Monte Carlo simulation with Geant4 for verification of ...€¦ · 4th Workshop on Geant4...

4th Workshop on Geant4 bio-medical developments and Geant4 physics validation - Genova, July 2005 - Christina Jarlskog 1

Monte Carlo simulation with Geant4 for verification of rotational total skin electron

therapy (TSET)

Christina JarlskogDepartment of Radiation PhysicsLund University, Malmö University Hospital (UMAS)

Collaborators:Lena Wittgren, Sven Bäck, Joakim Medin (UMAS) andErik Traneus (Nucletron)


Rotational total skin irradiation at UMAS

• Varian Clinac 2100 C/D• 6 MeV electrons• treatment SSD = 250 cm• 36 cm x 36 cm field• rotating platform


Aim of this study

• What is the absorbed dose in one rotation...?• measure dose rate in a stationary geometry• use the MC to relate to the treatment geometry

R dose in a stationary geometry for 100 MUdose in one rotation

----------------------------------------------------------------------------------------------------=

RMC100270---------

DMCstationary

DMCrotation

---------------------------=


Simulated geometry

- beam incidence angle wrt phantom x-axis: - the simulation scores energy deposition along x (‘z’)

0° ϑ 360°<≤

x

e


The rotational depth dose distribution

,

DMCrotation D z ϑ,( ) ϑd

ϑ∫=

DMCrotation w

0°D

MC

ϑ 0= ° 2 wϑDMC0° ϑ 90°≤<

ϑ∑ 2 wϑDMC

90° ϑ 180°< <

ϑ∑+ +≅ w

180°D

MC

ϑ 180= °+

DMCrotation w DMC

ϑ 0= ° 2 DMC0° ϑ 90°≤<

ϑ∑+⎝ ⎠

⎛ ⎞≅

w Δϑ360°-----------= Δϑ 5° 10°,=


The simulation

• geometry: homogeneous water phantom

• physics models: standard em and low-energy em

• production thresholds:

- air: 10 keV for photons/electrons- water: 10 keV photons, 100 keV/10 keV

• maximum allowed step length: not set


SteppingAction

• score energy distributions

• grid of 0.5 mm along x-axis

• scoring volume: cylinder of radius r (not built in the geometry)

• check if step is in a water volume

• check if step is within/crossing cylinder

• find energy deposition in scoring volume

• split energy deposition in the grid

• fill energy histogram


PrimaryGeneration: ‘OTP electrons’

• parametrization of a Varian 2100C/D accelerator with 25x25 applicator

• virtual source position: 85 cm upstream the isocenter

• exit phase space plane: 5 cm upstream the isocenter

• field size 36 cm x 36 cm (yz plane)

• y and z coordinates from flat distribution

• direction cosines corrected for scattering in accelerator

• correction independent of radial displacement

• only direct electrons simulated

• energy sampled from OTP spectrum


OTP energy spectrum

• extracted for PDD measured at SSD = 100 cm:

• precalculated with VMC++

• is the kinetic energy of the primary electrons

• assuming an energy spectrum with unknown constants

• PDD is given by:

• the unknown constants in are calculated by fitting

• by

Dmeas z( )

Dmono E z,( )

E

F E( )

D z( ) Dγ z( ) F E( )Dmono E z,( ) Ed∫+=

F E( )

Dmeas z( ) D z( )


Fitted PDD

the distribution is fitted for z > 4 mm to neglect low-energy electrons that deposit energy close to the surface of the phantom

0

20

40

60

80

100

0 1 2 3 4

depth (cm)

%D

measured

VMC++


Kinetic energy of primary electrons

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10

VMC++

Geant4

E (MeV)

F(E

)


Overview of the simulations (uniform thresholds)

Stationary geometries (r = 2 cm, cubic phantom 50 cm side):• SSD = 100 cm, standard em physics• SSD = 100 cm, low-energy em physics• SSD = 250 cm, low-energy em physics

Rotational geometries (SSD = 250 cm, low-energy em physics):

• test runs: r = ?, ?, N = ? if

ellipsoid phantom 35 cm in x, 50 cm in z, 25 cm in y - run 1: r = 2 cm, (10 M events/angle) - run 2: r = 0.5 cm, (100 M events/angle) - run 3: r = 0.5 cm, (100 M events/angle)

• ellipsoid phantom 20 cm in x, 40 cm in z, 25 cm in y• cylindrical phantom 30 cm diameter, 25 cm height

Δϑ = σRMC2%≈

Δϑ 10°=Δϑ 10°=Δϑ 5°=


Comparison MC/measurement at SSD=100 cm

0

0.01

0.02

0.03

0 10 20 30 40

measured

low energystandard

depth (mm)

D (

Gy/

MU

)


Comparison MC/measurement at SSD=250 cm

0

0.001

0.002

0.003

0.004

0.005

0 10 20 30 40

depth (mm)

D (

Gy/

MU

)

measured

low-energy


Results of test run 1 (r = 2 cm, )Δϑ 10°=

0

0.2

0.4

0.6

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)

stationary

10 degrees

20 degrees

30 degrees

0

0.2

0.4

0.6

x 10-13

0 10 20 30

depth (mm)D

(G

y/ev

ent)

40 degrees

50 degrees

60 degrees


Results of test run 1 (r = 2 cm, )

is the surface dose too high at large beam angles?

Δϑ 10°=

0

0.2

0.4

0.6

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)

70 degrees

80 degrees

90 degrees



0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)



0

1

2

3

0 10 20

depth (mm)

R

0

2

4

6

0 10 20

depth (mm)σ st

at (

%)

Rstationary depth doserotational depth dose


Test runs: difference in stationary depth dose

0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)

Dst

at (

Gy/

even

t)

2 cm radius

0.5 cm radius

-0.2

0

0.2

x 10-14

0 10 20 30 40 50

depth (mm)ΔD

stat

(G

y/ev

ent)


Test runs: difference in rotational depth dose

0

0.05

0.1

0.15

0.2

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)

r = 2 cm, Δθ = 10o

r = 0.5 cm, Δθ = 10o

r = 0.5 cm, Δθ = 5o

-0.05

0

0.05

0.1

x 10-14

0 10 20 30 40 50

run 1 - run 2

run 2 - run 3

depth (mm)ΔD

rot (

Gy/

even

t)


Test runs: difference in R

0

0.5

1

1.5

2

2.5

0 10 20

depth (mm)

R

r = 2 cm, Δθ = 10o

r = 0.5 cm, Δθ = 10o

r = 0.5 cm, Δθ = 5o

-0.2

-0.1

0

0.1

0 10 20

depth (mm)R

(run

1)

- R

(run

3)


Simulation parameters

Subsequent simulations were run with the following parameters:

- a scoring volume radius of 2 cm

- a step in angle of beam incidence

- 10 million events per angle for

- 40 million events for

Δϑ 10°=

ϑ 0°>

ϑ 0°=


Dependence on phantom cross-section

0

0.05

0.1

0.15

0.2

x 10-13

0 10 20 30 40

depth (mm)

Dro

t (G

y/ev

ent)

cylinder 30 cm

ellipse 20 cm x 40 cm


0.5

1

1.5

2

2.5

0 10 20

depth (mm)R



cylinder 30 cm


Thresholds by region for the cylindrical geometry

• inner phantom with 10 keV threshold for electrons:- outer radius: 15 cm, inner radius: 11 cm

- Δφ 30°=


Threshold dependence at SSD = 250 cm

0

0.1

0.2

0.3

0.4

x 10-13

0 10 20 30

depth (mm)

Dst

at (

Gy/

even

t)

10 keV

100 keV

0

0.05

0.1

0.15

0.2

x 10-13

0 10 20 30

10 keV

100 keV

depth (mm)D

rot (

Gy/

even

t)



0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)

θ = 10o

10 keV

100 keV

0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)D

(G

y/ev

ent)

θ = 20o

10 keV

100 keV



0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)

θ = 30o

10 keV

100 keV

0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)D

(G

y/ev

ent)

θ = 40o

10 keV

100 keV



0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20 30

depth (mm)

D (

Gy/

even

t)

θ = 50o

10 keV

100 keV

0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 10 20

depth (mm)D

(G

y/ev

ent)

θ = 60o

10 keV

100 keV



0

0.2

0.4

0.6

x 10-13

0 5 10 15 20

depth (mm)

D (

Gy/

even

t)

θ = 70o

10 keV100 keV

0

0.1

0.2

0.3

0.4

0.5

x 10-13

0 2.5 5 7.5 10

depth (mm)D

(G

y/ev

ent)

θ = 80o

10 keV

100 keV

depth (mm)D

(G

y/ev

ent)

θ = 90o


Ratio Dstat/Drot at SSD = 250 cm

Absorbed dose in one rotation at

the prescription depth (5 mm):

fraction dose = 2 Gy

- without degrader:

,

- with degrader:

, 0.5

1

1.5

2

2.5

0 10 20

depth (mm)

R

10 keV

100 keV

R 0.88=

Drot 0.42 Gy= Nrot 5=

Drot 0.23 Gy= Nrot 9=



0

0.01

0.02

0.03

0 10 20 30 40

measured100 keV10 keV

depth (mm)

D (

Gy/

MU

)



0

25

50

75

100

0 10 20 30 40

VMC++

100 keV

depth (mm)

%D

0

25

50

75

100

0 10 20 30 40

VMC++

10 keV

depth (mm)%

D


Conclusions

• the low-energy em physics model gives a good agreement both with measurements and with VMC++• the simulation has a dependence on the electron pro-duction threshold (when no limit on step length is set)• R can be calculated for 100 keV electron threshold• step length distribution? • dependence on beam angle?• set maximum step length?• ...

Date post:	09-Aug-2020
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

Monte Carlo simulation with Geant4 for verification of ...€¦ · 4th Workshop on Geant4...

Documents