8/12/2019 Cirro Ne

1/4

1

The GEANT4 toolkit capability in the hadron therapy field: simulationof a transport beam line

G.A.P. Cirrone a G. Cuttonea, F. Di Rosaa, L. Raffaelea, G. Russoa, S. Guatellib

and M.G. Piab

aLaboratori Nazionali del Sud, Istituto Nazionale di Fisica Nucleare, Via S. Sofia 62, Catania, Italy

bIstituto Nazionale di Fisica Nucleare, Sezione di Genova, Via Dodecaneso 33, Genova, Italy

At Laboratori Nazionali del Sud of the Instituto Nazionale di Fisica Nucleare of Catania (Sicily, Italy), the firstItalian hadron therapy facility named CATANA (Centro di AdroTerapia ed Applicazioni Nucleari Avanzate) hasbeen realized. Inside CATANA 62 MeV proton beams, accelerated by a superconducting cyclotron, are used for

the radiotherapeutic treatments of some types of ocular tumours. Therapy with hadron beams still representsa pioneer technique, and only a few centers worldwide can provide this advanced specialized cancer treatment.On the basis of the experience so far gained, and considering the future hadron-therapy facilities to be developed(Rinecker, Munich Germany, Heidelberg/GSI, Darmstadt, Germany, PSI Villigen, Switzerland, CNAO, Pavia,Italy, Centro di Adroterapia, Catania, Italy) we decided to develop a Monte Carlo application based on theGEANT4 toolkit, for the design, the realization and the optimization of a proton-therapy beam line. Anotherfeature of our project is to provide a general tool able to study the interactions of hadrons with the human tissueand to test the analytical-based treatment planning systems actually used in the routine practice. All the typicalelements of a hadron-therapy line, such as diffusers, range shifters, collimators and detectors were modelled. Inparticular, we simulated the Markus type ionization chamber and a Gaf Chromic film as dosimeters to reconstructthe depth (Bragg peak and Spread Out Bragg Peak) and lateral dose distributions, respectively. We validatedour simulated detectors comparing the results with the experimental data available in our facility..

1. Introduction

An increasing number of radiotherapy facilitiesaround the world use charged particles beams,namely protons or carbon ions to treat tumours[1]. Yet nowadays, radiotherapy with chargedions represents an innovative and pioneering tech-nique in the area of the battle against cancer.Ions have physical and radiobiological character-istics that differ from those of conventional ra-diotherapy beams (high energy photons and elec-trons) and offer a number of advantages. They

have a depth dose curve completely different fromthose of conventional beams, with a finite rangeof penetration in a tissue, negligible scatteringand a high amount of energy released at the endof their track, giving rise to the so-called BraggPeak. The depth and width of the Bragg peakdepend respectively on the initial beam energyand energy spread of the beam. Therefore vary-

ing the energy in a controlled way it is possi-ble to superimpose many narrow Bragg peaks toachieve a large Spread Out Bragg Peak (SOBP)that can tailor a highly conformal dose distri-bution, able to irradiate selectively tumour vol-umes sparing the adjacent critical normal tissues.Charged ions combine the advantage of a highphysical selectivity with that of high-LET radi-ations providing a higher Relative Biological Ef-fectiveness (RBE). For protons RBE is assumedto be 1.1 in clinical practice, although there is anevidence for an increased RBE that varies with

depth [2]. All these considerations led, in the lasttwenty years, a large number of medical physi-cists, radiotherapists and oncologists, to considerwith more and more interest ions with respect toelectron and photon beams. Today more than 25centres, which use proton and carbon ions for ra-diotherapeutic purposes, are active. Beside this,

8/12/2019 Cirro Ne

2/4

2

a huge number of studies is still necessary to im-prove the quality of an ion therapy treatment.

These studies regard both a better understand-ing of the dose distributions inside patients (alsotaking into account the contribution due to thenuclear non-elastic interactions) as well as thepossibility to improve the design of the trans-port beam line dedicated to the ion therapeutictreatments. On the basis of the experience wegained in the last years in the realization of theCATANA facility [3] [4], dedicated to the treat-ment of ocular melanoma using 62 MeV protonbeams, and considering the future proton therapyfacilities being planned in the world, we decidedto start the development of a Monte Carlo appli-cation dedicated to proton therapy and based onthe GEANT4 toolkit [5].

2. Simulation of the experimental setup

2.1. GEANT4: a Monte Carlo simulation

toolkit

GEANT4 is a Monte Carlo simulation tool [5].It is not a stand-alone executable but a toolkit oflibraries; it was designed and developed by an in-ternational collaboration, formed by individualsfrom a number of cooperating institutes involved

with High Energy Physics, space and medical ex-periments. It builds on the accumulated experi-ence of many contributors to the fields of MonteCarlo simulation and of the physics of particledetectors. The transparency of its physics imple-mentation permits a rigorous procedure in vali-dating the physics results. On the other handthe flexibility of its design permits the simula-tion of physical processes in such different fields ashigh energy physics, space and cosmic ray appli-cations, nuclear and radiation computations, aswell as heavy ion and medical applications. In thepast a large number of Monte Carlo codes have

been developed for application in physics, partic-ularly for the nuclear/high energy processes. Inthe previous version of GEANT the physical pro-cesses at low energy were not treated in sufficientdetail. Only in the fourth version, a specific li-brary for low energy processes was inserted. Itprovides a dedicated set of libraries for the elec-tromagnetic processes which extends the applica-

ble energy range of electromagnetic interactionsdown to 250 eV for photons and electrons and

down to 1 keV for protons, ions and antiprotons.These energy ranges are now ideal for medical ap-plications. Moreover hadronic processes for thespecific medical range can also be implemented.

2.2. GEANT4 set up: simulation of the

beam line

In the application we developed all the elementsof a generic proton-therapy beam line can be pre-cisely reconstructed. The application permits todefine all the typical elements of a transport beamline for proton therapy, from the point in whichbeam exits in air (typically three meters beforethe patient) up to the isocenter. Such elementsare the scattering and collimator system, themonitor chambers, the range shifter (to degradethe beam energy), the final collimator; the geo-metrical and physical characteristics of these ele-ments can be easily defined and changed. More-over it is possible to choose the characteristics ofthe incident beam: beam energy, energy spread,beam spot size and angular spread. Two differ-ent detector can be simulated: a Markus chamber(for the depth dose curves reconstruction) and aGafChromic (for the lateral dose distributions).

As far as the physics processes are concerned, weactivated inside our simulation the Low Energyelectromagnetic package coupled with the Pre-Compound hadronic model. We tested our appli-cation simulating the CATANA [4] proton beaminstalled at Laboratori Nazionali del Sud (Figure1).

Figure 2 shows the simulation output of theproton beam line we obtained using the developedapplication. The arrow indicates the proton beamdirection.

2.3. Experimental set up

We have available two types of experimentaldata: depth dose (Bragg curves) and lateral dosedistributions acquired using a Markus type ion-isation chamber and GafChromic films, respec-tively. The Markus is a plane parallel ionisationchamber with an electrode spacing of 2 mm, asensitive air-volume of 0.05cm3 and a collectorelectrode diameter of 5.4 mm. We measured the

8/12/2019 Cirro Ne

3/4

3

Figure 1. The proton therapy beam line for thetreatment of ocular melanoma, installed at Labo-ratori Nazionali del Sud - INFN in Catania, Italy.

depth dose distributions in different solid mate-rials (aluminium, copper, polymethyl methacry-late) and in liquid water. In the first case thechamber is positioned inside a phantom of that

material and layers of the same are placed infront of it in order to degrade beam energy untilthe Bragg curve is reconstructed. In these casesthe spatial resolution of the measurement was 100m. For the measurement in water, the chamberis placed inside a water phantom and its move-ments are performed using an high-precision step-ping motor, fully controlled by a software devel-oped by us. The spatial resolution achieved forthe in-water measurements is 200 m. In bothcases (for in-solid and in-water measurements)the ionisation charge produced between the inter-electrode space is collected using a Keithley 7517

electrometer. GafChromic (ISP-HS model) films,used for the lateral dose distributions reconstruc-tion, are self-developing film media. The films de-velop a distinctive and characteristic colour uponexposure to ionising radiation and become pro-gressively darker in proportion to the absorbeddose. The advantages of GafChromic for the de-termination of the lateral dose distributions in-

Figure 2. The CATANA proton therapy beamline simulated using our GEANT4 application.The arrow indicates the proton beam direcion.

clude a high spatial resolution and tissue equiv-alence. Experimental lateral dose distribution

measurements were carried out placing the filmsin a polymethyl methacrylate (PMMA) holderat isocenter. A non-modulated 62 MeV protonbeam, 25 mm in diameter, was used for the irra-diation.

3. Results and Conclusions

We simulated the Bragg curve distributionusing different combinations of the availableGEANT4 physic models: the Standard elec-tromagnetic model, and the Low Energy one.Each of these was coupled to the Pre Com-

pound hadronic model. The activation of thehadronic model produces a decrease of the heightof the Bragg peak of about 8%. The best re-sults were obtained using the Low Energypackagetogheter with thePre Compoundhadronic model,as shown in Figure 3; here the simulation resultis compared with an experimental Bragg curveacquired using the Markus ionisation chamber.

8/12/2019 Cirro Ne

4/4

4

In order to fully exploit the potentialities of the

0 5 10 15 20 25 30 350

1

2

3

4

5

6

Depth [mm]

NormalizedDose[a.u.]

Only Low Energy packageLow Energy and Precompound hadronic modelExperimental Markus chamber data

Figure 3. Experimental and calculated Braggpeaks comparison. The results from simulationare obtained using only the Low Energypackageand coupling it with the Pre Compoundhadronicmodel.

Bragg peak in the clinical practice of the cancertreatment, it is necessary to spread its widthso that the entire tumour can be irradiated withthe same dose. This is done inserting along thebeam path a PMMA rotating wheel of varyingthickness. The result is the so-called Spread OutBragg Peak (SOBP). Our application permits toan user to simulate such modulation and to recon-struct the corresponding dose distribution. Fig-ure 4 shows a comparison between a simulatedand an experimental SOBP. It is a very prelim-inary result but that demostrates the ability of

the application to simulate also time-dependentbeam line elemnts.

The results obtained permit us to conclude thatthe Monte Carlo GEANT4 toolkit can be power-fully applied in the hadron therapy field for thesimulation of a complete treatment beam line.We validated the outputs of the simulations com-paring them with the experimental data available

0 5 10 15 20 25 30 35 40

0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Depth [mm]

NormalizedDose[a.u.]

GEANT4 SimulationExperimental Markus Chamber Data

Figure 4. Experimental and calculated SpreadOut Bragg Peak (SOBP). The experimental re-sult was acquired with a Markus type Ionisationchamber

in the CATANA facility and the results encour-age us to continue and refine our work in the nextmonths.

REFERENCES

1. J. Sisterson, links on ion ther-apy facilities, Particles Newsletters,http://ptcog.mgh.harvard.edu

2. B.G. Wouters, G. K. Lam et al., Radiat. Res.146 (1996) 159-170

3. L. Raffaele, G.A.P. Cirrone et al., PhysicaMedica XVII, Supplement 3 (2001) 35-40.

4. G.A.P. Cirrone, G. Cuttone et al., IEEETrans. Nucl. Sc. 51, No.3, June 2004

5. S.Agostinelli, J. Allison et al., Nucl. Instrum.

Method. A506 (2003) 250-303