Radio! Oncol 1996; 3(): 138-41. Ionization gradient chamber in absolute photon and electron dosimetry Corey Zankowski and Ervin B. Podgorsak McGill University, Department of Medical Physics, Montreal General Hospital, 1650 avenue Cedar, Montréal, Québec H3G 1A4 A variable volume parallel-plate ionizalion gradienl chamber wax built to determine the absorbed dose in a polystyrene phantom. The sensilive volume of the gradient chamber is conlrolled by moving the chamber pislon by means of ci micrometer mourned to the phantom body. The displacement of the pislon is monitored by ci calibrated distance lmvel indicator which is accurate to within O.OJ mm. Irradialions were carried out wilh coball-60 gamma rays, photon beams rangingfmm 4 MV to 18 MV, and election beams between 5 MeV w 18 MeV. With the ionization gradienl chamber the calctt!ation of the absolute dose al ci given depth in phantom is simple and based on Jirsl principles using lhe slope of the measured ionization as a function of the electrode separation, i.e., lhe sensitive air volume. The discrepancies between lhe doses determined with our uncalibral-ed gradienl chamber and those oblained with a calibrated standard chamber are at most 1.08 % and 0.63 % for ¡photon and electron beams, respectively, ciI all clinical energies, indicating lhal the gradienl ionization chamber can be used as an abso!itle dosimeter. Key words: radiation dosage; polystyrene; photons; electrons; absolute dosimetry Introduction An accurate determination of the absolute dose rale produced by photon or electron machines is one of the most important componenls of modem radiotherapy. Radiotherapy clinics most commonly determine the absolute absorbed dose with parallelplate or cylindrical ionization chambers which are lirsl calibrated at, or trace their calibration factors to. a national standards laboratory. The dose is calculated from the measured ionizalion in air using the chamber calibration factor and following one of several available protocols (e.g., ICRU.1 AAPM-TG21 :2 AAPM-TG25;3 IAEA-WH0;4 elc.) These protocols are based on the standard Bragg-Gray5 '' or Spencer-Attix7 cavily theories and incorporate Correspondence lo: Prof. Ervin B. Podgorsak, Ph. D., FCCPM, Director. McGill Universily, Department of Medical Physics. Montreal General Hospital. 16.50 avenue Cedar, Montreal. Quebec H3G 1A4; Phone: +1 514 934 80.52; Fax + 1 514 934 8229. UDC: 539.166.08 various correction faclors, which are used to ac-counl for effects of chamber dimensions and wall materials as well as disruptions in the photon and electron fluence caused by lhe chamber. These correction factors make the dose determination cumbersome and inlroduce uncertainties in lhe final resull. The basic Bragg-Gray and Spencer-Attix cavity relationships for the dose Din medium are: D,„d= — w,„rsz'1 (l) m and Dmi,, = — Wu,r Lit, (2) m respectively, where Q is the charge collected under saturation conditions in the sensitive chamber air mass m, W,ii,- = 33.97 eV' is the mean energy re- Ionization gradient chamber in abso/u/e photon and election dosimetry 139 quired to produce an ion pair in air, and S,»"' and L','/i-'' are the ratios of unrestricted and restricted collisional stopping powers, respectively, for the medium and air tor the electron spectrum at the position of the cavity. Both the Bragg-Gray and the Spencer-Attix formalisms assume that the air cavity within the medium is sufficiently small such that it cloes not aller the electron fluence in the medium. The Bragg-Gray formalism uses unreslrict-ed slopping powers averaged over the slowing-down spectrum of only the primary electrons, while the Spencer-Allix formalism uses reslricted slopping powers averaged over the slowing-down spectrum of ali generations of electrons. It is evidenl from Equations (1) and (2) that the dose in medium is proporlional to the measured ratio Q/m which in principle should be straighl forward to determine. In actualily, Q is easy to measure accurately in clinical beams, however, in is al-mosl impossible to determine with an accuracy of better than 1 % required tor clinical use, precluding the direct use of Equations (1) and (2) in absolute dosimelry. The standard method tor obviating this problem is to. calibrate the cavily chamber response in a known reference radiation field which has been calibrated previously with a standard free air ionization chamber. This determination of the chamber calibration factor is actually an indirecl means of delermining the mass of air in the chamber sensitive volume. The chamber calibration factor in conjunction with various troublesome correction factors is then used to determine the dose to the medium. Invesligation of Equations (1) and (2) has revealed that al leasl tor small in the ratio of Q/m is a constanl allowing ils replacement wilh the derivate tlQ/dm, resulling in lhe following modified Bragg-Gray and Spencer-Allix relationships for lhe dose in medium: and (Iin fXn = Wm LZ' (tin (3) (4) developed an uncalibrated, variable volume, ionization gradient chamber (JGC) capable of measuring the absorbed close directly in an absolute manner. The chamber developed by Klevenhagen was made of Lucite and required the use of a water tenk for close measurement; therefore, corrections for the density and fluence differences between Lucite and waler had to be considered. Our chamber material is the same as the phantom material (polystyrene); consequently, there is no need for such corrections to the measured signal when determining the absorbed close in polystyrene. The determination of the absolute absorbed dose tor clinical photon and electron beams at a given depth in phantom with the IGC is based on first principles, is simple to evaluale, and agrees well with results obtainecl wilh standard calibrated ionization chamber techniques. Materials and methods A 7 cm diameter polystyrene piston was fashioned to move inside a cylinder bored along the center of a 30 x 30 x 8 cm1 polystyrene phantom. Graphite dag was painted on the top surface of lhe piston, and a 1.5 mm deep and 0.04 mm wide groove was cul through the graphite surface inlo the piston to form lhe 2.004 (1 ± 0.001) cm inner diameler measuring electrode and the guard ring of lhe chamber. The measuring electrode and lhe guard ring are bolh connected to ground (the measuring electrode ihrough an eleclromeler) wilh eleclronically shield- polystyrene entrance window ee:;:^IarizingeIectrode polystyrene phantom Figure l. .Schematic diagram of lhe ionization gradient chamber. The advantage to lhis approach is lhat, in conlrasl 10 Q/m, dQ/dm is relalively easily measured accurately making the modified Bragg-Gray and Spen-cer-Altix relationships direclly applicable in absolute dosimelry. Similarly lo Klevenhagen,9 we have ed cables. The polarizing electrode consisls of a 0.5 mm lhick polyslyrene disk painled wilh graphite dag and fastened to the top of the large phantom. The electronic potential of the polarizing electrode is maintained al ±400 V with respect to the collecting 140 Zaiikowski C and Poclgoršak EB electrode. The separation between lhc polarizing and measuring electrodes can vary between 0.5 mm and I O mm, and is controlled by a micrometer mounted to the phantom body. The movement of the piston (i.e., change in the air sensitive volume) is monitored by a calibrated distance travel indicator which is accurate to within O.OI mm. In Figure I we show a schematic diagram of the IGC. Irradiations of the gradient chamber were performed with a cobalt-60 gamma source, photon beams in the energy range from 4 MV to 18 MV. and electron beams in the nominal energy range from 9 MeV to 18 MeV. Results and discussion The specific design of our IGC allows us to determine dQ/dm of Eq. (4) with relative ease and a high degree of accuracy. Since dm is directly proportional to ihe change dz in electrode separation, we can write Eq. (4) as follows: ......... f *..... (5) with p the density of air at the ambient temperature and pressure, and A the area of the measuring electrode. 0 1 2 3 4 5 Electrode separation (mm) Figure 2. The rc.spon.sc of lhc ionization gradient chamber as a function of electrode separation. The chamber was exposed to cobalt-60 radiation (lield-size: 10 x 10 cm'; source-surface distance: 80 cm; dose rale: 86.7 cGy/min). The buildup region consisted of 3.7 mm of polystyrene. As shown in Figure 2, the response of our ioniza-lion gradienl chamber to cobalt-60 radiation varies linearly with eleclrode separalion (correlation coef-ficienl 0.99995), with dose (irradiation time) a parameter. The chamber response is represented by the measured change Q corrected for the chamber collection efficiency10 11 at the polarizing voltage of 400 V and given electrode separation z. All ionization response curves for positive and negative chamber polarities intersect at the same location on the x-axis indicating the true zero electrode separation. We purposely did not calibrate our electrode separation to this intersection point in order to emphasize that there is no need to determine the separation in an absolute manner: only the relative variation in- electrode separation is required in Eq. (5) The slopes dQ/dz obtained for the given doses in Figure 2 depend linearly on the dose as shown in Figure 3. Similar results were obtained in pulsed 50 O Q E E O O) O 2 3 4 5 6 Irradiation time (min) Figure 3. Ionization gradient dQ/dz as a function of thc irradiation lime. (dQ/dz was determined from data of Figure 2.). pholon and eleclron beams showing that (i) the chamber response is linear with dose and (ii) dQ/dz may be measured with a high degree of precision. In Tables l and 2 we show how the ionization gradient chamber meets its main objective: the absolute close determination in clinical photon and electron beams, respectively. Doses determined at a given depth in polyslyrene with a calibrated Farmer chamber and the AAPM-TG21 protocol for photon beams rnid AAPM-TG25 prolocol for eleclron beams are compared with closes determined at same depths in phanlom with our polystyrene ionization gradienl chamber. Tables I and 2 also give other relevant parameters used in the absolute dose measurements with lhe ionization gradient chamber. The discrepancies between doses determined with our uncalibrated gradient chamber and those obtained wilh lhe calibrated Farmer chamber are at most 1.08 % and 0.63 % for photon and electron beams, respectively, at all clinical energies indicaling that the ionization gradient chamber can be used as an absolute dosimeler. loniz.ation gradient dwmher in absolute /)hoton and e/ectro« dosimetry 141 Table l. Measurement of photon do.se with the ionization gradient chamber. 1 2 3 4 5 6 7 Photon Depth />*■ dQ/dz Dose (1GC) Dose (TG 21) % beam (mm) (nCm/ir') (cGy) (cGy) difference Co-60 3.7 1.113 8.274 83.41 83.18 + 0.27 4MV 10.1 1.108 9.640 97.26 96.79 + 0.49 6MV 50.1 1.103 8.313 83.81 84.23 0.49 1OMV 50.1 1.094 8.886 88.11 88.36 - 0.29 18MV 50.1 1.078 9.397 92.56 93.57 - 1.08 (1) photon beam energy; (2) depth d in phantom; (3) ratio of restricted stopping powers2 (A = 10 keV); (4) measured ionization gradient averaged over positive and negative polarities and corrected for charge recombination; (5) dose measured with ionization gradient chamber; (6) dose determined with the AAPM-TG21 protocol'; (7) percent difference between (5) and (6). Table 2. Measurement of electron dose with the ionization gradient chamber. 1 2 3 4 5 6 7 8 9 Electron Eu Depth E(d) Ï'ih/r dQ/dz.. Dose (1GC) Dose (TG 25) % beam (MeV) (111111) MMeV) (iiCiiim') (cGy) (cGy) difference 9MeV 8.1 15 5.24 1.017 10.469 95.56 96.01 0.47 l2 MeV 10.8 15 7.96 0.988 10.825 95.83 96.44 - 0.63 15 MeV 13.5 10 11.60 0.964 11.233 97.74 98.21 - 0.49 18 MeV 16. l 10 14.42 0.952 11.484 98.94 98.51 + 0.44 (1) electron beam nominal energy; (2) mean electron energy at phantom surface; (3) depth d in phantom; (4) mean electron energy at depth d; (5) ratio of restricted stopping powers2 (A = 10 keV) at E(d), (6) measured ionization gradient averaged over positive and negative polarities and corrected for charge recombination; (7) dose measured with ionization gradient chamber; (8) dose determined with AAPM-TG25 protocol'; (9) percenl difference between (7) and (8). Conclusions Uncalibrated ionization gradient chambers built as part of the phantom in which the dose is measured behave as Bragg-Gray cavities and can be used reliably in the determination of absolute dose. In contrast to the dosimetry with calibrated chambers, the dosimetry with ionization gradient chambers appears simple and requires no troublesome correction factors to account for chamber properties and for the unavailability of high energy photon and electron calibrations at standards laboratories. With our gradient chamber design, no cumbersome apparatus is required to measure the plate separation absolutely in order to determine the absorbed dose in an absolute manner. The charge per unit air mass gradient can be measured accurately (to within 1%) with relative ease in a carefully designed and precisely built gradient chamber. This implies that absolute dose measurements with ionization gradient chambers could be added to the other three currently known absolute dosimetry techniques: calori-metry, chemical (Fricke) dosimetry, and standard free air ionization chamber. References 2. Task Group 21, Radiation Therapy Committee, American Association ofPhysicists in Medicine.A prolocol for the determination of absorbed dose from high-energy photon and eleclron beams. Mled Phys 1983; 10: 1-31. 3. Task Group 25, Radiation Therapy Committee, American Association of Physicisls in Medicine. Clinical electron-beam dosimelry. Mled Phys 1991; 18: 73-109. 4. lnternational Atomic Energy Agency. Absorbed dose determination in photon and elec/mn beams. All international code of practice. lAEA Technical Report Series No. 277, 1987. 5. Bragg WH. Studies ill Radioactivity. New York: Mac-millan, 1912. 6. Gray LH. An ionization melhod for the absolute measurement of gamma-ray energy. Pmc R Soc 1936; Al56: 578. 7. Spencer LV, Attix FH. A theory of cavity ionizalion, Radiat Res 1955; 3: 239. 8. Boutillon M, Perroche A-M. Effect fa change ofstop-ping-power values oil the W value recommended by /CRUfor electrons ill dry air. Bureau lnternational des Poids el Mesures. Reporl. CCEMRl(l)/85-8, 1985. 9. Klevenhagen SC. Determination of absorbed dose in high energy electron and photon radiation by means of an uncalibrated ionization chamber. Phys Med Bio! 1991; .36: 239-53. 10. Allix FH. Determination of Aiu„ and Pim in the new AAPM radiotherapydosimetry protocol. Mled Phys 1984; 11: 714-6. 1. lnternalional Commission on Radiation Units and Mea- 11. Weinhous MS, Meli JA. Determining PK>„ the correc-suremenls. Radiation dosimetry: eleclron beams with tion faclor for recombination losses in an ionizalion energies between I and,50 Me V ICRU Report 35, 1984. chamber. Med Phys 1984; 11: 846-9.