Technological Transfer Technological Transfer from HEP to Medical Physicsfrom HEP to Medical Physics

How precise Brachytherapy MonteCarlo simulations can be applied in Clinics Reality

Problem: How to achieve accuracy and Problem: How to achieve accuracy and quickness?quickness?

Tests

Intermediate system between applications and

GRID

precision in calculation of dose distributionreproduction of the real geometry and tissues (CT ) calculation speedsimple to use ( for hospitals !)

Solution from HEP World : Geant4 + GRID + Solution from HEP World : Geant4 + GRID + WEBWEB

Functionalities (User Requirements)

Design

Software Process (USDP)A rigorous software process permits the development of reliable software:necessary feature for tools addressed to medical physics

Test from a microscopic point of view

S.Agostinelli1, F.Foppiano1, S.Garelli1, G.Ghiso2 , S.Guatelli 3, J.Moscicki4, M.G.Pia3,M.Tropeano5

1.Cancer Institute (IST), Genova,Italy 2.Servizio Sanitario Savona,Italy 3.INFN Genova,Italy, 4.CERN, Geneve,Switzerland, 5.University of Genova,Italy

DIANE R&D project: Application oriented gateway to GRID

DIANE permits the parallelisation on the

application and the access to GRID

The application developer is shielded

from complexity of underlying technology

Geant4 is an Object Oriented Toolkit for the simulation of the passage of particles through matter.Its application areas include high energy and nuclear physics experiments, astrophysics, medical physics,

radiation background studies, radioprotection and space science.Geant4 exploits advanced Software Engineering techniques and Object Oriented technology to achieve the

transparency of the physics implementation and hence provide the possibility of validating the physics results.

Geant4 has been developed and maintained by a world-wide collaboration of more than 100 scientists. The source code and libraries are freely distributed from the Geant4 web site www.cern.ch/geant4.

GRID is a project funded by European Union. The objective is to build the next generation computing

infrastructure providing distributed computering resources across the world

Brachytherapy is a medical therapy used for cancer treatmentcancer treatment

Radioactive sources deliver therapeutic dose to tumors,

preserving the surrounding healthy tissues

Interstitial brachytherapy

( prostate)Endocavitary brachytherapy

(lungs,vagina,uterus)Superficial brachytherapy

(skin)

3 m m ste e l c a b le

5.0 m m

0.6 m m

3.5 m m

1.1 m m

Ac tive Ir-192 C o re

Three devices:

BrachytherapyBrachytherapy

Calculation of dose delivered to tissues

(as for example Prowes, Variseed V7) used in

clinical practice

Analytical calculation methodsAll the tissues are approximated to water

Advantages High calculation speed

Disadvantages Approximated dose calculation density insensitivity The source is approximated to a point

Software characteristics for brachytherapySoftware characteristics for brachytherapy

Commercial software availableCommercial software availableSuch a software

does not exist for superficial

brachytherapy!

Speed is a fundamental requirement for software used in the clinical practice

The medical physicist sometimes has got to take decisions about the position

of the sources in few seconds

MC simulation were never been used in the clinical practice for the long calculations

MonteCarlo simulations are more accurate in the dose calculation but to slow for a realistic clinical use

AnalysisAIDA/Anaphe

Precision

Precision

Real geometry reproduction

Real geometry reproduction

Simple to use , Use in hospitals

Simple to use , Use in hospitals

Calculation speed

Calculation speed

Other user requirements

Other user requirements

MonteCarlo methodAccurate physical processes simulationTest to guarantee the quality of the software

Accurate description of the geometryPossibility to interface the software to CT

Graphic visualisation + user interfaceDose distribution analysis ( i.e. isodose curves)

Parallel systemCalculation shared resource access

Estensivity to new functionalitiesPublic access

Project : Project : development of a software for dose calculation which is accurate in the dose calculation and fast by a clinical use point of view

3D dose distribution calculation

Isodose curves

Calculation speedDIANE, GRID

SimulationGeant4

Tests on Geant4 e- and gamma electromagnetic processes :

CSDA range for e- and Gamma attenuation coefficient in different absorber materials

Comparison of different Geant4 physics models (Standard/LowEnergy/Penelope)

Comparison with protocol data ( ICRU 37 and ICRU 49)

Macroscopic TestTest about the dose distribution of brachytherapic

sources ( I-125, Ir –131) along the perpendicular to the

major axes on the source .

•G4LowE EPDL•G4Standard•NIST reference data

•Penelope gamma processes

•Reference data NIST

Geant4 results have been compared with

experimental measurements and protocol reference data (TG43 and

Italian Association of Medical Physics (AIFB) protocol)

The experimental dosimetric measurements of the seed Microselectron HDR (Ir-131)

have been performed with ionisation chambers at the Italian National Institute

of Cancer (IST), Genova (Italy)

The experimental measurements of the seed Bebig Isoseed I-125 have been

performed with films at the Medical Physics Institute of Savona (Italy)

G. Ghiso and S.Guatelli

At Medical Physics Institute in Savona(Italy)

Brachytherapy application Dose distribution calculation

Isodose curves

Development of the project

Sofware planning and

development

For all the brachytherapic devices

Generalisation of the softwareGeneralisation of the software

Each treatment planning software is specific to one brachytherapic technique.

Treatment planning software is expensive

(~ hundreds k Euro)

The software we developed is transparent to the particular brachytherapic source; this is possible because Geant4 simulates the involved physics without any approximation.

3D dose distribution Isodose curves Choice of the materials of the phantom Graphical visualization Possibility to interface the system to CT source compositiongeometry, materials, spectrum

Use of abstract classesRadioactive source definition thanks to the design pattern: Abstract Factory

Simulation result: energy depositAnalysis: dose distribution and isodose curves

Parametrization of the volumes in the geometry Parametrization function: volume -> material

Generalization+ Specific aspects of the source

Interface to CT

Dosimetry

FunctionalitiesFunctionalities

Some Results

The radioactive source is

positioned in the center of a phantom

filled with water. The results can be generalised to

more seed and in a human anatomy

Isodose curves

Dose distribution of a MicroSelectron- HDR source (Ir-131)

Isodose Curves

The plots describe the dose distribution in plans parallel to the one containing the source (y = 0. mm)

Dose distribution for Bebig Isoseed I-125 The Leipzig applicator is

positioned on the

phantom

Problem: The simulations are very long (compared to clinical use scale of time) in order to obtain results with high statistics.

Leipzig applicator

Performance

Endocavitary

brachytherapy

Interstitial

brachytherapy

Superficial

brachytherapy

1M events

61 minutes

1M events

67 minutes

1M events

65 minutes

On a average PIII machine,As an “average hospital” may have

The simulations are two

slow

The time required to obtain

results with high statistics

do not permit the use of MC

in clinical practice

Parallel cluster processing

Make fine tuning and customisation easy

Transparently using the GRID

Application independent

The application developer is shielded from complexity

and underlying technology

Not affecting the original code application Good separation of the subsystem

the application does not need to know that it runs in distributed environment

Performance in parallel mode

Endocavitary

brachytherapy

Interstitial

brachytherapy

Superficial

brachytherapy

1M events

4 minutes, 34 sec

1M events

4 minutes, 36 sec

1M events

4 minutes, 25 sec

On up to 50 workers, LSF at CERN,

PIII machine, 500/1000 MHz

It is not realistic to expect such

CPU resources in the “ average hospital”

Via DIANE

A hospital is not required to own and maintain extensive

Computing resources to exploit the scientific advantages

of MonteCarlo simulation of radiotherapy

Any hospital –even small ones- or in less wealthy countries,

that can not afford expensive commercial software systems –

may have access to advanced software technologies and

tools for radiotherapy

Work in progress: submission to the GRID and retrieval of

results from a web portal(to facilitate the usage by end-users)

Parallelisation and access to the GRIDParallelisation and access to the GRID Running on the GRID

Migration to distributed Migration to distributed environmentenvironment