A review of FLUKA
applications for medical
physics
G. Battistoni, INFN Milano
Contributions of:
T.T. Böhlen, F. Cappucci, P. Colleoni, M. Chin, A. Ferrari, A. Mairani,
S. Muraro, R. Nicolini, P. Ortega, K. Parodi, V. Patera, P. Sala,
V. Vlachoudis (and others)
FLUKA
Main authors:
A. Fassò, A. Ferrari, J. Ranft, P.R. Sala
Contributing authors: G. Battistoni, F. Cerutti, M. Chin, T. Empl, M.V. Garzelli, M.
Lantz, A. Mairani, V. Patera, S. Roesler, G. Smirnov,
F. Sommerer, V. Vlachoudis
Developed and maintained under an INFN-CERN
agreement Copyright 1989-2013 CERN and INFN
>5000 users
http://www.fluka.org
The FLUKA International Collaboration
M. Brugger, M. Calviani, F. Cerutti, M. Chin, Alfredo Ferrari, P. Garcia Ortega, A. Lechner, C. Mancini-Terracciano,
M. Magistris, A. Mereghetti, S. Roesler, G. Smirnov, C. Theis, Heinz Vincke, Helmut Vincke, V. Vlachoudis, J.Vollaire, CERN
G. Kharashvili, Jefferson Lab, USA
J. Ranft, Univ. of Siegen, Germany
G. Battistoni, F. Broggi, M. Campanella, F. Cappucci, E. Gadioli, S. Muraro, R. Nicolini, P.R. Sala, INFN & Univ. Milano, Italy
L. Sarchiapone, INFN Legnaro, Italy
G. Brunetti, A. Margiotta, M. Sioli, INFN & Univ. Bologna, Italy
V. Patera, INFN Frascati & Univ. Roma Sapienza, Italy
M. Pelliccioni, INFN Frascati & CNAO, Pavia, Italy
M. Santana, SLAC, USA
A. Mairani, CNAO Pavia, Italy
M.C. Morone, Univ. Roma II, Italy
K. Parodi, I. Rinaldi, LMU Munich, Germany
L. Lari, Univ. of Valencia, Spain A. Empl, L. Pinsky, B. Reddell, Univ. of Houston, USA
V. Boccone, Univ. of Geneva, Switzerland
M. Nozar, TRIUMF, Canada
K.T. Lee, T. Wilson, N. Zapp, NASA-Houston, USA
T. Boehlen, S. Rollet, AIT, Austria
A. Fassò, R. Versaci, ELI-Beamlines, Prague, CR
S. Trovati, PSI, Switzerland
M.V. Garzelli, Nova Gorica Univ., Slovenia
M. Lantz, Uppsala Univ., Sweden
P. Colleoni, Ospedali Riuniti di Bergamo, Italy
Anna Ferrari, S. Mueller HZDR Rossendorf, Germany
The Physics Content of FLUKA

60 different particles + Heavy Ions

Nucleus-nucleus interactions from Coulomb barrier up to 10000
TeV/n

Electron and μ interactions 1 keV – 10000 TeV

Photon interactions 100 eV - 10000 TeV

Hadron-hadron and hadron-nucleus interactions 0–10000 TeV

Neutrino interactions

Charged particle transport including all relevant processes

Transport in magnetic fields

Neutron multigroup transport and interactions 0 – 20 MeV

Analog calculations, or with variance reduction
RF
FLUKA Applications
Momentun
Cleaning
CMS
Point 4
Point 5
LHC Dum
Point 3.3
Point 3.2
The LHC
Loss Regions
Point 6
Regions of high losses
(e.g., Collimators,…)











Point 2
Regions with low losses
(e.g., due to residual gas)
Point 7
Cosmic ray physics
ALICE
Neutrino physics
LHCb
Accelerator design ( n_ToF, CNGS, LHC systems)
ATLAS
Particle physics: calorimetry, tracking and detector simulation etc.
( ALICE, ICARUS, ...)
ADS systems, waste transmutation, (”Energy amplifier”, FEAT, TARC,…)
Shielding design
Dosimetry and radioprotection
Radiation damage
Space radiation
Hadron therapy
Neutronics
Point 8
Point 1
Betatr
Cleani
Application for medicine: some examples
• Nuclear Medicine
o Dosimetry
• Radiotherapy
o Simulation of therapy devices
o Check of treaments
• Hadrontherapy
o Commissioning of facilities
o Treatment planning and forward checks
o Predictions for monitoring applications (imaging for
hadrontherapy)
o Design of instruments, dosimetry
o Calculation for shielding and rad. protection in facilities
The FLUKA voxel geometry
It is possible to describe a
geometry in terms of “voxels”,
i.e., tiny parallelepipeds (all of
equal size) forming a 3dimensional grid
anthropomorphic
phantom
You can import
a CT scan to a
FLUKA Voxel
Geometry
Now available the official ICRP Human Phantom
ICRP Publication 110: Adult Reference Computational Phantoms Annals of the ICPR Volume 39 Issue 2 7
CT stoichiometric calibration
CT segmentation into 27 materials of defined elemental
composition (from analysis of 71 human CT scans)
Air, Lung,
Adipose tissue
Soft tissue
Skeletal tissue
Schneider et al PMB 45, 2000
CT
stoichiometric
calibration
(II)
Assign to each material a “nominal mean density”, e.g. using the
density at the center of each HU interval (Jiang et al, MP 2004)
Schneider et al
PMB 45, 2000
But “real density” (and related physical quantities) varies
continuously with HU value: a HU-dependent correction on
density on each voxel is applied
Application in nuclear medicine
Radioactive source decay
FLUKA contains data about decaying schemes of radioactive isotopes,
allowing to select an isotope as radiation source. Complete databases are
generated from the data collected from National Nuclear Data Center
(NNDC) at Brookhaven National Laboratory.
Application in nuclear medicine
Calculation of absorbed dose at voxel level starting from 3D images of activity
distribution (SPECT, PET images)
Simulations in homogeneous water
Simulated 99Tc-SPECT of water phantoms (SIMIND code):
Dose calculation: Cylinder + spheres filled with 90Y
#1
#2
SPECT/PET - CT images handling
DOSE Maps
VOXEL
Dosimetry
Collaboration
INFN and IEO
MONTE CARLO
109 particles
With 109 particles simulated, FLUKA and VOXEL DOSIMETRY
(a standard analytic procedure in nuclear medicine)
results in water agree within 5%
Applications in radiotherapy
IORT
Simulation of a Linac for RadioTherapy
15
6 MeV Accelerator –photon fluence
Dosimetric validation
The Leksell Gamma Knife Perfexion:
The Leksell Gamma Knife Perfexion
(LGK-PFX) is a 60Co based medical
device, manufactured by Elekta AB
Instruments Stockholm, Sweden. The It
is emplyed in the cure of different brain
pathologies: small brain and spinal
cord tumors (benign
and malignant), blood
vessel abnormalities,
as well as neurologic
problems can be fully
treated.
Fabrizio Cappucci
INFN, Milan.
The Leksell Gamma Knife Perfexion:
The ionizing gamma radiation is
emitted from 192 60Co sources
(average activity ~1TBq each).
The sources are arranged on
identical sectors of 24 elements.
8
The sectors can be placed in
correspondence of
three different
collimation set able to focus the
gamma rays on a common spot, called
the isocenter of the field, having a
radial dimension of about 4, 8 and 16
mm respectively.
Fabrizio Cappucci
INFN, Milan.
Geometry Modelization: Materials
~ 1350 bodies
Protective shield;
LEAD.
Collimator channels;
TUNGSTEN.
Gammex 457
Solid Water
Isocenter of
the field.
Radiative sources
encapsulated in
stainless steel
bushings.
Thanks
to the collaboration with ELEKTA, which provided, under a confidential
Fabrizio
Cappucci
agreement
the detail of the geometry and all the involved material, has been
INFN,
Milan.
possible to implement an accurate model for the radiation unit.
Source Modeling: Geometry and materials
Metallic bushing.
60Co
cylindrical pellets
of 1 mm in diameter and
1 mm in length.
The β- electron (average
energy of about 315 keV) is
supposed to be absorbed
from the source or the
bushing itself, therefore, each
MC
primary
history
is
composed only by the two
photons.
Fabrizio Cappucci
INFN, Milan.
Relative dose distribution:
Relative dose profiles
16mm X profile
4mm Z profile
8mm Y profile
All within acceptance threshold as derived
from the Report of the American Association of
Physicists in Medicine for stereotactic
radiosurgery
Results:
We have investigated the relative linear dose distribution along the
three coordinated axes. Monte Carlo results have been compared
with standard treatment planning provided by Elekta in the same
homogeneous conditions of the target.
4∙109 primary histories (total calculation time of about 20h on 26
nodes) have been performed for each simulation.
Relative Output Factors (ROF):
ROFs are the ratio between the dose given by a set of collimators and the dose given
by the largest collimators, i.e. the 16 mm.
Collimator
size
Elekta
ROF
FLUKA
ROF
MC
Statistical
Error
∆
8 mm
0.924
0.920
0.88%
0.43%
4 mm
0.805
0.800
0.92%
0.63%
Δ is the percentage difference between the results
from Monte Carlo calculation and the Elekta values:


ELEKTA
OF




1
·
100


FLUKA
OF


Applications in hadrotherapy
Recent Physics Developments





Accurate stopping power calculation with all relevant high order
corrections
Continuous development Interactions of hadrons and nuclei from few
tens MeV/u to several hundreds/MeV/u
Refiniment of nuclear models for de-excitation, production of relevant
isotopes, see P.Sala Varenna 2012, Proc. of int. conf. on Nuclear
Reaction Mechanisms.
All e.m. physics important for gamma imaging: Compton and
annihilation on bound electrons see JINST 2012 JINST 7 P07018
doi:10.1088/1748-0221/7/07/P07018
Other work in progress: interactions of He and light ions…
CERN, 10 Jan 2013
27
Present application of MC calculations
in hadron therapy
• Commissioning of infrastructures
• Commissioning of Treatment Planning
•
•
•
•
System
(TPS in the following) (Heidelberg, CNAO)
Calculation of input physics databases (for
example: the case of TPS developed within the
INFN-IBA collaboration)
Check of TP predictions (and possibly provide
corrections)
Calculation of secondary particle production
Data analysis in dosimetry experiments
Example: use of FLUKA @CNAO to provide input databases
Beam delivery
Required parameters
Scanning with active energy variation
 147 Energy steps (30-320 mm)
 1 Focus size @ ISO
FLUKA calculated FWHM at the isocentre as function of the proton beam energy
S. Molinelli et. al.
Phys. Med. Biol. 58 (2013) 3837
CNAO Med Phys Group
Use of FLUKA @CNAO to provide input databases
FLUKA calculated depth-dose distribution in water
S. Molinelli et. al.
Phys. Med. Biol. 58 (2013) 3837
Dosimetric checks
FLUKA recalculation of a patient PLAN
•
•
Capability to import a CT scan to build 3D voxel geometry
Capability to assigm materials and composition according to HU numbers from CT
scan (now automatic!)
• Capability of coupling to radiobiological models based on Dual
Radiation Approach Theory  Calculation of RBE-weighted dose
(DRBE)
Treatment planning and Monte Carlo
• Currently treatment planning for hadron therapy
are commonly based on fast analytic dose
engines using Pencil Beam algorithms.
• MC calculation of doses and fluences are
superior in accuracy because they take into
account heterogeneities, large densities,
geometry details. They can predict secondary
particle production. However they require much
longer execution times…
Towards a new TPS approach based on MC
• Can we build a TPS using the accuracy
achievable by a detailed MC calculation?
An integrated MC+optimization tool:
• to explore the possibility of a treatment planning
which overcomes the “water-equivalent” approach
• to take into account all details about geometry and
materials
• which can be applied to realistic treatment
conditions with acceptable CPU time
• That can be applied in planning for ions with 1<Z<8:
today’s talk will be focused on protons only
Components and program flow
A 3-port chordoma case
The Syngo TPS
prescription
A. Mairani et al.
PMB 58 (2013) 2471
MC fw simutation of
TPS prescription
Result of our MC
Optimization
QA in hadrontherapy
Use of detection of b+ activity (PET) or of prompt g’s (or charged particle)
produced in the patient. MC is the only possible tool to achieve a reliable
prediction of the observables.
In the literature:
• Potentiality of FLUKA: F. Sommerer et al Phys. Med. Biol. 54 2009
• K. Parodi et al. pioneered the application of PET as a tool to check
hadrontherapy treatments comparing measurements with FLUKA
• Work going on to achieve a true “in-beam” application of the technique to
minimize problems such as metabolic washout
The case for prompt g (nuclear de-excitation)
• Large flux, maybe enough stats for in-beam
• Collimation like Anger camera in SPECT
• Well known technique, robust, compact
• Wide g energy
spectrum 
careful design
• Neutron
background
rejection? TOF
not so easy to
exploit.
• Collimation
reduces stats
Photon yield
GANIL: 90 deg photon yields by 95 MeV/n 12C in PMMA
Blue: Fluka
Red: data
Green: dose profile
Eg> 2 MeV, within
few ns from spill
Z (mm)
[sketch and exp. data taken from F. Le Foulher et al IEEE TNS 57 (2009), E. Testa et al, NIMB 267 (2009)
993. exp. data have been reevaluated in 2012 with substantial corrections]
ENVISION WP6, June 2013
39
Photon yields by 160 MeV p in PMMA: final
Pb
Collimator
NaI detector
PMMA
target
Schematic layout
(dimensions mm)
from J.Smeets et al., IBA
Photon yields by 160 MeV p in PMMA: final
Absolute comparison
Energy spectrum of “photons” after background subtraction (collimator open –
collimator closed) for 160 MeV p on PMMA. FLUKA red line (with exp. resolution
folded in), data black line (J.Smeets et al., IBA, ENVISION WP3)
Univ. Pisa, Roma, Torino and INFN
Project in collab. with CNAO
•
•
Agreement with CNAO to build a Full
in-beam (full-beam) PET system able
to sustain annihilation and prompt
photon rates during the beam
irradiation.
FLUKA strongly used for the design
The FLUKA interface: importing DICOM files
Dicom sets
Slices
Slice Information
Vasilis.Vlachoudis@cern.ch
2D projections
3D
FLUKA interface: superimposing results to CT images
Dose
2 Beams
b+ Activity
1 Beam
g g Emission Map
1 Beam
New tool for PET detector simulation
Generation of PET scanner geometry and
management of FLUKA output and signal reconstruction
Development of new facilities
The TOP-IMPLART Project
Usingf a linac for proton-therapy:
Introduction of the use of FLUKA in shielding calculation (S. Muraro)