Validating Monte Carlo as a clinical tool
for commissioning the UCLH proton therapy
centre
Presented by: Vasilis Rompokos,
Malet Place - UCL,
22nd of April 2015
(work by: Ilina Vasileva Korn)
Proton Therapy
Monte Carlo
Both sexy
Reality
Monte Carlo is
neither accurate
nor precise
Project aim:
The primary and most important aim is to treat patients
by 2018!
Project aim:
•
Check how bad Geant4 is doing and optimise it
•
Simulate the nozzle commissioning data that are
required by the treatment planning system
•
Commission the treatment planning system
•
Create the local protocols
Background:
Effect of range cut and step size on 230MeV protons
Grevillot L et al. “Optimization of GEANT4 settings for proton pencil beam scanning
simulations using GATE” Nucl Instrum Methods Phys Res B 268:3295-3305
Background:
Effect of precompound Nuclear models (Physics-lists)
Grevillot L et al. “Optimization of GEANT4 settings for proton pencil beam scanning
simulations using GATE” Nucl Instrum Methods Phys Res B 268:3295-3305
Background:
Nuclear halo
Dose:
- core (primary p)
- halo ~ 1/3 beam
range (secondary p)
- aura (neutrons, γ)
Gottschalk B et al. “Nuclear halo of 177MeV proton beam in water” Med Phys submitted 2014
Geant4 code
The Monte Carlo code uses Geant4 v10.0.02. The code is installed
on the plus1 HEP cluster at UCL.
Beam parameters
position, energy, energy
spread, spot shape and
size, direction and angular
dispersion
Phantom
Water, voxels
Output
Scored absorbed dose
in voxels, text and root files
Test the physics list and dose scoring algorithms
Phantom
Water tank *
- 30x30x30 cm, ionization potential=75eV (ICRU37 and 49)
- rows of 85 voxels: 0.4x0.4x0.35 cm (sensitive volume of Extradin
T1) at 10 transverse distances (0-10 cm)
Beam parameters *
- Position: at the water tank wall
- Mono-energetic beam of 177 MeV,
no energy spread
- Radius: 5.948 mm
- Angular dispersion: 31.55 mrad
* “Nuclear halo of a 177 MeV proton beam in water: theory, measurement and
parametrization” (arXiv:1409.1938v1, 2014, MGH Burr Proton Therapy Center)
Physics list
G4VUserPhysicsList
- G4EmStandardPhysics_option3
- G4HadronPhysicsQGSP_BIC
- G4EmExtraPhysics *
- G4HadronElasticPhysics
- G4StoppingPhysics
- G4IonBinaryCascadePhysics
- G4NeutronTrackingCut *
- G4DecayPhysics
Geant4
Hadrontherapy
advanced
example
physics
list
(QGSP_BIC_EMY). The Hadrontherapy example simulates the
passive transport beam line installed at Laboratori Nazionali del Sud
(INFN) in Catania, Italy.
* TOPAS, Paganetti
GATE, Physics List for Proton Therapy Applications
Scoring absorbed dose
Dose in 10 transverse
distances in 85 voxels scored
in two ways:
1) G4ScoringManager
and
2) G4UserSteppingAction
G4UserRunAction
- The maximum step size is 0.1 mm – distance to the next interaction
or to the next geometrical boundary (5% of voxel thickness)
- Production thresholds for e-, e+, γ is 0.1 mm *
* “Optimization of Geant4 settings for proton pencil beam scanning simulations using
GATE”( Nuc.Inst. and Methods in Phys.Research B 268, 2010)
Simulation vs Measurement
(Preliminary results)
Simulation of 106 protons
Simulation vs Measurement
(Preliminary results)
Simulation vs Measurement
(Preliminary results)
Simulation vs Measurement
(Preliminary results)
Simulation vs Measurement
(Preliminary results)
Simulation vs Measurement
(Preliminary results)
Future work
Implementation:
- study the physics list
- more variables (LET, fluence, dose per particle type, dose per
physics process)
- implement sensitive detector
Validation:
- understand the 177 MeV measurements
- data from the PBS nozzle at the Hospital of the University of
Pennsylvania
PBS data from the Hospital of
the University of Pennsylvania
100 MeV – 226.7 MeV
PBS data from the Hospital of
the University of Pennsylvania
Take home message:
If User <> “theoretical background understanding” Then
If CodeSetup <> “optimised for the specific use” Then
MonteCarlo <> “Accurate”
MonteCarlo <> “Precise”
End If
End If
Questions?
Email: vasilis.rompokos@uclh.nhs.uk
Monte Carlo data
Solid line: Nuclear reactions turned-off
{Proton Therapy Physics (Harald Paganett - p.54)}
Moyers MF et al “Calibration of a proton beam energy monitor”, Medical
Physics, 2007 34(6): 1952-1966
{Although there are small differences in various correction factors used to calculate the Janni and
ICRU stopping power tables, the primary difference is the mean excitation energy or I value.}