International Journal of Radiation Research, January 2014 Volume 12, No 1

Three-dimensional gel dosimetry for dose volume histogram verification in compensator-based IMRT

INTRODUCTION

Some tissues in human body areradiobiologicallydifferentfromwaterandtheseinhomogeneities must be considered in dosecalculationinordertoachieveanaccuratedosedelivery. In other words, to maximizetherapeutic bene it of radiation therapy,absorbed dose that would be delivered in thepresence of inhomogeneity must be predicted

accurately (1). Investigation of coincidence ofpredicted 3D dose distribution by treatmentplanning calculation with corresponding toactual delivered is one the most importantstages in radiation therapy (2, 3). Treatmentveri ication can be ful illed by many dosimetrytools.Dosimeterslikeionizationchambers,TLD,diode and ilm are dimensionally limited. Gelhave more dose sensitivity (slope of thecalibration curve) and higher dose sensitivity

M. Keshtkar1,2*, A. Takavar1, M.H. Zahmatkesh3, H.A. Nedaie1, A. Vaezzadeh1, M. Naderi4

1DepartmentofMedicalPhysicsandBiomedicalEngineering,TehranUniversityofMedicalSciences,Tehran,Iran

2DepartmentofMedicalPhysics,IsfahanUniversityofMedicalSciences,Isfahan,Iran3NovinInstituteofMedicalRadiation,ShahidBeheshtiUniversityofMedicalSciences,Tehran,Iran

4DepartmentofRadiotherapy‐Oncology,TehranUniversityofMedicalSciences,Tehran,Iran

ABSTRACT Background: Some ssues in human body are radiobiologically different from water and these inhomogeneity must be considered in dose calcula on in order to achieve an accurate dose delivery. Dose verifica on in complex radia on therapy techniques, such as intensity‐modulated radia on therapy (IMRT) calls for volumetric, ssue equivalent and energy independent dosimeter. The purpose of this study is to verify a compensator‐based IMRT plan in anthropomorphic inhomogeneous phantom by Dose Volume Histograms (DVH) using polymer gel dosimetry. Materials and Methods: An anthropomorphic pelvic phantom was constructed with places for gel inserts. Two a ached cubic inserts for prostate and bladder and a cylindrical insert for rectum. A prostate treatment case was simulated in the phantom and the treatment was delivered by a five field compensator‐based IMRT. Gel dosimeters were scanned by a 1.5 Tesla magne c resonance imaging (MRI). Results were analyzed by DVH and difference of differen al DVH. Results: Results showed for 3D compensator‐based IMRT treatment plan for prostate cancer, there was overall good agreement between calculated dose distribu ons and the corresponding gel measured especially in planning target volume (PTV) region. Conclusion: Our measurements showed that the used treatment plan configura on has had clinically acceptable accuracy and gel dosimetry can be considered as a useful tool for measuring DVH. It may also be used for quality assurance and compensator‐based IMRT treatment verifica on. Keywords: Heterogeneity, polymer gel dosimetry, compensator-based IMRT, dose verification.

*Correspondingauthor:Mr.MohammadKeshtkar,Fax:+982188973653E‐mail:Keshtkar.dmohammad@yahoo.com

Submitted: July 2013 Accepted: Oct. 2013

Int. J. Radiat. Res., January 2014; 12(1): 13-20

► Original article

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

Keshtkar et al. / 3D verification in compensator-based IMRT

dosimetry is capable to capture dosedistribution in three dimensions. Furthermore,geldosimetersaretissueequivalentandhavenosigni icant energydependence (4, 5). Thehistoryof development of gel dosimetry has beenmentioned in many papers (4), but the mostnotable development was in 2001, when Fonget al. introduced a new polymer gel dosimeterthat was fabricated in the presence of oxygenknownasMAGIC(MethacrylicandAscorbicacidin Gelatin Initiated by Copper) (6). Thisdevelopmentingeldosimetrypavedthewayforfabricatinggelsonthebenchtopinlaboratory.

There are many techniques for deliveringradiotherapytreatmentandintensitymodulatedradiation therapy (IMRT) is one of them.Although MLC‐based IMRT techniques are thewidely accepted techniques nowadays,compensator‐based IMRT is an alternativewayto deliver the intensity modulated treatment.Compensator‐based IMRT has advantages overMLC‐based IMRT such as simplicity, makingcontinuously varying intensity modulation,shortertreatmenttime,simpleandrapidqualityassurance, but the main disadvantage of thistechnique is lack of automation (7‐14).Compensators produce an optimized primaryluence pro ile at the patient's surface andperturbsbeambyhardeningtheprimaryphotonspectrumandgeneratingscatteredphotonsandelectrons(15, 16).Fromtheattenuationequationswith the consideration of beamdivergence andbeamhardening,thecompensatorthicknesscanbecalculatedandthenconstructioncanbedonebymillingmachine(10).Dose veri ication by gel dosimetry and its

applicationinIMRTandtomotherapyhavebeeninvestigatedbymanyresearchers (17‐25,3).Totheauthor'sknowledge,despiteofmanystudiesforMLC‐based IMRT, there are a few publicationsabout application of gel dosimetry incompensator‐basedIMRT (26,27)and itshouldbenoted that inhomogeneous phantom has notbeen used in these publications. In this paper,the goal is to verify three dimensionally acompensator‐based IMRT plan in cause moredose resolution (5); MAGIC gel and MRI wereemployedinthiswork.

MATERIALSANDMETHODS

PhantomdesignAn anthropomorphic inhomogeneous pelvic

phantombasedonCTslicesthatobtainedfromapatient study was designed and fabricated( igure1).Thephantomconsistsofsliceswhichare integrated to form a human pelvis.Individual slices were machined withcorresponding body contour, related organs(prostate,bladderandrectum)andpelvicbone( igure 2). Two attached cubic inserts withdimensionsof3.5×3.5×4cm3and6×6×7cm3forprostate and bladder respectively andcylindricalonewithdiameterof3cmand7cmheight for rectum (herein we call themorgan‐speci ic inserts) determined for geldosimetry.Itshouldbenotedthatdimensionsoforgan‐speci ic inserts were manufactured insuch a way that they would it in the pelvicphantom. The phantom and gel inserts weremade of Poly methyl‐methacrylate (PMMA),whilethepelvicboneandfemursweremadeofbone equivalent material,polytetra luoroethylene(PTFE).

Gelmanufacturing

ThecompositionofMAGICgelandprocedurefor its manufacturing was the same asmentionedinFongetal.(6).Duetothetoxicityofsomematerials;gelpreparationwascarriedoutin a fume cupboard. After preparation, theMAGIC gel was poured into organ‐speci icinsertsandcalibrationvials.Inordertoperformdose response calibration, a set of gel illedscrew‐top glass vials (inner diameter 14 mm,length 10 cm) was employed. All gels wereallowedtosetinarefrigerator.

Compensator‐basedIMRTtreatmentcaseandirradiation

For treatment planning, organ‐speci icinsertswere illedwithwaterandtheninsertedanthropomorphic phantom by polymer geldosimetry. Since the MAGIC gel‐MRI methodinto pelvic phantom. Because the gel is nearlywaterequivalent(6),itcanbeassumedthatithasno signi icant effect on the absorbed dose

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 14

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

Figure 1. Components of the fabricated phantom, note that cubic insert for prostate is posi oned and a ached to

bladder one in such away they make one integrated volume.

Figure 2. The opened slice anthropomorphic phantom.

calculations. Pelvic phantom was imaged by acomputerized tomography (CT) scanner. Threeiducialmarkerswerestuckon thephantomtosimplify positioning during CT scanning andIMRTdelivery.ThethicknessofCTsliceswas5mmand 48 imageswere acquiredwithout anygapbetweenslices.

For this work, TiGRT treatment planningsystem was employed. After importing the CTimages into the treatment planning software(TPS), structures were delineated; rectum incylindrical insert,bladderandprostate incubicinserts. Fractional doses were 8 Gy andtreatment with 18 MV photons was selected.

Then a compensator‐based intensitymodulatedtreatment plan (intensity‐map basedoptimization) with ive coplanar beams wasgenerated by TPS ( igure 3). Data related tocompensators were exported; these data foreachcompensatorconsistofaspreadsheetthatthe number value into each point shows theheight of compensator in that point. Formanufacturingcompensatormolds,thesevalueswere imported intoAutoCAD softwarepoint bypoint and the map of each compensator wasplannedina10×10cmplane.Accordingtothesemaps, laser cutting machine (CO2) cut slices ofmoldswithmaterialofPerspex;Perspexslicesofeach compensator were attached togethertightly, and then were illed by meltedcerrobend. After enough cooling, compensatorswereextracted.

Figure 3. Dose distribu on in a slice of pelvic phantom (Up) and isodose curves obtained by gel dosimeter (down).

Keshtkar et al. / 3D verification in compensator-based IMRT

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 15

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

Keshtkar et al. / 3D verification in compensator-based IMRT

Before irradiation, gels were stored inaccelerator room for ive hours. Irradiationaccording to the treatmentplanwasperformedadayaftergelpreparationusingaVarian2100linear accelerator and 18 MV photons. Thecalibration vials were placed in a water tank(40×40×40)andirradiatedwith18MVphotonsusing 20×20 open ield and SSD 100 cm.Delivered dose to calibration vials for doseresponseevaluationwas;1(2vials),2,4,6,7,8(2 vials), 9, 10 Gy. A pair of vials was leftunirradiated to serve as a control. Afterirradiation all gels were stored in refrigeratoragain.

MRIacquisition

Among many feasible methods for geldosimeter readout and dose mapping, MRI ismost popular. During irradiation spin‐latticerelaxation rate (R1) and spin‐spin relaxationrate (R2) change in gels as a function ofabsorbed dose, but R2 has a larger sensitivityanddynamicrange(4).

MRI imaging was performed 2 days afterirradiation using 1.5T scanner (Siemens,symphony) to ensure that polymerizationreactionwascompleted.GelswereplacedinMRIroomtoreachthermalequilibriumandimagingstarted5hours later.Thecalibrationvialswereattached to organ‐speci ic inserts and wereplaced in small water tank for increasing SNRthenwerepositioned at the center of head coiland imaged at the same time. The selectedimaging parameters for gels are as follows: theield‐of‐view (FOV) = 256mm × 256mm, slicethickness =5 mm without gap, TR = 5000 ms,echo spacing ∆TE= 22ms, voxel size = 1mm ×1mm × 5mm, NEX = 2, and the number ofechoes=32.

Imageanddataprocessing

After omitting the irst echo of the 32‐echotrain,theR2valueswerecomputedbyassumingan exponential decayof theMR signal using anin‐house MATLAB code (Mathworks, Inc.). TheR2valuesoftheimagesconvertedtodoseusingcalibrationequation.Full3Ddosedistributionofboth measurements and calculations wereprepared. On the one hand, because of the

different size of organ‐speci ic inserts andcalibration vials which cause difference intemperature rise during MRI scanning, andbecause of potentially higher oxygencontamination insmaller tubes (28), dosescalinginourstudyisneededtoadjustthedifferenceinradiation response of the gels with differentsizes. It should be noted that dose scaling wasapplied on measured dose by comparing themeasured dose distribution and correspondingtocalculatedinsidethetargetvolumeforrelativedosimetry(29).

Animportantwaytoevaluatecalculateddosedistribution, is dose‐volume histogram (DVH),and also criteria for the optimization of IMRTtreatment plans are often based on DVHconstraints (19). In thisstudy, theunique featureof gel dosimetry (i.e. true 3D dosimetry) wasused by calculating dose‐volume histogram(DVH)anddifferenceofdifferentialdose‐volumehistogram (DDDVH) for analyzing. DVH andDDDVH of all de ined organs both in calculatedandmeasureddatawaspreparedandcomparedtogether. For computing DVH, volumes ofinterestwerede ined inboth3Dmeasuredandcalculateddosedistributionsandvoxels thereinwere evaluated and then a histogram wasmounted.

RESULTS

Calibration data, obtained by the analysis ofR2mapsofthecalibrationscrew‐topglassvials,are shown in igure 4. After the regressionanalysis, obtained values are as followings: theslope a=0.838 (±3.02%)and the offset=5.17 (±2.28%). Also the coef icient of the determinantR2 and the standard deviation of the R2 valuewas0.9973andapproximately1%,respectively.

As aforementioned, we have limited ourresultstoDVHanddifferenceofdifferentialDVH(DDDVH). Differential dose‐volume histogram(DDVH)showsfrequencyofvoxelsasafunctionof speci ic dose. DDDVH is obtained bysubtractingthenumberofvoxelsindosebinsforthe measured dose from those for thecalculated dose (3). Note that in DDDVH,normalizationwasperformedasdosedifference

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 16

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

agreement between gel‐measured andcalculateddata.Figure(b)showsthatmeasureddose was greater than the calculated dosebetween0%(0Gy)to4%(0.32Gy)andwaslessthan thecalculateddosebetween4%(0.32Gy)to 10% (0.8 Gy). Beyond 10% (0.8 Gy) dosedifference in the DDVH is negligible. Thecalculatedmeanrelativedosetotherectumwas46±3% (1 SD); while corresponding gelmeasuredvaluewas44.8±3.3%(1SD).

Another critical organ is bladder that it'sDVHsandDDDVHshown in igure7.Accordingto DVHs, there are very good agreementbetween gel‐measured and calculated data.Figure 7(b) shows minor deviations thatmeasured dosewas greater than the calculateddose between 0% (0 Gy) to 10% (0.8 Gy).Beyond 10% (0.8 Gy) dose difference in theDDVHisnegligible.

ineachbindividedwithtotalnumberofvoxels.If bothmeasured and calculateddosehave thesame number of voxels with a speci ic doserangeinabin,itresultsavalueof0forthatbin.A bar or column height that has value of 0.01,means that there is 1% difference betweenmeasuredandcalculateddoses.

Figure 5 shows DVHs and DDDVH for PTV.Accordingtothese igures,therewasverygoodagreement between gel‐measured andcalculated data. In DDDVH igure, dosedifferences are approximately less than 1%.Notathat100%dosecorrespondstoprescribeddose8Gy.Thecalculatedmeanrelativedose tothe PTV was 100.1±2% (1 SD); whilecorresponding gel measured value was101±2.2%(1SD).

TheDVH andDDDVH related to rectum areshown in Figure 6. The DVHs show close

Figure 4.R2 as a func on of absorbed dose, the standard devia on of the R2 value was approximately 1%.

Figure 5. DVH (a) and difference in differen al dose‐volume histograms (DDDVH) (b) for PTV.

Keshtkar et al. / 3D verification in compensator-based IMRT

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 17

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

Keshtkar et al. / 3D verification in compensator-based IMRT

Figure 6. DVH (a) and difference in differen al dose‐volume histograms (DDDVH) (b) for rectum.

Figure 7.DVH (a) and difference in differen al dose‐volume histograms (DDDVH) (b) for bladder.

DISCUSSION

Comparison between calculated dosedistributionandcorrespondingtomeasurementcan be carried out by many ways. For currentstudy,weusedDVHandDDDVHforcomparison.According to results for PTV, there was highdegree of agreement between measured andcalculateddataforPTV.Alsoobtainedresultsfororgan at risks show overall good agreementbetweenmeasuredandcalculateddata.

Asmentioned in introduction, application ofgeldosimetry for treatmentplanveri ication inMLC‐basedIMRThasbeeninvestigatedbymanyresearchers in homogeneous phantom.Gustavsson etal. (19) investigated the feasibilityofusingnewtypegelforIMRTveri ication.They

illed a cylindrical glass lask with gel asphantom, a kidney‐shaped target was de inedand planning was based on sliding windowtechnique.According to their results, therewasgood agreement between measured endcalculateddosedistribution,discrepancieswerefoundinhotspotstheupperandlowerpartsofPTV and this was attributed to sub optimalscatterkernelsusedinTPS.

Sandilosetal.(24)forvalidatingTPS,captureddosedistributionbygeldosimetryforaprostateMLC‐based treatment plan con iguration andtheir results showed gel‐measured dosedistributions were adequately matched withcorrespondingTPScalculations.

Vergote et al. (25) used thorax phantom forvalidatingTPS inpresenceofair inhomogeneityandtheirresultsshowedanunderdosageof the

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 18

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

verify a compensator‐based IMRT plan ininhomogeneous phantom. According to theresults, itcanbeconcludedthatusedtreatmentplan con iguration in the presence ofinhomogeneity has clinically acceptableaccuracy and gel dosimetry is a useful tool formeasuring DVH to compare it with treatmentplanningsystemresults,henceitcanbeusedforcompensator‐basedIMRTtreatmentveri icationandqualityassurance.

REFERENCES

1. Papanikolaou N, Ba sta J, Boyer A, Kappas C, Klein E, Mackie T, Sharpe M, Van Dyk J (2004) Tissue inhomogeneity correc ons for megavoltage photon beams. AAPM Task Group, 65: 1‐142

2. Ezzell GA, Galvin JM, Low D, Palta JR, Rosen I, Sharpe MB, Xia P, Xiao Y, Xing L, Cedric XY (2003) Guidance document on delivery, treatment planning, and clinical implementa on of IMRT: Report of the IMRT subcommi ee of the AAPM radia on therapy commi ee. Medical Physics, 30:2089‐115.

3. Watanabe Y and Gopishankar N (2010) Three‐dimensional dosimetry of TomoTherapy by MRI‐based polymer gel technique. Journal of Applied Clinical Medical Physics, 12: (1) 3273.

4. Baldock C, De Deene Y, Doran S, Ibbo G, Jirasek A, Lepage M, McAuley KB, Oldham M, Schreiner LJ (2010) Polymer gel dosimetry. Physics in Medicine and Biology, [Review]. Mar 7;55 (5):R1‐63.

5. Deene YD (2004) Essen al characteris cs of polymer gel dosimeters. J Phys: Conf Ser, 3: 34‐57.

6. Fong PM, Keil DC, Does MD, Gore JC (2001) Polymer gels for magne c resonance imaging of radia on dose distribu ons at normal room atmosphere. Physics in Medicine and Biology, 46: 3105.

7. Brahme A (1988) Op miza on of sta onary and moving beam radia on therapy techniques. Radiotherapy and Oncology, 12(2):129‐40.

8. Chang S (2006) Compensator‐intensity‐modulated Radiotherapy—A tradi onal tool for modern applica on. US Oncological Disease, 115: 1‐4.

9. Chang SX, Cullip TJ, Deschesne KM (2000) Intensity modula on delivery techniques: "Step & Shoot” MLC auto‐sequence versus the use of a modulator. Medical Physics, 1(2): 80‐85.

10. Chang SX, Cullip TJ, Deschesne KM, Miller EP, Rosenman JG (2004) Compensators: an alterna ve IMRT delivery technique. Journal of Applied Clinical Medical Physics, 5(3): 15‐36.

11. Chang SX, Deschesne KM, Cullip TJ, Parker SA, Earnhart J (1999) A comparison of different intensity

PTVandconcluded thatpolymergeldosimetryisavaluabletechniquetoverifydosecalculationalgorithms for IMRT in heterogeneouscon igurations.

Farajollahi (26) assessed the accuracy of thepredicted dose distribution in prostatecompensator‐based IMRT plan using both geland ilm. A cylindrical homogeneous phantomwas used. The 90% isodose line from the ilmmeasurement was superimposed on the gelresults together with the calculated 90%isodose line. There was reasonable agreementbetweentwomethods,butbothshowedsimilardiscrepancies in some regions compared withthe calculations. Possible reasons for thesediscrepancieswereattributed toslighterror inthe settings of the beams or the positioning ofthe compensator, or more obviously in thealgorithmusedinthecalculation.

Haas (27) investigated to verify ancompensator‐basedIMRTplaninheadandneckphantom using ilm and gel dosimetry.There was good agreement between the dosedistribution obtained with ilm and geldosimetry, some discrepancies were observedbetween calculated and measured dosedistribution. These discrepancies wereattributed to compensator manufacturingprocess; this was because of limitation of themilling machine used to manufacture thecompensators in producing beam pro ileaccurately.

Comparisonbetweenour results and otherscon irms the applicability of gel dosimetry for3D dose veri ication in IMRT and showspredicted dose distribution by TiGRT in thepresenceofbone‐equivalentinhomogeneityhasacceptableaccuracy.

ItshouldbenotedthattheuseofglassvialsreducetheuncertaintyofR2calculationclosetovial edges and the use of smaller vials reducethecostrelativetousinglargervials.

CONCLSIONS

An experimental procedure based on theMRI‐polymergeldosimetrymethodwasusedto

Keshtkar et al. / 3D verification in compensator-based IMRT

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 19

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20

modula on treatment techniques for tangen al breast irradia on. Interna onal Journal of Radia on Oncology Biology Physics, 45(5):1305‐14.

12. Srivastava RP and De Wagter C (2007) The value of EDR2 film dosimetry in compensator‐based intensity modulated radia on therapy. Physics in medicine and biology, 52:N449.

13. Wendt TG, Abbasi‐Senger N, Salz H, Pinquart I, Koscielny S, Przetak SM, Wiezorek T (2006) 3D‐conformal‐intensity modulated radiotherapy with compensators for head and neck cancer: clinical results of normal ssue sparing. Radia on Oncology, 1(1): 1‐18.

14. Xu T, Shikhaliev PM, Al‐Ghazi M, Molloi S (2002) Reshapable physical modulator for intensity modulated radia on therapy. Medical physics, 29: 2222‐2229.

15. Jiang SB and Ayyangar KM (1998) On compensator design for photon beam intensity‐modulated conformal therapy. Medical physics, 25: 668‐75.

16. Vaezzadeh S, Allahverdi M, Nedaie HA, Ay M, Shirazi A, Yarahmadi M (2013) EBT GAFCHROMICTM film dosimetry in compensator‐based intensity modulated radia on therapy. Medical Dosimetry, 38(2): 176‐83.

17. Crescen RA, Scheib SG, Schneider U, Gianolini S (2007) Introducing gel dosimetry in a clinical environment: customiza on of polymer gel composi on and magne c resonance imaging parameters used for 3D dose verifica ons in radiosurgery and intensity modulated radiotherapy. Medical physics, 34: 1286‐97.

18. Gopishankar N, Vivekanandhan S, Rath G, Laviraj M, Senthilkumaran S, Kale S, Thulkar S, Bisht R, Subramani V (2013) Indigenously developed mul purpose acrylic head phantom for verifica on of IMRT using film and gel dosimetry. Journal of Applied Clinical Medical Physics, 14(2): 62‐76.

19. Gustavsson H, Karlsson A, Bäck SÅJ, Olsson LE, Haraldsson P, Engström P, Nyström H (2003) MAGIC‐type polymer gel for three‐dimensional dosimetry: Intensity‐modulated radia on therapy verifica on. Medical physics, 30: 1264‐71.

20. Karlsson A, Smulders B, Gustavsson H, Bäck SÅJ, editors. (2004) IMRT prostate dosimetry using a normoxic polymer gel and MRI. J Phys; Conf. Ser, 3: 284‐287.

21. Low DA, Dempsey JF, Venkatesan R, Mu c S, Markman J, Haacke EM, Purdy JA (1999) Evalua on of polymer gels and MRI as a 3‐D dosimeter for intensity‐modulated radia on therapy. Medical physics, 26: 1542‐51.

22. Oldham M, Gluckman G, Kim L, editors (2004) 3D verifica on of a prostate IMRT treatment by polymer gel‐dosimetry and op cal‐CT scanning: J Phys; Conf Ser, 3: 293‐296.

23. Pavoni J, Pike T, Snow J, DeWerd L, Baffa O (2012) Tomotherapy dose distribu on verifica on using MAGIC‐f polymer gel dosimetry. Medical Physics, 39: 2877‐84.

24. Sandilos P, Angelopoulos A, Baras P, Dardoufas K, Karaiskos P, Kipouros P, Kozicki M, Rosiak JM, Sakelliou L, Seimenis I (2004) Dose verifica on in clinical IMRT prostate incidents. Interna onal Journal of Radia on Oncology* Biology* Physics, 59(5):1540‐7.

25. Vergote K, De Deene Y, Claus F, De Gersem W, Van Duyse B, Paelinck L, Achten E, De Neve W, De Wagter C (2003) Applica on of monomer/polymer gel dosimetry to study the effects of ssue inhomogenei es on intensity‐modulated radia on therapy (IMRT) dose distribu ons. Radiotherapy and Oncology, 67(1):119‐28.

26. Farajollahi AR (1998) An inves ga on into the applica ons of polymer gel dosimetry in radiotherapy. Medical Physics, 26: 493

27. Haas O (2003) An intelligent oncology worksta on for the 21st century. NOWOTWORY, 53(4): 389‐97.

28. MacDougall N, Miquel M, Keevil S (2008) Effects of phantom volume and shape on the accuracy of MRI BANG gel dosimetry using BANG3TM. Bri sh Journal of Radiology, 81(961):46‐50.

29. Watanabe Y (2005) Variable transforma on of calibra on equa ons for radia on dosimetry. Physics in Medicine and Biology, 50(6): 1221‐34.

Keshtkar et al. / 3D verification in compensator-based IMRT

Int. J. Radiat. Res., Vol. 12 No. 1, January 2014 20

Dow

nloa

ded

from

ijrr

.com

at 2

2:47

+04

30 o

n T

hurs

day

June

11t

h 20

20