Home

Search

Collections

Journals

About

Contact us

My IOPscience

Electron depth absorbed dose distribution for a 10 MeV clinical microtron

This content has been downloaded from IOPscience. Please scroll down to see the full text. 1975 Phys. Med. Biol. 20 39 (http://iopscience.iop.org/0031-9155/20/1/003) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 192.236.36.29 This content was downloaded on 01/10/2015 at 09:32

Please note that terms and conditions apply.

PHYS. MED. BIOL., 1975, VOL.

20, NO. 1, 39-46. @ 1975

Electron Depth Absorbed Dose Distribution for a l 0 MeV Clinical Microtron A.BRAHME,

M.SC.

Institute of Radiation Physics, 104 01 Stockholm 60, Sweden

G. HULTGN,

M.SC.,

and H. SVENSSON,

PH.D.

Radiation Physics Department, University of Ume&, 901 85 Umeb, Sweden Received 17 June 1974, in final form 28 August 1974 ABSTRACT. Central axis depth absorbed dose distributions of the electron beam from a medical microtronacceleratorhave been measured. Themeasured distributions differ from those of existing betatrons and linear accelerators. A larger absorbed dose build-up and a sharper dose fall-off are obtained in close agreement with theoretical calculations for monoenergetic electron beams. The differences from other accelerators are explained by the narrowenergy spectrum, the clean geometry and the small amount of scattering material in the electron beam of the microtron.

1. Introduction Electrondepthabsorbed dose distributionsfromtherapy accelerators (NACP 1972, HPA 1972) usually differ significantly from theoretical distributions calculated for monoenergetic parallel broad beams (Berger and Seltzer 1969). These differences are due to the broad spectral and angular distributions of the electrons in the clinical beams. The agreement between distributions measuredon point monodirectional and monoenergeticbeamswithalarge ionization chamber (Harder and Schulz 1971) and the theoretically calculated distributions are, however, rather good (Caswell and Berger 1972). The shapes of the theoretical curves are often more suitable for radiation treatment t'han those obtained from clinical accelerators because the theoretical curves have a more pronounced dose build-up, a higher dose level maintained to a greater depth foragivenelectronenergy,asharper dose fall-off and less X-ray contamination. Thebroadspectralandangulardistributions of the electronsin therapeutic beams are partly due to intrinsic accelerator factors (acceleration and extraction principles) and partly due to scattering and energy degradation in materials between the accelerator and the patient (e.g. vacuum window, air, foils, transmissionchambers and collimators).Simple modifications of the latter factors will improve, to some degree, the dose distribution from acceleratorsalreadyinhospital use (Loevinger, Karzmarkand Weissbluth 1961, Svensson 1971). Recently published depth dose curves from a linear accelerator with a scanning beam system without scatteringfoils indicate that the electron beam from this accelerator is fairly clean (Almond and de Almeida 1973). I n this paper, depthdose distributions are presentedfor a new t'ype of therapeutic

A . Brahme et al.

40

accelerator, the microtron.Themeasureddept'habsorbed dose distributions agree more closely with the theoretical distributions than those published for medical betatrons or linear accelerators. 2. Experimental

2.1. The accelerator Thetechnicalperformance of the accelerator,a fixed energymicrotron consisting of a stationary accelerator feeding a beam transport system placed in an isocentric gantry, has been describedearlier by Brahme and Reistad (1972).Thepresentmeasurements were carried outat a fixed accelerator energy, E,, of 10.2 0.05 MeV. Thefull width of the energydistribution is 100 keV (35 keV a t half maximum) as measured by a magnetic spectrometer. The additional energy straggling of the electrons from the vacuum window of the accelerator to the patient is kept low by the small amount of scattering material in the bea'm (table 1). Table1.Electronbeamparametersfor the microtronaccelerator, MM 10. The energy spread, t'he most probable energy loss and the RMS angles of divergence are calculated according to Brahme (1972).

Source

Full width at half maximum of energy distribution (keV)

Xost probable energy loss (keV)

Root mean square angle of divergence (radian) ______---

Crude beam Vacuum window, transmission chambers, mirror Scatt,ering foil Air

35 25

-

95

330 180

0.47

40

100 195

260 590

0.15

80

0.02 0.1 1

0.10

-

Total without foil Total with foil

0.49

The crude electron beam has its focus just outside the vacuum window with a focal spot diameter of 1.3 mm. It is rotationally symmetrical with a diameter a t half maximum of 14 cm at theisocentre distance of 100 cm. A scatteringfoil (0.2 mm Au) has been used to increase the energy spread in the beam during some of the measurements. The foil is placed at thefocus of the crude electron beam.Themain part of the energy degradation, AE,, and the energy and angular spreadof the electron beam are caused by the scatteringfoil (table 1). 2.2. Depth absorbed dose measurements

Themeasurements were madeinlargephantoms of polystyrene(density 1.057 g cm-2) and water. The walls of the water phantom were made of 0.5 cm Lucite. One of the walls was equipped with a thin window in order to make depth dose measurements possible a t very small phantom depths. The absorbed

Depth Microtron

Dose Curves

41

dose measurements were madewithliquidionizationchambers(Wickman 1973, 1974). The design of the liquid chamber is similar to that of a parallel plate ionization chamber in which the air is replaced by a 0.3 mm thick layer of a dielectric liquid. The chamber is made of polystyrene (Rexolite). It has a front wall thickness of 0 . 3 mm and theconductive electrodes consist of very thin layers of beryllium. The response of the liquid ionization chamber has been compared with that of the ferrous sulphate dosemeter a t different phantom depths and electronenergies between 10-30 MeV (Svensson, Hulten, Hettinger andWickman 1973). Therelative responses agreefor the twodosemeters, within experimental accuracy which is approximately 0.5% near dose maximum and 1-2% at thesteep slope of the depthdose curves. The liquid ionization chamber has therefore been regarded as energy independent for the absorbed dose measurementspresented.Theuncertaintyin the relativeabsorbed measured dose distributions is less than 2%. 3. Results 3.1. Comparisons with other accelerators

Fig. 1 shows the first part of depthabsorbed dose curvesfromdifferent accelerators. 811 the curves were measured with the technique described above except for curve ?io. 3 which wasdeterminedby Almond andde Almeida (1973) using an ionizationchamber method. Onlyaminorportion of the

(E,),,=9~4-10~0 MeV SSD = 100 cm Field sizes 2 IO cm x IO cm

l 'OO

l

I

~

1

j 2

3 Depth in water (cm) Fig. 1. Depthabsorbed dose curves for the build-up region. Experimentalpointsare given for curves Nos 1, 2, 4, 5 and 6. Curve No. 3 is taken from Almond and de Almeida ( 1 9 7 3 ) . Curve No. 1 : Microtron MM 10, no foil. Curve No. 2 : Microtron MM 10, with foil. Curve No. 3: Linear accelerator, Sagittaire, scanning beam. Curve h-0. 4 : Betatron, BBC, Uppsala. Curve No. 5 : Linear accelerator, MEL 75-10, Uppsala. Curve No. 6 : Microtron MM 10, with filter.

42

A . Brahme et al.

differences in the curve shapes should be due to the differences in accelerator electron energy, field sizes, SSD and dosimetry. Theabsorbed dose near the surface is higher and the dose maximum is reached a t a smaller depth for the conventional accelerators, Nos 4 and 5 , than for the scanning beam linear accelerator,No. 3, and the microtrons,Nos 1 and 2. This is explained by the fact that the former accelerators havetreatment beams withbroaderspectralandangularelectrondistributions. For the betatron, No. 4,the intrinsic narrow energy spread of k 50 keV (Kollath 1967) is often increased considerably in passing through the wall of the doughnut and the scattering foils. The width of the energy distribution for the scanning beam linear accelerator is limited to k 3% by a slit system (Aucouturier, Huber and Jaoven 1970). I n more recent specifications thisvalue is decreased to 2% or 200 keV a t 1 0 MeV which is approximately the same as for the microtron beam when the 0.2 mm Au foil is used, due to thestraggling in the foil (table 1 ) . The angular spreadof the electrons at theisocentre is also approximately the same for these two accelerators as the elastic scattering in the foil does not contribute significantly to the angular spread at the isocentre when the foil is situated at the focal spot of the crude electron beam. The main portion of the angular spread for these two accelerators thus results from elastic scattering in the air (table 1). Curves Kos 2 and 3 should therefore be expected to have similar shapes. Smalldifferences are, however, observed in the build-up and the fall-off parts of the curves. Almond and de Almeida (1973) have pointed out that the electron beams from their accelerator are contaminated with electrons scattered from the collimator which, together with a somewhat larger energy spread, could explain the differences. The crude microtron beam, curve KO.1, gives, as might be expected, a still more pronounced absorbed dose build-up. Fig. 2 shows depth absorbed dose curves measured in water. The same beam geomet,ries were used as in fig. 1. The dose fall-off is somewhat sharper for the microtron than for the other accelerators. 3.2. Comparisons with theory

Figs 3 and 4 compare experimental depth dose distributions with calculated distributions (Berger and Seltzer 1969, Berger 1974, private communication). The depth is given as zIR, where z is the depth in the phantom and R, the practical extrapolated range. Thecalculated curves originally given for infinite SSD are recalculated to SSD 100 cm by application of the inverse square law correction. The root mean square error of the calculated distributions amounts to about 3%. The figures clearly indicate that themicrotron with the foil, curve No. 1, differs more from the calculated curve than does the microtron without the foil, curve No. 2. Curves Nos 1 and 2 have been used for extrapolation to a monoenergetic beam, N o . 3 . The extrapolation has been made with respect to the full width a t half maximum of the energy distribution at the phantom surface. This measureof the energy spread is, to a first approximation, proportional t'o the relative change in the fall-off slope of the depth dose distribution dose fall-off, a (Brahme 1972). Theext'rapolation thus resultsinsharper

Microtron Depth Dose Curves

43

Large fields Microtron MM IO (foil) --v-Betatron BBCJppsala -cLinac,MEL75-IO,Uppsala

0

2

4

6

Depth in water (cm)

Fig. 2 . Depth absorbed dose curves for different accelerat'ors. The first part of the curves are given in detail in fig. 1.

m

Mkmtrm a2rnmAu foil

3

-

2""

p

0

02

mnoenergetic ele Calculated curve

0 4 08

06

Fig. 3. Depth absorbed dose curves for water. Curves Nos 1 and 2 represent measureNo. 3 was extrapolated from thesecurves. ments made on the microtron and curve The calculated curve was taken from Berger and Seltzer (1969). Experimental points are also taken from Almond and de Almeida (1973) who measured on a scanning beam linear accelerator. From curve No. 2 , R, was determined to 4.94 f 0.02 cm a t E, = 9.94 f 0.07 MeV. The t'heoretical curve for 10 MeV has R, = 4.95 f 0.02 cm.

A . Brahme et al.

44

displacement of the dose maximum to a somewhat great'er depth and consequently a lower dose level at thesurface (curve ?To. 3 in figs 3 and 4).The small differences in most probable electron energy have a negligible influence on the curve shapes infigs 3 and 4 as they are plotted against zlR,.

ZIR,

Fig. 4. Depthabsorbed dose curvesforpolystyrene. Thesameirradiation geometries were used with the microtron as in fig. 3. The calculated curve is taken from Berger (1974, private communication). From curve No. 2 R, was determined t o be 5.24 & 0.02 g cm-2 at E , = 9.94 0.07 MeV. The theoret'ical curve for 10 MeV has R, = 5.18 0.02 g cm-z.

For polystyrene as phantom material (fig. 3) as well as for water (fig. 4) the extrapolated curves differ somewhat fromthe calculated ones for small phantom depths but agree well for depths beyond dose maximum. Part of the difference is explained by the low energy secondary electrons produced in the air and the angular spread of the electrons at thephant'om surface for which no correction ismadeinthe aboveextrapolation. It is however notable thatboththe experimental results of Harder and Schulz (1971) a t 20 MeV and the present measurements a t 10 MeV have dose maxima a t rather smaller depths than those predicted by the Monte Carlo results of Berger and Seltzer (1969). A comparison of the measuredrelativedistributionsinwater and polystyrene show good agreement over the entireelectronrangeexcept for the somewhat sharper dose fall-off in polystyrene. This difference is explained by the lower mass scattering power in polystyrene compared to water. However, the differences in most cases are small enough to justify the use of polystyrene as a water-equivalent phantom material. 4.

Discussion

Opinions differ as to whatis the most suitable shapeof a depthabsorbed dose distribution. A large absorbed dose build-up may in manycases be regarded as favourable. However, a large build-up region may be undesirable in cases in

Alicrotron Depth Dose Curves

45

which thetumourextendsto small depths.Themicrotron acceleratoris equipped with aspecial depth dose flattening filter that can be used if a constant dose level from a small depth to the dose maximum is desired (Brahme 1974). This filter does not change the depth absorbed dose curve beyond the dose maximumnor does it deterioratethe field uniformity. It does,however, increase the absorbed dose a t small phantom depths (fig. 1, curve No. 6). It is thus possible to change the shape of the dose distribution for the treatment of a particular patient. The difference in dose fall-off for the various accelerators is in some cases of little significance forradiationtreatment at an electronenergy of 10 MeV (fig. 2). The difference between the curve shapes is more important a t higher energies. To obtain the 90% dose level a t 7 cm depth, approximately 19 MeV monoenergetic electrons are thus needed (Berger and Seltzer 1969). The most probable electron energy at the phantom surface necessary for the scanning beam linea,r accelerator is 20 NeV and for some betatrons 28-35 MeV (KACP 1972). These figures are valid for SSD = 100 cm and large field sizes. It appears that t’he microtron with its low intrinsic energy spread and clean beam geometry a t thesehigh energies should give veryfavourabledepth dose curves. The intrinsicenergyspread of amicrotronbeam at higher energies is expected theoretically to be even smaller than for a 10 NeV machine. A 20 MeV medical microtron is now being constructed for the University Hospital a t Umeb. 5.

Conclusions

Central axis depth absorbed dose distributions of the electron beam from a medical microtron accelerator have been measured. The measured distributions differ from thoseof existing betatrons andlinear accelerators. A larger absorbed dose build-up and a sharper dose fall-off are obtained, in close agreement Tvith theoreticalcalculationsformonoenergeticelectronbeams,Thedifferences fromotheraccelerators are explained by the narrowenergy spectrum,the clean geometry and the small amount of scattering material in the electron beam of the microtron. We wish to thank Professor Gunnar Hettinger, Mr. Dag Reistad, Professor Rune Walstam and Professor Olle Wernholm for valuable discussions. This study was supportedbygrants from the Swedish Board for Technical Development.

RBSUNE La distribution de la dose Blectronique profonde absorbbe pour un Microtron clinique de 10 MeV On a mesure les distributions de ladose absorbee en profondeurle long de l’axe central pour un faisceau d’Qlectrons provenant d’un accelerateur medical microtronique. Les distributions mesurbes different de celles pour les betatrons et accelerateurs linthires existants. On obtient une plus grande accumulation de la dose absorbbe ainsi qu’une chute plus rapide de la dose, en bon accord avec lea calculs theoriques pour des faisceaux monoenergetiques d’electrons. Les differences trouvees en comparaison avec d’autres accelerateurs s’expliquent par le spectre d’energie etroit, la geometric nette et la petite quantite de la matiere diffusante dans les faisceaux 6lectronique du microtron.

46

Nicrotron Depth Dose Curves ZUSAMMENFASSUNG

Die Verteilung der absorbierten Elektronentiefendosis fur ein klinisohes 10-MeV-Mikrotron Es sind die Verteilungen der absorbierten Tiefendosis entlang der Zentralachse fur den Elektronenbundel aus einem medizinischen Mikrotron-Beschleuniger gemessen worden. Die gemessenen Verteilungen unterscheiden sich von denen fur die vorhandenen Betatronen und Linearbeschleuniger. Man erhalt einen grosseren Zuwachs der absorbierten Dosis, sowie einen scharferen Abfall der Dosis, in guter Ubereinstimmung mit dentheoretischen Berechnungen fur monoenergetische Elektronenbiindel. Die Unterschiedeanderen Beschleunigern gegeniiber werden erklart durch des schmale Energiespektrum, die saubere Geometrie und die kleine Menge des Streumaterials in dem Elektronenbundel des Mikrotrons.

REFERENCES P. R., and DE ALMEIDA, C. E., 1973, in Radiation Therapy withthe Electron Beam, ClinicalApplications of 7-30 *VeVElectronBeam, E d . hT.duV.Tapley, P. R. Almond and C. E. de Almeida (Chicago: Radiological Society of North America) p. 1. AUCOUTURIER, J., HUBER, H.,a n d JAOVEN, J., 1970, Rev. Tech. Thomson-CSF, 2, 655. BERGER, M. J., and SELTZER, S. M., 1969, Ann. N.Y. Acad. Sci., 161, 8. BRAHME, A., 1972, On the Optimal Choice of Scattering Foilsfor Electron Therapy, TRITAEPP-72-17 (Stockholm: Royal Institute of Technology). BRAHME, A., 1974, Swedish pat. &pp.7404930-5. BRAHME, A., a n d REISTAD,D., 1972, The Microtron, a N e w Accelerator f o r Radiation Therapy. Digest of t h e 3rd International Conference on Medical Physics, Gothenburg, 1972. (Abstract, Phys. Med. Biol., 17, 692.) CASWELL,R. S., and BERGER,M. J., 1972, Theoretical Aspects of Radiation Dosimetry. Paperpresented at Conf. onParticleAcceleratorsinRadiationTherapy,Los Alamos Scientific Laboratory, 2-4 October, 1972. HARDER, D., a n d SCHULZ, H. J., 1971, International Congress Series KO.249 (Amsterdam: Excerpta Medica) p. 475. HPA (HOSPITAL PHYSICISTS ASSOCIATION), 1972, Central A x i s Depth Dose Data for Use in Radiotherapy. Br. J . Radiol., Suppl. 11, p. 83. KOLLATH, R.,1967, Particle Accelerators (London: Pitman). LANZL, L. H., 1969, Ann. N.Y. Acad. Sci., 161, 112. LOEVINGER, R., KARZMARK, C. J.,a n d WEISSBLUTH, M,, 1961, Radiology, 77, 906. NACP (XORDIC ASSOCIATION OF CLIKICAL PHYSICS), 1972, Acta Radiol. Ther. Phys. Biol., 11, 603. SVENSSON, H., 1971, Acta Radiol. Ther. Phys. Biol., 10, 443. SVESSSON,H., HULTEN, G., HETTINGER, G., and WICKMAN,G., 1973, International Congress Series No. 301 (Amsterdam: Excerpta Medica) p. 439. WICKMAN, G., 1973, International Congress Series No. 301 (Amsterdam: Excerpta Medica) p. 440. \\’ICKMAN, G., 1974, Phys. Med. Biol., 19, 66. ALMOND,