Maria Grazia Pia, INFN Genova Technology transfer from HEP computing to the medical field F....

Maria Grazia Pia, INFN Genova

Technology transfer from HEP Technology transfer from HEP computing to the medical fieldcomputing to the medical field

F. Foppiano3, S. Guatelli2, J. Moscicki1, M.G. Pia2, M. Piergentili2

CERN1

INFN Genova2

National Institute for Cancer Research, IST Genova3

Topical Seminar on Innovative Radiation Detectors Siena, 23-26 May 2004

http://www.ge.infn.it/geant4/talks

S. Agostinelli, S. Garelli (IST Genova)L. Archambault, L. Beaulieu, J.-F. Carrier, V.-H. Tremblay (Univ. Laval)

M.C. Lopes, L. Peralta, P. Rodrigues, A. Trindade (LIP Lisbon) G. Ghiso (S. Paolo Hospital, Savona)

Including contributions from:


A real life case

A dosimetric system for A dosimetric system for brachytherapy derived from HEP brachytherapy derived from HEP

computingcomputing(but all the developments and applications presented in this talk are general)

Technology transferTechnology transfer

Activity initiated at IST Genova, Natl. Inst. for Cancer Research (F. Foppiano et al.)– hosted at San Martino Hospital in Genova (the largest hospital in Europe)

Collaboration with San Paolo Hospital, Savona (G. Ghiso et al.)– a small hospital in a small town


The goal of radiotherapyThe goal of radiotherapy

Delivering the required therapeutic dose to the tumor area with high precision,

while preserving the surrounding healthy tissue

Dosimetry system precision accurate model of the real configuration (from CT)

speed adequate for clinical use user-friendly interface for hospital usage

Calculate the dose released to the patient by the

radiotherapy system

Accurate dosimetry is at the basis of radiotherapy treatment planning


The realityThe reality

Treatment planning is performed by means of commercial software

The software calculates the dose distribution delivered to the patient in a given source configuration

Open issues

PrecisionPrecision CostCost

Commercial systems are based on approximated analytical methods,approximated analytical methods, because of speed constraints

Approximation in geometry modelinggeometry modeling

Approximation in material modeling material modeling

Each treatment planning software is specific to one techniquespecific to one technique and one type of sourceone type of source

Treatment planning software is expensiveexpensive


Commercial factorsCommercial factorsCommercial treatment planning systems are governed by commercial rules (as any other commercial product...)

i.e., they are produced and marketed by a company only if the investment for development is profitable

No commercial treatment planning systems are available for non-conventional radiotherapy techniques such as hadrontherapyhadrontherapy

or for niche applications such as superficial brachytherapysuperficial brachytherapy

0

20

40

60

80

100

-20 -15 -10 -5 0 5 10 15 20

Distance from the centre (mm)

Sig

nal %

Film X

Microcubes X

Treatment planning systems for hadrontherapy are quite primitive

not commercially convenient so far


Monte Carlo methods in radiotherapyMonte Carlo methods in radiotherapy

Monte Carlo methods have been explored for years as a tool for precise dosimetry, in alternative to analytical methods

de facto,

Monte Carlo simulation is not used in clinical practice

(only side studies)

The limiting factor is the speedspeedOther limitations: reliable? for “software specialists only”, not user-friendly for general practice requires ad hoc modeling


CT-simulation with a Rando phantomExperimental data with TLD LiF dosimeter

CT images used to define the geometry:

a thorax slice from a Rando

anthropomorphic phantom

Comparison with commercial treatment

planning systems

Comparison with commercial treatment

planning systems

M. C. LopesIPOFG-CROC Coimbra Oncological Regional Center

L. Peralta, P. Rodrigues, A. TrindadeLIP - Lisbon

Central-Axis depth dose

Profile curves at 9.8 cm depth

PLATO overestimates the dose at ~ 5% level


M. C. Lopes1, L. Peralta2, P. Rodrigues2, A. Trindade2

1 IPOFG-CROC Coimbra Oncological Regional Center - 2 LIP - Lisbon

Head and neck with two opposed beams for a 5x5 and 10x10 field size

A more complex set-upA more complex set-up

An off-axis depth dose taken at one of the slices near the isocenter

PLATO fails on the air cavities and bone structures and cannot predict accurately the dose to tissue that is surrounded by air

Deviations are up to 25-30%

Beam planeSkull bone

Tumor

Air

Bone

In some tumours sites (ex: larynx T2/T3-stage) a 5% underdosage will decrease local tumour

control probability from ~75% to ~50%


The challenge


Develop a Develop a general purposegeneral purpose

precise precise dosimetric system

with the capability of

realistic geometryrealistic geometryand material modelingand material modeling

interface to CT imagesinterface to CT images

with a user-friendly interfaceuser-friendly interface

atat low costlow cost

adequate adequate speedspeed for clinical usage for clinical usageperforming atperforming at


PrecisionPrecision

Accurate model of the real experimental set-upAccurate model of the

real experimental set-up

Easy configuration for hospital usage

Easy configuration for hospital usage

SpeedSpeed

Calculation of 3-D dose distribution3-D dose distribution in tissueDetermination of isodose isodose curves

Based on Monte CarloMonte Carlo methodsAccurate description of physicsphysics interactionsExperimental validationvalidation of physics involved

Realistic description of geometrygeometry and tissuetissuePossibility to interface to CT images

Simple user interface + Graphic visualisation Elaboration of dose distributionsdose distributions and isodosesisodoses

ParallelisationParallelisationAccess to distributed computing resourcesdistributed computing resources

Other requirementsOther requirementsTransparentTransparentOpen to extension extension and new functionalityPublicly accessiblePublicly accessible

RequirementsRequirementsRequirementsRequirements


PrecisionPrecisionPrecisionPrecision

Based on Monte Carlo methodsMonte Carlo methods

Extension of electromagnetic interactions down to low energies (< 1 keV)

Microscopic validation of the physics modelsMicroscopic validation of the physics modelsComparison Comparison with experimental data experimental data

specific to the brachytherapic practice

Accurate description of physicsphysics interactions

Experimental validationvalidation of physics involved


Code and documentation publicly distributed from web

1st production release: end 1998– 2 new releases/year since then

Developed and maintained by an international collaboration of physicists and computer scientists

Run, Event and Track management PDG-compliant Particle management Geometry and Materials Tracking Detector response User Interface Visualisation Persistency Physics Processes


Barkas effect (charge dependence)models for negative hadrons

e,down to 250 eV

EGS4, ITS to 1 keVGeant3 to 10 keV

Hadron and ion models based on Ziegler and ICRU data and parameterisations

Based on EPDL97, EEDL and EADL evaluated data libraries

Bragg peak

shell effects

antiprotons

protonsions

Fe lines

GaAs lines

Atomic relaxation Fluorescence

Auger effect

Based on Penelope analytical models


Validation

Microscopic validation:

verification of Geant4 physicsverification of Geant4 physics

Dosimetric validation: in the experimental contextin the experimental context


proton straggling

ions

e-, Sandia database

Al

NISTGeant4-LowEGeant4-Standard

Stopping power

Microscopic validationMicroscopic validation

many more

validation results

available!

2N-L=13.1 – =20 - p=0.87

NISTGeant4-LowEGeant4-Standard

Photon attenuation coefficient

Al

2N-S=23.2 – =15 - p=0.08


Dosimetric validationDosimetric validation

0 10 20 30 40 500,0

0,2

0,4

0,6

0,8

1,0

1,2 Simulazione Nucletron Misure

Dose %

Distanza lungo Z (mm)Distance along Z (mm)

SimulationNucletronData

F. Foppiano et al., IST Genova

Comparison to

manufacturer data, protocol data,

original experimental data

experimental mesurements

G. Ghiso, S. Guatelli S. Paolo Hospital Savona

Ir-192 I-125


General purpose systemGeneral purpose systemGeneral purpose systemGeneral purpose system

Object Oriented technologySoftware system designed in terms of Abstract Interfaces

Abstract Factory design patternSource spectrum and geometry transparently interchangeableSource spectrum and geometry transparently interchangeable

For any brachytherapy technique

For any source type


Flexibility of modelingFlexibility of modeling

CT DICOM interface

through Geant4 parameterised volumesGeant4 parameterised volumes parameterisation function: materialparameterisation function: material

Abstract Factory

Configuration of

any brachytherapy technique any brachytherapy technique

any source type any source type

through an Abstract FactoryAbstract Factory to define geometry, primary geometry, primary spectrumspectrum

Phantom

various materialsvarious materials water, soft tissue, bone, muscle etc.

General purpose software system for brachytherapy

No commercial general software exists!


Realistic model Realistic model of the experimental set-upof the experimental set-up

Realistic model Realistic model of the experimental set-upof the experimental set-up

Spectrum (192IrIr, 125II)Geometry

Phantom with realistic material modelPhantom with realistic material modelPossibility to interface the system to CT imagesPossibility to interface the system to CT images

Radioactive source

Patient


Modeling the source geometry

Modeling the source geometry

Precise geometry and material model of any type of source

Iodium core

• Iodium core• Air• Titanium capsule tip• Titanium tube

Iodium core:Inner radius :0Outer radius: 0.30mmHalf length:1.75mm

Air:Outer radius:0.35mm half length:1.84mm

Titanium tube:Outer radius:0.40mmHalf length:1.84mm

Titanium capsule tip:Box Side :0.80mm

I-125 source for interstitial brachytherapy

Ir-192 source + applicator for superficial brachytherapy3 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


Effects of source anisotropyEffects of source anisotropyEffects of source anisotropyEffects of source anisotropy

LongitudinalLongitudinal axis of the source axis of the sourceDifficult to make direct measurements

TransverseTransverse axis of the source axis of the sourceComparison with experimental data

Plato-BPS treatment planning algorithm makes some crude

approximation ( dependence,

no radial dependence)

-40 -30 -20 -10 0 10 20 30 400,0

0,5

1,0

1,5

2,0

2,5

Simulazioni Plato Misure

Dose %

Distanza lungo X (mm)Distance along X (mm)

SimulationSimulationPlatoPlatoDataData

-40 -30 -20 -10 0 10 20 30 400,0

0,5

1,0

1,5

2,0

2,5 Simulazioni Plato

Dose %

Distanza lungo Z (mm)

Distance along Z (mm)

Effects of source

anisotropySimulationSimulation

PlatoPlato

Rely on simulation for better accuracy than

conventional treatment planning software


Modeling the patientModeling the patient

source

Modeling a phantomModeling a phantom

of any material (water, tissue, bone, muscle etc.)thanks to the flexibility of Geant4

materials package

Modeling geometry and materials from CT Modeling geometry and materials from CT datadata

Acquisition of CT image 3D patient anatomy

file

DICOM is the universal standard for sharing resources between heterogeneous and

multi-vendor equipment

3-D view

Geant4-DICOM interface developed by L. Archambault, L. Beaulieu, V.-H. Tremblay

(Univ. Laval and l'Hôtel-Dieu, Québec)


User-friendly interfaceUser-friendly interfaceto facilitate the usage in hospitalsto facilitate the usage in hospitals

User-friendly interfaceUser-friendly interfaceto facilitate the usage in hospitalsto facilitate the usage in hospitals

Graphic visualisation of dose distributionsElaboration of isodose curves

Application configurationApplication configurationJob submissionJob submission

Dosimetric analysis

Web interface


DosimetryDosimetryDosimetryDosimetry

AIDA + AnaphePython

Analysis of the energy deposit in the phantom resulting from the simulation

Dose distribution

Isodose curves

for analysisfor interactivity

+ any AIDA-compliant analysis system

Simulation of energy deposit through Geant4 Low Energy Electromagnetic package

to obtain accurate dose distributionProduction threshold: 100 m

2-D histogram with energy deposit

in the plane containing the source

Abstract Interfaces for Data Analysis


MicroSelectron-HDR source

DosimetryEndocavitary brachytherapy

DosimetrySuperficial brachytherapy

Leipzig applicator

Dosimetry Interstitial brachytherapy

Bebig Isoseed I-125 source


Application configurationApplication configuration

Fully configurable from the web

Type of source

Phantom configuration

# events

Run modes:

demo

parallel on a cluster (under test)

on the GRID (under development)


Speed adequate for clinic useSpeed adequate for clinic useSpeed adequate for clinic useSpeed adequate for clinic use

Transparent configuration in sequential or parallel mode

Transparent access to the GRID through an Transparent access to the GRID through an intermediate software layerintermediate software layer

Parallelisation

Access to distributed computing resources


PerformancePerformance

Endocavitary brachytherapy

1M events

61 minutes

Interstitial brachytherapy

1M events

67 minutes

Superficial brachytherapy

1M events

65 minutes

on an “average” PIII machine, as an “average” hospital may own

Monte Carlo simulation is not practically conceivable for clinical application, even if more precise


DIANE DIstributed ANalysis EnvironmentDIANE DIstributed ANalysis Environment

prototype for an intermediate layer between applications and the GRID

Hide complex details of underlying technology

Developed by J. Moscicki, CERN

http://cern.ch/DIANE

R&D in progress forR&D in progress forLarge Scale Large Scale

Master-WorkerMaster-Worker ComputingComputing

DIANEDIANE

Parallelisation Access to the GRID

Transparent access to a distributed computing environment


Performance: parallel mode on a local cluster

Performance: parallel mode on a local cluster

1M events

4 minutes 34’’

5M events

4 minutes 36’’

1M events

4 minutes 25’’

on up to 50 workers, LSF at CERN, PIII machine, 500-1000 MHz

Performance adequate for clinical application, but…

it is not realistic to expect any hospital to own and maintain a PC farm

Endocavitary brachytherapy

Interstitial brachytherapy

Superficial brachytherapy

preliminary: further optimisation in progress


Running on the GRIDRunning on the GRIDVia DIANE

Same application code as running on a sequential machine or on a dedicated cluster

– completely transparent to the user

A hospital is not required to own and maintain extensive computing resources to exploit the scientific advantages of Monte Carlo simulation for radiotherapy

Any hospital

– even small ones, or in less wealthy countries, that cannot even small ones, or in less wealthy countries, that cannot afford expensive commercial software systemsafford expensive commercial software systems –

may have access to advanced software technologies and tools for radiotherapy


Traceback from a run on CrossGrid testbed

Traceback from a run on CrossGrid testbed

Current #Grid setup (computing elements):5000 events, 2 workers, 10 tasks (500 events each)

- aocegrid.uab.es:2119/jobmanager-pbs-workq- bee001.ific.uv.es:2119/jobmanager-pbs-qgrid- cgnode00.di.uoa.gr:2119/jobmanager-pbs-workq- cms.fuw.edu.pl:2119/jobmanager-pbs-workq- grid01.physics.auth.gr:2119/jobmanager-pbs-workq- xg001.inp.demokritos.gr:2119/jobmanager-pbs-workq- xgrid.icm.edu.pl:2119/jobmanager-pbs-workq- zeus24.cyf-kr.edu.pl:2119/jobmanager-pbs-infinite- zeus24.cyf-kr.edu.pl:2119/jobmanager-pbs-long- zeus24.cyf-kr.edu.pl:2119/jobmanager-pbs-medium- zeus24.cyf-kr.edu.pl:2119/jobmanager-pbs-short- ce01.lip.pt:2119/jobmanager-pbs-qgrid

Spain

Poland

Greece

Portugal

Resource broker running in Portugal

matchmaking CrossGrid computing elements


Extension and evolutionExtension and evolution

General dosimetry system for radiotherapyGeneral dosimetry system for radiotherapy extensible to other techniques

plug-ins for external beamsplug-ins for external beams

((factories for beam, geometry, physics...)

Configuration of

any brachytherapy technique any brachytherapy technique

any source type any source type

Plug-ins in progress

System extensible to any source configuration

without changing the existing code

treatment headtreatment head hadrontherapyhadrontherapy ......


A medical accelerator for IMRTA medical accelerator for IMRT

Build a simulation tool which determines the dose distributions given in a phantom by the head of a linear accelerator used for IMRT.

Many algorithms were developed to estimate dose distributions, but even the most sophisticated ones resort to some approximations. These approximations might affect the outcome of dose calculation, especially in a complex treatment planning as IMRT.

step and shoot

IMRT generates tightly conforming dose distributions.

This microscopic control allows IMRT to produce dose distribution patterns that are much closer to the desired patterns

than possible previously


The user can choose the energy and standard deviation of the primary particles energy distribution (Gaussian)

The primary particles (e-) leave from a point source with random direction (0˚< θ < 0.3˚) and a gaussian distribution

The head components modeled include: target, primary and secondary collimators, vacuum window, flattening filter, ion chamber, mirror, vacuum and air

Each pair of jaws can be rotated through an axis that is perpendicular to the beam axis

The actual analysis produces some histograms from which the user can calculate the Percent Depth Dose (PDD) and the flatness at the following depths in the phantom: 15 mm, 50 mm, 100 mm and 200 mm.

Work in progress...Work in progress...


DesignDesign


(very) Preliminary results(very) Preliminary results

FlatnessPercent Depth

Dose


Real hadron-therapy beam line

GEANT4 simulation

CATANA hadrontherapy

CATANA hadrontherapy

talk by P. Cirrone on Monday


Dosimetry in interplanetary missions

Aurora Programme

Dose in astronaut resulting from Galactic Cosmic Rays

vehicle concept


ConclusionsConclusions

Physics & software technology from HEP have a potential to address key issueskey issues in medical physics

The social impactsocial impact of technology transfer from HEP computing may be significant

What is the support of HEP to technology transfer?

Geant4 + AIDA/Anaphe/PI

+ WWW + DIANE + GRID

=

precise, versatile,

fast, user-friendly,

low-cost dosimetry


Thanks!Thanks!

G. Cosmo (CERN, Geant4)

L. Moneta, A. Pfeiffer(Anaphe/PI, CERN)

J. Knobloch (CERN/IT)

S. Agostinelli, S. Garelli (IST Genova)

G. Ghiso, R. Martinelli (S. Paolo Hospital, Savona)

G.A.P. Cirrone, G. Cuttone (INFN LNS, CATANA project)

M.C. Lopes, L. Peralta, P. Rodrigues, A. Trindade (LIP Lisbon)

L. Archambault, J.F. Carrier, L. Beaulieu, V.H. Tremblay (Univ. Laval)

This project has fostered a collaborative aggregation of contributions from many groups all over the world

the authors F. Foppiano (IST) – medical physicistmedical physicistS. Guatelli, M. Piergentili (Univ. and INFN Genova) – studentsstudents

J. Moscicki (CERN) – computer scientistcomputer scientistM.G. Pia (INFN Genova) – particle physicistparticle physicist

Maria Grazia Pia, INFN Genova Technology transfer from HEP computing to the medical field F....

Documents

Transcript of Maria Grazia Pia, INFN Genova Technology transfer from HEP computing to the medical field F....