and VH3 applicators the manufacturer�Nucletron BV, Veenendaal, The Netherlands� fixated the po-sition of the source in the applicator cup in such a way, thatits accuracy is within 0.2 mm of the stated value.

The procedure used to design the flattening filters isdescribed to a great extent in the study of Jeraj et al.13

The method uses the fact that the prime contribution of theadded filter to the dose rate distributions will be the multi-plication of the original dose distribution by the exponentialexp�−�x�, � being the effective attenuation coefficient of thefilter material for the mean energy of the Ir-192 spectrum andx being the required filter thickness. Thus the dose will beattenuated by the exp�−�x� factor and by inverting it we willobtain a good approximation to the desired filter thickness.The material used in the filters are the tungsten alloy Den-simet 17 �W 90%, Ni 6%, Cu 4%, �=17 g cm−3�, that is, thesame material as the Leipzig applicators.

The final resulting shape of the filters and a schematicview of the Leipzig applicator with the filter are shown inFig. 1 and photography of the VH2 and VH3 applicators with

FIG. 1. Schematic view of the two Valencia horizontal-type applicator withthe source traveling parallel to the treatment surface. The coordinate systemwith the origin located at the center of the outer surface of the protectiveplastic cup is also shown.

the filter removed is shown in Fig. 2.

497 Granero et al.: HDR skin applicators with flattening filter 497

II.B. Monte Carlo study of the Valencia applicators

The 4.7.0 version of the MC GEANT4 code14 was used toobtain the air-kerma strength and the dose rate in wateraround the mHDR-v1 �classic� and mHDR-v2 source mod-els. This code and the complete process involved in analyz-ing the data were described in more detail in previouspapers.15–17 Compton scattering and the photoelectric effectfrom the standard electromagnetic package, as well as Ray-leigh scattering from the low energy package of GEANT4,were used. The cutoff energy was 10 keV for photons.

The reference system used in this study is shown in Fig.1, with the origin at the intersection of the applicator axisand the contact surface. The x and y axes are parallel to thesurface and the positive z axis is directed toward the phan-tom. In all simulations kerma was scored instead of dosebecause at the 192Ir energy kerma approximates dose for dis-tances greater than 1 mm from the source.18,19 The lineartrack-length kerma estimator20 was used.

II.C. Dose rate distributions in water of the VH2 andVH3 applicators

The air kerma strength of the mHDR-v1 �classic� andmHDR-v2 sources were obtained in a previous article.5 Ap-plicators were placed in contact with the base of a 20 cmhigh and 20 cm diameter cylindrical liquid water phantom torepresent a typical clinical situation. Although the applicatorsdo not present cylindrical symmetry, we showed in a previ-ous study5 that the dose rate distributions delivered by the

FIG. 2. Photography of the manufactured Valencia VH3 applicator showingthe tungsten filter. The filter is located in the window of the old Leipzigapplicator.

Leipzig H-type applicators closely approximate to cylindri-

Medical Physics, Vol. 35, No. 2, February 2008

cal symmetry with respect to the z axis. Therefore, if we alsotake into account that source-guide orientation cannot becontrolled in clinical routine, cylindrical symmetry has beenassumed for the VH2 and VH3 applicators. The MC dose

rate distributions D�y ,z� were obtained simulating 1.6�109

�VH3� and 1.4�109 �VH2� photon histories obtaining stan-dard deviations �k=1� of the mean dose rate values of lessthan 0.5%.

II.D. Experimental verification of MC application

The MC GEANT4 code14 is suitable for use in brachy-therapy applications.15–17,21 However, in order to validatethis specific application and to guarantee that a realisticsource-applicator-filter geometry has been considered, an ex-perimental verification procedure was included in this study,in accordance with the TG-43U1 recommendations.22 TLD-100 chip dosimeters, radiochromic film, and a pin-point ion-ization chamber were used to measure dose in a polymeth-ylmethacrylate �PMMA� phantom composed of slabs with atotal volume of 20�20�20 cm3. Irradiations were per-formed using a microSelectron-HDR unit with the mHDR-v2source. The air kerma strength Sk was measured using a spe-cific insert inside the HDR-1000 well chamber �Standard Im-aging� with an electrometer Max-4000 �Standard Imaging�,calibrated at ADCL Wisconsin traceable to NIST.

In order to compare experimental results with those ob-tained by the MC method, an additional MC study of themHDR-v2 source and the VH2 and VH3 applicators on thePMMA phantom was carried out. For this simulation, thesame electromagnetic processes, cutoff energy, number ofhistories, and grid system as described previously �Sec. II C�were considered.

II.E. Thermoluminescent dosimeter measurements

The TLD system used consists of a Harshaw 6600 TLDreader, together with Harshaw XD-100 extremity �EXT-RAD� dosimeters. The reader employs a hot gas heatingtechnique. Optimum heating cycle parameters were deter-mined as follows: preheat temperature, 160 °C; preheattime, 9 s; temperature rate, 12 °C /s; maximum temperature,300 °C; acquire time, 30 s; anneal temperature, 300 °C; andanneal time, 33 s. After readout, dosimeters were kept atroom temperature for 2 h followed by heating at 80 °C for18 h. The TL elements are LiF:Mg,Ti �TLD-100� hot pressedchips, with an area of 3�3 mm2 and a thickness of100 mg /cm2 �0.38 mm�. The chip is hermetically bonded toa substrate to which a unique barcode label is attached foridentification purposes.

The reader calibration was done using dosimeters chosenat random from the same batch. Those dosimeters were irra-diated at Co-60 energies and known doses around 50 cGy.Calibration dosimeters were read and annealed together withthe rest of the batch. The response of individual dosimeterswas corrected for differences in their relative sensitivity, us-ing an element correction factor determined by previous

measurements of the whole batch to a common dose.

498 Granero et al.: HDR skin applicators with flattening filter 498

For each applicator, dosimeters were irradiated at six cen-tral axis depths of 2.2, 5.2, 8.2, 11.2, 15.2, and 17.2 mm withone dosimeter at each depth. This procedure was repeatedseven times to evaluate the uncertainties at each measure-ment point. The doses to the dosimeters are in the range10–50 cGy for an irradiation time of 30 s. The transit sourcedose was evaluated measuring the dose at a fixed point withthe ionization chamber described below for various dwelltimes. It was concluded that the transit dose is negligible.The length parameter �the distance from the source to theindexer of the microSelectron-HDR unit� used was 1321 mmfor which the readings are on the center of the maximumplateau �Leipzig Applicators Instruction Manual, NucletronBV�.

To correct the experimental TLD readings for the overresponse of TLD at low energies, the energy spectrum in thePMMA phantom was estimated by means of the MC method.With the energy spectrum obtained at each dosimeter posi-tion and the TLD-100 dependence with energy data providedby Pradhan et al.23 correction factors due to energy depen-dence of the LiF dosimeters were calculated. This readingcorrection factor is only about 2.5% in relation to the readinggiven by the dosimeters calibrated in 60Co.

II.F. Radiochromic film measurements

Radiochromic film measurements have been done for theVH2 and VH3 applicators using the new EBT films �ISP,Wayne, NJ�. The EBT film has important advantages com-pared with other commercially existing films: �i� it is sensi-tive to lower doses �between 1 and 800 cGy�, �ii� its opticaldensity stabilizes in about 2 h when the film reaches theequilibrium state,24 �iii� it is not dependent on radiation en-ergy in the range from keV to MeV and thus degradation ofthe energy spectrum caused by scattered radiation is not animportant fact for this type of film,25 �iv� the film presents aspatial uniformity better than 1.5% and thus the use of thedouble exposition technique is not necessary, �v� it is insen-sitive to room light, and �vi� it can be used in a wide range oftemperatures �ISP product information�.

The EBT films have been irradiated using the cubicPMMA phantom described above in horizontal position withrespect to the VH2 and VH3 applicators at depths of 2, 4, 6,8, and 10 mm and performing three measurements with dif-ferent film pieces at each point to check if the dose ratedistributions are as flat as obtained with the MC simulations.The calibration of the film has been done by irradiating 40small pieces of EBT film in a Co-60 Phoenix machine todoses ranging from 0 to 1000 cGy. The film was read about70 h after the irradiation with a flat-bed document scannerAgfa Duoscan T 1200 using the same procedure as describedin other studies.26,27 Scanning of the films was done in trans-mission mode and using the 48 bit RGB mode. Only thechannel red of the RGB image has been used because itprovides better response than green and blue channels. Theanalysis of the images obtained with the scanner was done

using the software Adobe Photoshop and MATLAB.

Medical Physics, Vol. 35, No. 2, February 2008

II.G. Ionization chamber measurements

Measurements with the pin-point ionization chamber�PTW, Freiburg, Germany� were done at depths of 5, 7, 9, 11,13, and 15 mm in the PMMA phantom. This chamber has anactive volume of 0.015 cm3. The charge was integrated in anUNIDOS electrometer �PTW, Freiburg, Germany�. Crosscalibration of the chamber was done with a cobalt unit. Noenergy correction was applied.

II.H. Output verification of the Valencia applicatorswith a well chamber

To independently verify the output factors of the H-typeLeipzig applicator in the clinic, a measurement method basedon the use of a well chamber was proposed in a previousstudy.10 A correspondence was established between the doserate output in water of the Leipzig applicators �cGy/hU� andthe reading from a well chamber with a special insert, nor-malized to the HDR calibration factor with the standardHDR insert and the source strength. The CF were obtainedfor the same HDR-unit and H-type Leipzig applicators withwhich the experimental and MC study were performed.5

The correspondence factors CF �A2 h2 Gy−2 m−4� weredefined �10� as

CF =R � ��p,T�

f � Sk, �1�

where R �A� is the measurement using the special insert inthe well chamber entrance with the source in the position ofmaximum reading, ��p ,T� is the correction factor for theatmospheric conditions, f �Gy m2 h−1 A−1� is the well cham-ber calibration factor, and Sk �Gy m2 h−1� is the actual sourcestrength.

This method has been applied to verify the output of theVH2 and VH3 applicators using two well chambers: theHDR1000-plus model �Standard Imaging, Middleton, WI�and the TM33004 model �PTW, Freiburg, Germany�. An in-sert similar to that used for the HDR1000-plus chamber wasdesigned to be used with the TM33004 chamber.

II.I. The influence of the uncertainty of the sourceposition on the dose rate distributions

The filter has been designed assuming that the source islocated in its central position. Although the Valencia appli-cators are accessories of the microSelectron-HDR afterloaderwhich incorporate a precise system for positioning thesource, we considered the changes in the dose rate distribu-tion caused by a possible source displacement. Using MonteCarlo, the dose rate distributions have been estimated dis-placing the source 0.5 and 1 mm from its central position.The dose was scored in a three-dimensional �3D� way using

cubic cells.

499 Granero et al.: HDR skin applicators with flattening filter 499

III. RESULTS AND DISCUSSION

III.A. Monte Carlo dose rate distributions in liquidwater

MC simulations in a liquid water medium have been per-formed for both VH3 and VH2 applicators of themicroSelectron-HDR.

In Fig. 3 the dose rate distributions obtained by MC andnormalized to the dose at a depth of 0.3 cm are shown forboth VH2 and VH3 applicators.28 It is apparent from Fig. 3that the isodoses are nearly as flat as aimed by the design ofthe filters. In Fig. 4, dose distributions in a plane perpendicu-lar to the axis of the applicators are shown at differentdepths, showing a dose heterogeneity within 5%. The aniso-

FIG. 3. Relative dose distribution in water for the VH3 and VH2 Valenciaapplicators obtained by MC GEANT4 code. The isodose curves shown corre-spond to 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%,110%, and 120% with respect to values at a depth of 0.3 cm.

tropy in the dose rate distribution due to deviations in source

Medical Physics, Vol. 35, No. 2, February 2008

position has been estimated performing additional MonteCarlo simulation. Approximately, deviations of 0.5 and 1 mmin the source position produce asymmetry in the dose ratedistribution of 3% and 5%, respectively.

In Table I, output factors for both applicators are pre-

FIG. 4. Relative dose profiles at selected depths for the Valencia VH3 andVH2 applicators.

TABLE I. Output factors, per unit kerma strength, for the two Valencia ap-plicators and the mHDR-v2 source at 0.3 cm depth on the applicator centralaxis. Also the correspondence factors to verify the output of both applicatorsare given.

Dose @ 0.3 cm depth CF�cGy h−1 U−1� ��106 nA2 h2 Gy−2 m−4�

HDR1000-plus TM33004VH2 0.2245 0.737 0.144VH3 0.1659 0.822 0.128

500 Granero et al.: HDR skin applicators with flattening filter 500

sented. These output factors are defined as the dose incGy h−1 U−1 in the central axis of the applicator at a depth of0.3 cm.

III.B. Comparison of dose distribution for Leipzig andValencia applicators

In Fig. 5 the dose rate distributions obtained by MC andnormalized to the dose rate at a depth of 0.3 cm of the H3and H2 Leipzig applicator are compared to those of the VH3and VH2 applicators, respectively. Figure 6 compares thedose distribution of the Leipzig and Valencia applicator in aplane at a depth of 0.3 cm. Both figures show the improve-ment in the dose distribution of the Valencia applicators with

FIG. 5. Comparison of the relative dose distributions in water for the VH3and H3 �top� and VH2 and H2 �bottom� applicators obtained by MC GEANT4

code. The isodose curves shown correspond to 10%, 15%, 20%, 30%, 40%,50%, 60%, 70%, 80%, 90%, 100%, 110%, and 120% with respect to valuesat a depth of 0.3 cm.

respect to the Leipzig ones. Figure 7 compares the absolute

Medical Physics, Vol. 35, No. 2, February 2008

dose profiles along the symmetry axis of the applicators. Ob-serve that relative dose profiles of VH3 and H3 �or VH2 andH2� are nearly identical.

III.C. Experimental validation of the GEANT4 codesimulation

Figure 8 shows the dose calculated by means of MC alongthe symmetry axis of the applicators and the values obtainedin the measurements using TLD dosimeters, the pin-pointionization chamber, and EBT films in order to validate theMC algorithm.

The measurements using TLD dosimeters at differentpoints along the applicator axis are shown in Fig. 8. Thesystematic uncertainties of the TLD measurements were es-timated assuming an uncertainty of a 2.3% for the calibrationof TLD dosimeters with 60Co and an uncertainty of 3% for

FIG. 6. Comparison of relative dose profiles at z=0.3 cm depth for the VH3and H3 �top� and VH2 and H2 �bottom� applicators.

the measurement of air-kerma strength with a well chamber.

501 Granero et al.: HDR skin applicators with flattening filter 501

These estimations were made with the aid of MC calcula-tions. The statistical uncertainties due to successive measure-ments with TLD are between 1% and 4%. The quadraturesum of all of these uncertainties gives a global uncertainty ofbetween 6% and 8%. The measured doses in the PMMAphantom are in good agreement with the dose obtained bymeans of the MC GEANT4 code calculations, the differencesranging between −2.3% and 3.2%.

Figure 8 shows as well the dose measured using the pin-point ionization chamber at different points along the appli-cator axis. The uncertainties of the measurements are esti-mated to be of the order of 3%. The measurements are invery good agreement with the dose obtained by means of theMC GEANT4 code calculations, the differences ranging be-tween −2.2% and 2.6%.

Figure 8 also shows the results of the measurements with

FIG. 7. Comparison of absolute dose rate profiles along the applicator axisof the VH3 and H3 �top� and VH2 and H2 �bottom� applicators.

the EBT films, the differences with respect to MC varying

Medical Physics, Vol. 35, No. 2, February 2008

between −7.9% and 6.3%. The statistical uncertainty of thesemeasurements ranges from 1% to 9% depending on the mea-surement point.

Finally, Fig. 9 shows the dose in a plane at a depth of z=0.3 cm measured with the radiochromic EBT film showingthe very high degree of flatness of the dose rate distribution.

In conclusion, all these measurements verify that GEANT4

is validated for this study and that the geometrical designused to describe the VH2 and VH3 applicators for GEANT4 issuitable.

III.D. The influence of the uncertainty of the sourceposition on the dose rate distributions

The 3D dose distribution results of the Monte Carlo study

FIG. 8. Comparison of MC calculations �full lines�, EBT radiochromic film�circles�, TLD �triangle� and pin-point ionization chamber �inverted triangle�dose rate results along the central axis, all of them obtained in a 20�20�20 cm3 PMMA phantom.

show that if the source is displaced by 0.5 mm using the

502 Granero et al.: HDR skin applicators with flattening filter 502

VH3 applicator, an asymmetry of approximately 3% is ob-served in the dose rate distribution. This corresponds to anincrease of the dose in the forward direction and a decreasein the backward direction of the source. As for a displace-ment of 1 mm the asymmetry is about 5%. With the VH2applicator, the differences are lower: approximately 2% for a0.5 mm displacement and 4% for 1 mm. Figure 10 shows thevariations obtained for the VH3 applicator in the case of 0.5and 1 mm displacements.

One can note that the variations are produced in the di-rection of the source displacement in the applicator. Thus,although the positioning system of the microSelectron-HDRafterloader to which the Valencia applicators are connected isvery precise, the user should verify that the source is locatedin its central position.

III.E. Correspondence factors

The correspondence factors deduced from our measure-ments are presented in Table I. These factors provide asimple method to verify the output for the users of the Va-lencia applicators.

IV. CONCLUSIONS

In this study, tabulated dose rate distributions for ValenciaVH2 and VH3 applicators �which have an internal diameterof 2 and 3 cm with the source traveling parallel to the contactsurface� have been obtained using the MC GEANT4 code.

This application has been validated experimentally and arealistic geometry has been verified by means of TLD, radio-chromic, and ionization chamber measurements for the twoapplicators and the mHDR-v1 �classic� and mHDR-v2

FIG. 9. Dose rate distribution for the Valencia VH3 applicator in a plane at0.3 cm depth in the PMMA phantom obtained with the EBT film.

sources.

Medical Physics, Vol. 35, No. 2, February 2008

FIG. 10. Relative dose distributions obtained by MC for the VH3 applicatorin water. The y� axis represents the direction of displacement of the source.The isodose curves shown correspond to 40%, 60%, 80%, 100%, and 120%with respect to values at a depth of 0.3 cm. The figure includes the isodosecurves for the source displaced 0.5, 1 mm, and with the source centered.

503 Granero et al.: HDR skin applicators with flattening filter 503

Percentage depth dose and off-axis profiles normalized toa depth of 0.3 cm on the applicator central axis were ob-tained. A table of output factors at that depth is presentednormalized to 1 U of the source kerma strength. Moreover,correspondence factors to verify the output of the Valenciaapplicators have been deduced for two widely used wellchambers.

The matrix data obtained, with a grid separation of 0.5mm, have been used to construct a data set in a convenientformat to allow the modeling of these distributions in a treat-ment planning system.28 These data sets provide the requireddosimetry information needed for clinical treatment plan-ning.

ACKNOWLEDGMENTS

This work was supported by Nucletron which manufac-tured the Valencia applicator prototypes. This study was sup-ported in part by Ministerio de Educación y Ciencia �Spain�Project Nos. FPA2006-12120-C03-02, DPI2004-04268-C02-01, and FIS2004-05713; European Community Project No.FP6-2005-LIFESCIHEALTH-7 FP6-037555. The authors ac-knowledge the Fundación IVO �Valencia, Spain� for its re-search support.

a�Electronic mail: fballest@uv.es1M. D. Evans et al., “Dosimetric characteristics of surface applicators forhigh dose-rate brachytherapy �Abs�,” Med. Phys. 22, 671 �1995�.

2I. M. Hwang and H. W. C. Leung, “Dosimetry characteristics of Leipzigapplicators,” in Proceedings of the 1st Far East radiotherapy treatmentplanning workshop, edited by F. Mould and M. W. Gurtler, 1996, pp.88–89 �unpublished�.

3I. M. Hwang, S. Y. Lin, L. C. Lin, K. S. Chuang, and H. J. Ding, “Alter-native effective modality of Leipzig applicator with an electron beam forthe treatment of superficial malignancies,” Nucl. Instrum. Methods Phys.Res. A 508, 460–466 �2003�.

4H. Niu, W. C. His, J. C. H. Chu, and M. C. Kirk, “Dosimetric character-istics of the Leipzig surface applicators used in the high dose rate brachyradiotherapy,” Med. Phys. 31, 3372–3377 �2004�.

5J. Pérez-Calatayud, D. Granero, F. Ballester, V. Puchades, E. Casal, A.Soriano, and V. Crispín, “A dosimetric study of the Leipzig applicators,”Int. J. Radiat. Oncol. Biol. Phys. 62, 579–584 �2005�.

6M. D. C. Evans, M. Yassa, E. B. Podgorsak, T. N. Roman, L. J. Schreiner,and L. Souhami, “Surface applicators for high dose rate brachytherapy inaids-related Kaposi’s sarcoma,” Int. J. Radiat. Oncol. Biol. Phys. 39, 769–774 �1997�.

7J. F. Williamson and Z. Li, “Monte Carlo aided dosimetry of the microse-lectron pulsed and high dose-rate 192Ir sources,” Med. Phys. 22, 809–819�1995�.

8G. M. Daskalov, E. Löffler, and J. F. Williamson, “Monte Carlo-aideddosimetry of a new high dose-rate brachytherapy source,” Med. Phys. 25,2200–2208 �1998�.

9G. M. Daskalov, “Erratum: Monte Carlo-aided dosimetry of a new highdose-rate brachytherapy source �Med. Phys. 25, 2200–2208 �1998��,”

Medical Physics, Vol. 35, No. 2, February 2008

Med. Phys. 27, 1999 �2000�.10J. Pérez-Calatayud, D. Granero, F. Ballester, V. Crispín, and R. van der

Laarse, “Technique for routine output verification of Leipzig applicatorswith a well chamber,” Med. Phys. 33, 16–20 �2006�.

11J. Pérez-Calatayud, F. Ballester, D. Granero, V. Crispín, and R. van derLaarse, “Erratum: ‘Technique for routine output verification of Leipzigapplicators with a well chamber �Med. Phys. 33, 16–20 �2006��’,” Med.Phys. 33, 2654 �2006�.

12T. Kron, M. Haque, K. Foulkes, and R. Jeraj, “A flattening filter forbrachytherapy skin irradiation,” Phys. Med. Biol. 47, 713–722 �2002�.

13R. Jeraj, A. Sarvary, and T. Kron, “Optimal flattening filter shape of asurface Brachytherapy applicator,” Phys. Med. Biol. 47, 723–735 �2002�.

14S. Agostinelli et al., “GEANT4—A simulation toolkit,” Nucl. Instrum.Methods Phys. Res. A 506, 250–303 �2003�; see also http://geant4.web.cern.ch/geant4/.

15J. Pérez-Calatayud, D. Granero, F. Ballester, V. Puchades, and E. Casal,“Monte Carlo dosimetric characterization of the Cs-137 selectron/LDRsource: Evaluation of applicator attenuation and superposition approxima-tion effects,” Med. Phys. 31, 493–499 �2004�.

16J. Pérez-Calatayud, D. Granero, E. Casal, F. Ballester, and V. Puchades,“Monte Carlo and experimental derivation of TG43 dosimetric param-eters for CSM-type Cs-137 sources,” Med. Phys. 32, 28–36 �2005�.

17F. Ballester, D. Granero, J. Pérez-Calatayud, E. Casal, and V. Puchades,“Monte Carlo dosimetric study of Best Industries and Alpha Omega Ir-192 brachytherapy seeds,” Med. Phys. 31, 3298–3305 �2004�.

18R. Wang and X. A. Li, “A Monte Carlo calculation of dosimetric param-eters of 90Sr / 90Y and 192Ir SS sources for intravascular brachytherapy,”Med. Phys. 27, 2528–2535 �2000�.

19F. Ballester, C. Hernández, J. Pérez-Calatayud, and F. Lliso, “MonteCarlo calculation of dose rate distributions around 192Ir wires,” Med.Phys. 24, 1221–1228 �1997�.

20J. F. Williamson, “Monte Carlo evaluation of kerma at a point for photontransport problems,” Med. Phys. 14, 567–576 �1987�.

21J. Pérez-Calatayud, D. Granero, and F. Ballester, “Phantom size inbrachytherapy source dosimetric studies,” Med. Phys. 31, 2075–2081�2004�.

22M. J. Rivard, B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G.S. Ibbot, M. G. Mitch, R. Nath, and J. F. Williamson, “Update of AAPMtask group no. 43 report: A revised AAPM protocol for brachytherapydose calculations,” Med. Phys. 31, 633–674 �2004�.

23A. S. Pradhan and U. Quast, “In-phantom response of LiF TLD-100 fordosimetry of 192Ir HDR source,” Med. Phys. 27, 1025–1029 �2000�.

24A. Rink, I. A. Vitkin, and D. A. Jaffray, “Characterization and real-timeoptical measurements of the ionizing radiation dose response for a newradiochromic medium,” Med. Phys. 32, 2510–2516 �2005�.

25S. T. Chiu-Tsao, Y. Ho, R. Shankar, L. Wang, and L. B. Harrison, “Energydependence of response of new high sensitivity radiochromic films formegavoltage and kilovoltage radiation energies,” Med. Phys. 32, 3350–3354 �2005�.

26S. Devic, J. Seuntjens, E. Sham, E. B. Podgorsak, C. R. Schmidtlein, A.S. Kirov, and C. G. Soares, “Precise radiochromic film dosimetry using aflat bed document scanner,” Med. Phys. 32, 2245–2253 �2005�.

27G. Thomas, R. Y. L. Chiu, and F. Rabe, “A study of GafChromic XR typeR film response with reflective-type densitometers and economical flatbedscanners,” J. Appl. Clin. Med. Phys. 4, 307–314 �2003�.

28See EPAPS Document No. E-MPHYA6-35-005802 for detailed table re-sults �Excel for Windows spreadsheet�. This document can be reachedthrough a direct link in the online article’s HTML reference section or viathe EPAPS homepage �http://www.aip.org/pubservs/epaps.html�.