Water-equivalent plastic scintillation detectors for high-energy beam dosimetry: II. Properties

and measurements

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

1992 Phys. Med. Biol. 37 1901

(http://iopscience.iop.org/0031-9155/37/10/007)

Download details:

IP Address: 143.111.80.27

The article was downloaded on 16/11/2011 at 16:51

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

Phys. Med. Biol., 1992, Vol. 31, No IO, 1901-1913. Printed in the LJK

Water-equivalent plastic scintillation detectors for high- energy beam dosimetry: 11. Properties and measurements

A S Beddart, T R Mackie and F H Attix Department of Medical Physics, University of Wisconsin Medical School, Madison, Wl 53706, USA

Received 21 February 1992

AbslracL ’Ihe properties of a new scintillation detector system far use in dosimetry of high-energy beams in radiotherapy have k e n measured. ?he most important properties of t h e e detectors are their high spatial resolution and their nearly water-equivalence. Measurements have shown that they have acellent reproducibility and stability, and a linear response versus dose-rate. I t Is also shown that u l q , have k i te r spat ia l resolution than ionization chambers and have much le= energy or depth dependence in electron fields due to the removal of the infiuence of the palar imion effect. Dose distributions in water, using miniature plastic =intillation detectors, have k e n measured for different high-energy photon and electron beams.

1. Introduction

A water-equivalent plastic scintillation detector system has been used for dosimetry measurements of high-energy beams and electron beams. The physical characteristics and additional theoretical considerations of these dosimeters have been discussed previously (Beddar er a1 1992a). This paper will present dosimetric properties of the scintillation detector in comparison with commonly used detectors.

Since the scintillation detector system has been discussed in detail elsewhere (Beddar er a1 1992a), only a brief description will follow. The detector consists of a plastic scintillator embedded in a small polystyrene probe and optically coupled to an optical fibre light guide. A second identical parallel fibre light guide which is not coupled to the scintillator is used for subtraction of background and stem effect (Beddar er a1 1989) as will be discussed below. The two fibre light guides are coupled to identical photomultiplier tubes (PMTS).

Results are presented for two different-sized scintillators. The first is 2.5 mm diameter by 4.0 mm in length (‘medium-sized‘) and was used without the second background fibre light guide. Therefore, this detector was only used for photon beams as discussed below. The other is 1.0 mm diameter by 4.0 mm in length (‘small-sized’) and was used for both photon and electron beams.

t Present address: Ontario Cancer InstituteiF’rincess Margaret Hospilal, Clinical Physics Department, 500 Sherbourne Street. ?bronto, Ontario M 4 X IK9, Canada.

MMI-915Ji92/1019oL+13$04.50 @ 1992 IOP Publishing Ud 1901

1902

2. General properties

2.1. Srem efecl The sources of the background signal in this system arise from the inherent dark current of the PMT and from mostly Cerenkov light generated in the fibres. The dark current produced by the PMT is approximately three orders of magnitude smaller than the scintillator output signal, under good optical coupling conditions and for typical radiotherapy beams. This was the case for both prototype detectors. The second source of hackground, which acts as a stem effect and is due primarily to Cerenkov light emission, must be considered carefully when deciding whether it can be neglected. For relative depth-dose and beam profile measurements for photon beams, the Cerenkov background is negligible for several reasons. With good optical coupling, the Cerenkov contribution is always small when compared to the light output produced by the scintillator, usually less than 3% of the output signal of the detector. In this case, the Cerenkov contribution to the total output signal of the detector (for a field size of 10 x 10 cmz and for a 10 MV x-ray beam) was equal to 1.9% and 2.670 for the medium-sized and the small-sized detectors, respectively. In addition, as shown in figure 1, the Cerenkov light output induced in the fibres varies smoothly with depth and follows the general shape of the depth4ose curves given by an ionization chamber. A maximum deviation of about 10% in rhe shape of the depthaose curves is obtained between the normalized responses of the fibres and the ionhi ion chamber sparring [ram 2 d q ~ h in w ~ i e r q u a i 20 clll a!w b r ~ y s p~aciicaiiv constant thereafter. Overall, a maximum error of 0.3% is made in the relative depth- dose curve by neglecting this 'stem effect' in a photon beam. This is valid only for photon beams and only when the stem of the scintillation detector is perpendicular to the central axis of the radiation beam because the angular distribution of electrons set in motion by the photon beam is nearly depth independent beyond d,,,. This error will increase with increasing radiation field sizes, reaching a value of as much as 1% for a 40 x 40 cmz field size.

This will not be the case when dealing with electron beams. Since Cerenkov radiation is produced when a charged particle passes through a medium of refractive index n with a velocity greater than that of light in the medium, the amount of Cerenkov radiation produced by an electron beam will be significantly larger than that produced by a photon beam. For example, for 6 and 12 MeV electron beams, the maximum CKrcnkov cnntrihution cnmpared to the light output produced by the small-sized scintillator is of the order of 12% and occurs when exposed at d,,, in a 10 x 10 cm2 fieid size as is shown helow. Previous work (Beddar er al 1992b) has characterized this effect for electron beams and has shown that the use of a parallel- fibre-bundle light pipe, identical to the one that carries light from the scintillator, offers a suitable means of generating a similar background Cerenkov light signal that can be subtracted to obtain the output from the scintillation dosimeter alone.

Therefore, for photons, the measurements presented below reflect the detector response with no dark current or Cerenkov stem effect correction since this correction would have minimal effect. However, for the electron beams, all data have been corrected for the stem effect and dark current.

2.2. Reproducibilily and stability

The small-sized scintillation detector was tested for reproducibility in a 10 x 10 all2 CO-60 gamma ray beam in a water phantom at a source-to-surface distance (SSD) Of

A S Beddar et ai

, ~. .

Plastic scintillation deiectors for dosimeiy: II 1903

01 I

0 5 10 15 20 25 30

0 5 10 15 20 25 30

Depth in Water (em)

Figure 1. Camparisan of the radiation-induced lighl inlensity in the fibre light guide io ioniulion chamber response when exposed 10 a 6 MV and 10 M V pholon beam. Curves are normalized 10 their maximum.

80 cm and at a depth of 2 cm. The reading consisted of the charge accumulated over a 1 minute period of time. A percentage standard deviation of 0.08% was obtained for a series of 20 consecutive readings. The readings relative to the average reading are contained well within the f0 .2% envelope as shown in figure 2 Similar measurements using 6 MV and 10 MV x-ray beams from a Varian Clinac 2100C yielded a percentage standard deviation equal to 0.05% and 0.09%, respectively.

The reproducibility of the detector when exposed to electrons was tested using a Varian Clinac 2100C at 6 and 18 MeV The measurements were performed in a water phantom at a depth of 1.5 cm with an SSD of 100 cm and a field size of 10 x 10 an2. Single measurements consisted of the integrated charge obtained from an irradiation of 200 monitor units (MU). Since the scintillation response of the detector for electron beams is obtained by subtracting the output signal of the ‘background’ fibre guide from the output signal of the ‘signal’ fibre guide, both outputs were separately tested for reproducibility. This is shown for the 6 MeV beam in figure 3. Percentage standard deviations equal to 0.08% and 0.06% were obtained for the signal channel output and the background channel output, respectively. For the 18 MeV beam, a percentage standard deviation of 0.09% was obtained for both outputs. Therefore, when added in quadrature, a net percentage standard deviation of the scintillator reading equal to 0.1% and 0.13% would be obtained for the 6 MeV and the 18 MeV beams, respectively.

1904 A S Beddar et al

1.010

W 2 1.005 0

1.000 9 P $ 0.985

a D: 0.980

- W

0.885

CO-80 I

t o . 2 x un. Sld. 0.r. - 0.08 x

I 5 10 15 20

Reading #

Figure 2. Reproducibility of the scintillation detector in a beam.

Sld. Dsr. - 0.08 X

0 5 10 20 F 2.48

Std. U-. - 0.08 x

5 10 15 20 6 2.83

0 Reading No.

Figure 3. Reproducibilily of (a) lhe 'signal' and (b) the 'background' fibre guide outputs of the scintillation detector system in a 6 MeV electron beam.

Daily readings over the course of 15 days were performed using the cobalt-60 unit to test the reproducibility of set-up and stability of the small-sized scintillation detector system including the high-voltage Supply and electrometers that were used

Plastic scintillation detectors for dosimetry: II 1905

in the processing of the detector response. A percentage standard deviation equal to 0.3% was obtained.

2.3. Linearify

The same detector was used in the 6 MV x-ray beam to check the linearity of its output signal from the electrometer. Measurements were performed in water at d,,, for a field size of 10 x 10 cmz at 100 cm SSD at two different dose-rates, namely 240 and 400 cGy min-I. The integrated detector output signal was found to be linear with dose from 40 cGy to 400 cGy as shown in figure 4. The slopes were equal to 3.48 nC cGy-' at both dose-rates. No dose-rate effect is seen, and shows again that the scintillator output is free of any fatigue effect for the relatively short exposure times and accumulated doses used above.

15.0 - U

r- I 12.5 0 5 10.0

B 0 7.5 e

5.0 k Y 2.5

DI

m

d " CI

cGy/min eGy/min

"." 0 40 80 120 160 200 240 280 320 360 400

Dose (cCy)

Figure 4. Signal linearity of the s'inlillalion deleclor in a 6 Mv x-my k a m at Two different dose-rates.

10.0 - U E = 12 M e V I 8.0 - 0

E

Y - 6.0 -

0 80 160 240 320 400 480 560 840 720 800

Dose (cCy)

Flgure 5. Signal lineanly of the scinliilalion detector in a 12 MeV eleclron beam. A, 'signal' fibre guide ttsponse, M, 'background' Bhre guide response, e, net scintillalor ttsponse.

1906 A S Beddar et a1

The output signal linearity of the detector at the electrometer was checked using a 12 MeV electron beam from the Varian Clinac 21ooC. The measurements were performed in water at d,,, for 100 cm SSD and a field size of 10 x 10 cmz with a dose-rate at the detector of 400 cGy min-I. The response of the ‘signal’ fibre guide channel and the ‘background’ fibre guide channel, respectively, as a function of absorbed dose is shown in figure 5. The output signal of both channels was found to be linear with dose within the dose range indicated. The slopes were equal to 1.223 nC cGy-’ and 0.132 nC cGy-’ for the ‘signal’ fibre guide and the ‘background’ fibre guide, respectively. Notice that, as expected, the dark current of the photomultiplier tube (PMT), which was not subtracted from the output signal of the ‘background’ channel, does not show up in the ordinate of the curve shown in figure 5. This is due to the fact that the Cerenkov light output is about two orders of magnitude higher than the dark current. The linearity of the net response of the scintillator as a function of dose in the Same range (40 cGy-800 cGy) is also shown in figure 5. The slope was found to be equal to 1.091 nC cGy-I.

2.4. Dose-rate proportionalitj

The dose-rate proportionality of the detector was tested further using the Varian Clinac 2100C at dose-rate settings of 80, 160, 240, 320 and 400 MU min-’. The mon- itor chamber is calibrated for all dose-rate settings for both photon energies (6 MV and 10 MV) according to the AAPM a s k Group 21 protocol (AAPM 1983) using a Farmer ionization chamber. The respoilse ot the detectoi dctermmcd by mtcgiating charge corresponding to an irradiation of 200 cGy is shown in figure 6. This response was normalized at 240 cGy min-’ and is equal to unity within f0.05% showing that the detector response k independent of dose-rate in the range investigated.

1.002

.-A BMVX-Rays

1 . . . . . , . . . . . . . . 1 % ffi 1,000 ..... ... .. . ~ . . . . . . -../.. \_ )r.

0.990 I 0 100 ZOO 300 400

Dose Rate (cGy/min)

Figure 6. Do=-mle independence of Ihe scintillarim detector lor x-rays

The dose-rate proportionality for electron beams was tested using the Varian Clinac 2100C at the same dose-rate settings. The same calibration procedure was fol- lowed as above. The response of the scintillation detector consisted of the integrated charge corresponding to a 1 minute irradiation time for the different dose-rates. The output response of both channels (‘signal’ fibre guide and ‘background‘ fibre guide) and the scintillator as a function of dose-rate (cGy min-I) are shown in figure 7(a )

Plastic scintillation detectors for dosinferry: II 1907

and (b) for the 6 MeV and the 12 MeV electron beam energies, respectively. The response of the scintillation detector is linear with dose-rate in the range indicated above. Also, the ‘signal’ fibre guide and ‘background’ fibre guide outputs are found linear as a function of dose-rate. For the 6 MeV electron beam the slopes were equal to 1.236 nC cGy-’ and 0.135 nC cGy-’ for the ‘signal’ fibre guide and the ‘background’ fibre guide, respectivcly. Similarly, for the 12 MeV beam the slopes were equal to 1.225 nC cGy-’ and 0.129 nC cGy-’ for the ‘signal’ fibre guide and the ‘background’ fibre guide, respectively. The net response of the scintillator yielded nearly equal slopes; namely 1.101 nC cGy-’ and 1.096 nC cCy-’ for the 6 MeV and the 12 MeV beam, respectively. The dose response is nearly the Same as that performed with either the 6 MeV or 12 MeV electron beams. The linear regression correlation coefficients for all of these responses were found to be equal to 1.ooO.

Dose Rate (cCy/min)

Figure 7. Dose-rate proponionalily of the scintillalion detector for (a) 6 MeV electrons and (b) 12 MeV elCCtrons A, ‘signal’ fibre guide response, H, ‘background’ fibre guide response, 0 , ne1 scintillalor response.

2.5. Spaiial resolution

The medium-sized scintillator detector was used in the measurement of a half-blocked 6oCo beam profile and was compared to those obtained by a 0.6 cm3 Farmer cham- ber (Victoreen, Model #30-352) and a 0.1 cm3 P”W chamber (Victoreen, Model

1908 A S Beddar et a1

#30-350-1). Both ionization chambers and the scintillator detector were oriented perpendicularly with respect to the radiation beam so that their longitudinal axes were orthogonal to the beam central axis and to the scanning direction. The mea- surements were performed in water for a 1 0 x 10 cm2 field size at 80 cm SSD. The results are shown in figure 8. In addition, a dose-calibrated film was used to see how its resolution compares to the plastic scintillator. Although it is known that film has an undesirable energy dependence in high gradient regions, it is still commonly used in mapping beam profiles. The spatial resolution of film is limited by the aperture of the densitometer used to read its optical density. The scintillation detector ex- hibits the highest resolution when compared to both ionization chambers as shown in figure 8. This was expected since the scintillation detector has a smaller detecting volume than the 0.6 cm3 and t h e 0.1 cm3 ionization chambers.

1.0

0.8

CI 5 0.4

B m -

0.2

n n I"

-10 -5 0 5 10

Transverse Distance (mm)

Figure 8. A '"CO half-blocked gamma-ray k a m profile in water using llie scintillation deteclor mmpared to two dineren1 size ionization chambers and dose-calibrated film.

The small-sized scintillator detector (1.0 mm in diameter and 4.0 mm in length for a volume of 0.003 an3) was used in the measurement of transverse dose profiles of a 6 MV x-ray beam in water a t 5.0 em depth for a 10 x 10, 6 x 5 and 3 x 3 cm2 field size at I00 cm SSD. These profiles are shown in figure 9 and are compared to those obtained using a 0.1 cm3 PTW ionization chamber and a Scanditroniw p- type silicon diode. (While the dimensions of the diode are not well defined by the manufacturer, its sensitive volume is specified to be about 0.2 to 0.3 mm3, with an 'effective detection area' equal to 2.5 mm in diameter and a 0.45 mm thick p s i silicon substrate.) The scintillation detector had a higher spatial resolution when compared to the 0.1 an3 ionization chamber. However, the resolution obtained with the diode and the plastic scintillator were found to be comparable. Other measurements performed with the same detectors as above using electron energy beams (6, 12 and 18 MeV) are presented elsewhere (Beddar 1990). These measurements also showed no significant differences between the diode and the scintillation detector for a 10 x 10 cm' field size.

3. Depth-dnse distributions in water

For x-ray beams, depth4ose measurements in water were performed using the

Plnslic scinfillniion deiectors for dosimety: II

- W

U 10 20 30 40 50 60 70

Transverse Distance (mm)

1909

Figure 9. Transvene dose profiles of a 6 MV x-ray beam for a 10 x 10. 5 x 5 , and 3 x 3 cm2 field sizes using the Yinlillalion dclcclor, an ionimlion cliamber and a photon diode.

medium- and small-sized scintillators. For electron beams, only the small-sized scin- tillator detector was used. All percentage depthdose curves were normalized at dn,,,.

3.1. Pholon beanis

The medium-sized scintillator was used in the measurement of depth4ose profile for a soCO beam. The radiation field size was equal to 10 x 10 an2 for an SSD of 80 cm. The output signal of the detector consisted of the integrated charge corresponding to a 100 MU irradiation. The normalized response of the scintillation detector versus depth in water is shown in figure lO(a). The standard deviation of the measurements at each depth was within the symbol size and were all less than 0.2%. This response is seen to overlap the depth4ose curve obtained with a Farmer ionization chamber (Victoreen, Model #30-352). Similar measurements were performed using an AECL Therac 20 which provides 6 MV and 18 MV photon beams. The radiation field size was 10 x 10 cm2 at an SSD of 100 cm. The normalized response of the scintillation detector closely follows the percentage depth4ose obtained when using a Wrmer ionization chamber. This is shown in figure 1O(b) for the 18 MV photon x-ray beam.

Similar measurements, using the small-sized scintillation detector, were performed on a Varian Clinac 4 providing a 4 MV photon beam and a Varian Clinac 2100C providing 6 MV and 10 MV photon beams. These measurcments were performed in a water phantom for a 10 x 10 an2 field size at an SSD of 80 cm for the 4 MV beam and 100 cm for the 6 MV and 10 MV beams. Excellent agreement was found between the depthaose measurements obtained with the ionization chamber to those obtained with the small-sized scintillation detector. The results obtained for the 10 MV x-ray beam are shown in figure lO(c).

3.2. Electron bennu

Depthdose measurements in water using the small-sized scintillator detector were performed on a Varian Clinac 2100C providing a 6, 12 and 18 MeV electron beam. These measurements were compared to the depthdose curves obtained using the

i

I

i !

1910 A S Beddar e1 a1

100 '... CO-60

- Ion Chamber . Scintillatar

BO

40

" 0 2 4 8 8 10 12 14 18 I

01 0 5 10 15 20 25 30

0 5 10 15 20 25

Depth in Water (em)

Figure 10. Percentage dcpth4lose in water for (0 ) ''CO gamma mys, (b) I8 Mv x-lays using the medium-sized scintillation delector and (c) 10 MV x-ray using the small-size scintillation deleclor in mmparison with an ionizalion chamber.

Farmer ionization chamber and a Scanditronk p-type Si diode. When acquiring data for the electron beam dcpthdose curves, the scintillation detector signal was corrected for stem-effect background as discussed above. No corrections were applied to the diode. The ionization measured using the ionization chamber was converted to dose following the Bsk Group 21 protocol (AAPM 1983). This conversion was done in a three step process: (i) the measurement depths were corrected for the chamber- shift correction, (ii) then the absolute reading (ionization) of the ionization chamber was corrected for the water/air stopping power ratio at all depths, (iii) finally, the electron h e n c e correction was applied at all depths for the corresponding nominal

Plaslic scinlillarion deteclors for dosintelry: 11 1911

0 5 10 15 20 25 30 35 40 45 50 55 80

E P 12 MeV

0 0 10 20 30 40 50 80 70 80 90 100

100 e’ E = 18 YeV

40

20 ~~~ 0 0 20 40 80 80 100 120 140 180

Depth in Water (mm)

Figure 11. Percentage depih-dose in waler for (0 ) 6 MeV electrons, (b) I2 MeV electrons and (c) 18 MeV electmns in comparison wilh an ionization chamber and an elcclron diode.

electron energy of the electron beam being evaluated (6, 12, and 18 MeV beam). The d e p t h d o s e C U N ~ S for the three different electron energies using the scintillation detector and compared to the Farmer ionization chamber and the diode are shown in figure 11. The d e p t h d o s e curves obtained with the scintillation detector arc found to be in excellent agreement with those obtained with the ionization chamber for the three different energies at all depths. The measurements performed with the diode showed some disagreement with the other detectors in the build-up region, especially at lower energies and in the tail region a t 6 MeV: This behaviour is not unexpected.

1912

4. Discussion

The properties of this detector as shown in this and previous work (Beddar ec a1 19YZa) demonstrate the potential usefulness of these new detectors in radiation dosimetry. They may be used for field mapping in water phantoms, or for in vivo insertions. They offer the method of choice for electron inhomogeneity and inter- face measurements as well as for dosimetly of stereotactic radiosurgery (Beddar et al 19%~).

At the present time, plastic scintillation detectors as described here will most likely not lend themselves to any kind of traceable calibration (National Institute of Standards and Rchnology or Accredited Dosimetly Calibration Laboratories), since the PMT is strongly dependent on the stability of the high-voltage supply and has no plateau where one can operate. However, if the PMT is replaced by some other photosensitive device (i.e. solid state devices or photodiodes) that converts the light into an electrical signal with a constant gain factor, then these plastic scintillator detectors wuld lend themselves to traceable calibration. (This work is presently under investigation.) If this were the case, plastic scintillator detectors would have additional advantages over the presently used dosimeters in reducing the number of correction factors that need to he calculated or evaluated when calibrating high- energy photon and electron beams. This may simplify the calibration procedures as. recommended hy existing protocols.

A S Beddar et a1

5. Cnnclusinns

This work has shown that these new scintillation detectors have excellent reproducihil- ity and stability. In addition, they are linear with dose and are dose-rate independent. They have excellent spatial resolution and when stem-effect background subtraction is performed have shown themselves to he ideal devices for characterizing either a photon or electron beam.

Acknowledgments

The authors would like to thank Charles Lescrenier and Radiation Measurements Inc. for their support.

Rhsume

D6tecleun B sinlillatians en plastique &quivalenI eau pour la dosimilrie des faisceaw: de haute energie: 11. Proprii16s et mcsurcs.

Les autcun on1 diterminC la amctirisliques d'un noweau syslkme de detectcun & sfintillations utilise pour la dosim6lrie des iaisceaux de hmle encrgie en radiolhirapie. Les propri&& lcs plus inl6ressanles de ces ditecieurs mnt leur grande dsolulion spatiale et leur composition proclie de cellc de I'eau. Des mewres on1 monld que ces d6iccteurs pr6sentent dcxcellenles repmductibiliti et stabilit6, el une dponse liniaire en fonction du dibit de dose. Lc5 auteun on1 aussi m o m 6 que oes dClecteun pdsentent unc meilleure &.solution spatiale que les chambres dionisalion el wont kauwup mains dependanls de I'energie ou de la profonder dam dcs champs d'electronr, i cause de la suppression de ?influence de Yeffel de plarisalion. Les distributions de dose dans I'eiiu, ulilisanl des dCtecteun miniature B sinlillations, en plastique, on1 616 mesurks pour diff6rents iaisceaux de pholons et d'ilectrons de baulc incrgic.

Plastic scinlillation detectors for dosimeiry: I1 1913

Zusammenfassu ng

Msser-iquivalenle Plaslikrzinlillalionsdetektoren fur die Dosimeirie hochenergelischer Strahlen: ' E l 11. Eigenschaflen und Mesungen.

Die Eigenschaften eines neuen Szintillalionsdetcklonyslems fiir die Dosimelrie hochenergelischer Slrahlen in der Strahlentherapie wurden kslimml. Die wichiigsle Eigenschaft dieser Delektoren kt ihre hohe idumliche AuABsung und die Paache, daB sie nahau Wdsserdquivaleni sind. Mesungen haben gezeigt, daQ sic eins ausgezeichnete Repmduzierharkeit und Slabilil it besitzen, sowie ein lineares Verhalien gegeniikr der Dosisleislung. Sie haben cine &re idumliche AuRBsung aB lonisalion- skammern und ksi tzen eine vie1 gelingere Energie- und l iefenahhingigkeit in Eleklronenfeldern weil der EinAuO von PolaCsalions-effeklen weglilli. Dosisveaeilungen in Waser wurden mil Hilfe cines Miniatur-Plastiliszintillationsdeleklon fiir verschiedene hochenergetische Photonen- und Elektro- nenstrahlen gemessen.

References

AAPM I983 A protocol lor the determination of absorkd dose Irom high-energy pholon and electron kams Med P@s. 10 741-71

Beddar A S 1990 Mter-equivalenl plastic winli l lal ion detecton for high energy photon and eleclron k a m PhD lhesis Univenity of Wisonsin (Madison, WI: Medical Physics Publishing Corporalion)

Beddar A S, AttU F H and Mackie T R 1989 On the nature of the lighl induced by radiation in ransparen1 media used in radiotherapy Med. Phys. 16 683

Beddar A S , Mackie T R and All& F H 1992a Water-equivalent plastic sh t i l l a l ion detecton lor high energy beam dosimelxy: I. Physical characteristics and t h e o r e l i d mnsiderations Phys. Med BioL 37

Beddar A S . Mackie T R and Allix F I4 19921, Cerenkov l ighl generated in optical fibres and olher light pipes irradiated by electron \rams Php. Mcd B i d 37 925-35

Beddar A S, Mason D L and OBr ien P F 1992c Dosimetry of slereolaclic radiation fields using a miniature plastic scintillation delcctor Md. Phys 19 190