Geant4 Workshop 2011 Wollongong, Australiaweb/@eng/@phys/... · Centre for Medical Radiation Physics, University of Wollongong, Australia. Geant4 Workshop 2011. Wollongong, Australia

Ashley Cullen, Michael Lerch, Marco Petasecca, Susanna Guatelli, Anatoly RosenfeldCentre for Medical Radiation Physics, University of Wollongong, Australia

Geant4 Workshop 2011Wollongong, Australia

Introduction to MRT

Dosimetry for MRT

The Silicon Microstrip Detector

Energy Dependence Problem

Monte Carlo Simulations

Conclusions

In the early days of radiotherapy Co-60 and kilovoltage x-ray units were the dominant radiotherapy modality.

Superficial doses were high, causing excessive skin burn

Discovery that iron grids pressed hard against the skin caused regions shielded from radiation healed adjacent burns

In 1960’s, experiments investigate the effect of 22.5 MeV 25μm deuteron microbeams on mice cerebrum

◦ Doses in excess of 3,000 Gy caused little damage

◦ Range of 1.5mm in tissue was of little clinical use

Microtomography experiments conducted with 30μm synchrotron pencil beams on mouse heads had little contrast at 10 Gy, so dose was increased to 200 Gy.

◦ A month later, a histopathological search failed to find any radiation damage.

A study investigated the effect of 25μm-wide planar 50-150 keV synchrotron x-rays on murine brains

◦ Thousands of Gy were survived

◦ Microbeam radiation therapy (MRT) was born

MRT under investigation as an experimental treatment for untreatable paediatric brain tumours.

A “white” synchrotron beam is used (ID17 @ ESRF, Grenoble)

A striated planar array of beams tens-of μmwide and hundreds-of μm centre-to-centre

Peak-to-valley dose ratio very important parameter

[1] E.A Siegbhan, et al., Dosimetry for synchrotron x-ray microbeam radiation therapy, ESRF[2] E. Braüer-Krisch, et al., Effects of pulsed, spatially fractionated, microscopic synchrotron X-ray beamson normal and tumoral brain tissue

Dosimetry for MRT is particularly challenging:◦ Very high dose rates (~20,000 Gy/s)

Varies with decrease of storage ring current

◦ Very short treatment times (~30ms)

◦ Very small beam sizes (50μm)

◦ Radiobiological importance of radiation field structure

Ideal dosimetry system requirements:◦ High radiation hardness

◦ High temporal resolution

◦ High spatial resolution

◦ Real-time readout

Several methods of dosimetry for MRT are in current use:◦ Ion chambers

Real time measurement

No spatial information

◦ High Dose Radiochromic film

High spatial resolution

Dose range from peak to valley are outside dynamic range of film

Requires two separate irradiations to obtain measurements of both

Requires post-irradiation digitisation

Experimental dosimetry methods:◦ MOSFETs Provides high-spatial resolution when scanned

through beam Not radiation hard to intense synchrotron radiation◦ Fibre-optic◦ OSL

Epitaxial construction◦ 26 μm thick epitaxial layer◦ On 370 μm Si substrate

Sensitive volume is 20 µm × 500 μm

Surrounded by a guard-ring

10 100 10000

1

2

3

4

5

6

7

8

Photon Energy (keV)

γ Mas

s En

ergy

Abs

. Coe

ff. R

atio

(Si/W

ater

)

0.0

0.2

0.4

0.6

0.8

1.0

ID17

Nor

mal

ised

Phot

on S

pect

rum

Monochromatic synchrotron x-rays

ID17 @ ESRF, Grenoble Produced by a double-crystal

monochromator Dose to microstrip compared

to tissue-eq. ionisationchamber results

Displays low energy over-response

0 10000 20000 300001E-5

1E-4

1E-3

0.01

0.1

1

Norm

alise

d In

tens

ity

Distance (arb units)

Left Right

The radiation transport simulation toolkit Geant4 (v9.3) was used for all Monte Carlo simulations described herein

The Low Energy Electromagnetic Physics Package was used◦ Provides accurate tracking of photons to energies of 250 eV

◦ Agrees with NIST data Amako, et al., Comparison of Geant4 Electromagnetic Physics Models Against the NIST Reference

Data, IEEE Trans. Nuc. Sci., Vol. 52, No. 4, August 2005

To improve simulation speed, beamline components were not modeled

Microbeams were assumed to have a homogeneous energy and fluence distribution

Polarisation not modeled◦ see G. Takacs, Photon polarisation, magnetic fields, and electron trajectories: the

irrelevance of the electron mean free path

Detector at depth 5 mm in 10x10x10 cm3

phantom. Monochromatic photons

fired upon phantom central on detector.

Dose scored in sensitive volume of epitaxial layer

Repeated for water with same scoring geometrical boundaries

40 50 60 70 801

2

3

4

5

6

7

Ratios (Si/Water) MC Dose Ratio - 26 µm Epi-Layer MC Dose Ratio - 6 µm Epi-Layer Mass-Energy Absroption Ratio Electron Stopping Power

Ratio

(Si/W

ater

)

Energy (keV)

Microstrip placed at various depths in water phantom

Irradiated with the MRT spectrum in a homogeneous field

Dose to microstrip and water in same geometric boundaries scored

Water dose equivalence obtained from ratio

0 20 40 60 802

4

6

8

10

12

14

16

18

Dose

per

Prim

ary

Phot

on x

10-1

1 (Gy/

n)

Depth in Phantom (mm)

Microstrip Detector Water

0 20 40 60 80

1.45

1.50

1.55

1.60

1.65

1.70

1.75

Wat

er E

quiva

lenc

e Ra

tio (D

Si/D

w)Depth (mm)

A simulation was produced to determine the photon spectrum outside of microbeams

An infinitesimally thin pencil beam was fired into a 10×10×10 cm3 water phantom normal to a central surface

The spectrum of photons entering a 5×5×5 mm3

sensitive volume were binned in 1 keV increments.

Performed at various positions

1 10 1001

10

100

1000

10000

100000

Cou

nts

Photon Energy (keV)

Distance fromBeam Centre

0 mm 1 mm 5 mm 20 mm

20 25 30 35 40 45 501

10

100

1000

10000

100000

Counts

Photon Energy (keV)

Distance fromBeam Centre

0 mm 1 mm 5 mm 20 mm

Previous simulation repeated with monochromatic photons

To isolate which part of the input spectrum is most responsible

Energies:◦ 40 keV

◦ 80 keV

◦ 100 keV

◦ 400 keV

0 10 20 30 40 50 60 70 801

10

100

1000

10000

100000

1000000

1E7

Coun

ts

Photon Energy (keV)

Lateral Pos. 0 mm 1 mm 5 mm 20 mm

0 100 200 300 4001

10

100

1000

10000

100000

1000000

1E7

Coun

ts

Photon Energy (keV)

Lateral Pos. 0 mm 1 mm 5 mm 20 mm

0 10 20 30 Comptonmin.

401

10

100

1000

10000

100000

1000000

1E7

Coun

ts

Photon Energy (keV)

Lateral Pos. 0 mm 1 mm 5 mm 20 mm

0 20 40 60 80 1001

10

100

1000

10000

100000

1000000

1E7

Coun

ts

Photon Energy (keV)

Lateral Pos. 0 mm 1 mm 5 mm 20 mm

40 keV

400 keV

80 keV

100 keV

The photon spectrum out-of-field due to being as a result of Compton scatter

A 25-35 keV low-energy photon component is present which does not attenuate with distance◦ This is the region Si is most sensitive to

◦ Produced by Compton scattering of previously scattered photons.

80 – 100 keV primary photons most responsible◦ Requires more investigation

A new detector has been developed to reduce the low energy over response.

Silicon-on-Insulator construction

◦ 7μm Si detector on 2μm SiO2 insulator

◦ Constructed on 370μm Si substrate

Ion-beam-induced charge-collection studies have been performed, but yet to be analysed.

Will be tested with a synchrotron MRT field at ESRF, France in October

Microstrip detector is over sensitive to low energy photons

Response will vary with spectrum

◦ i.e. depth, in-field/out-of field

The use of Geant4 has been used to design a less energy-dependent detector

◦ Has yet to be experimentally tested