1. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 1 Results from the recent carbon test beam at HIMAC Koichi Murakami Statoru Kameoka KEK CRC supported by

2. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 2 Introduction A joint project among Geant4 developers, astro- physicists and medical physicists in Japan Development of software framework for simulation in radiotherapy ≫ funded by the Core Research for Evolutional Science and Technology (CREST) program organized by Japan Science and Technology Agency (JST) from 2003 to 2008 The project goal provides a set of software components for simulation in radiotherapy (especially hadrontherapy), ≫ well designed general purpose software framework ≫ DICOM/DICOM-RT interface ≫ application of GRID computing technology ≫ visualization tools In addition, physics validation is one of key issues.

3. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 3 Physics Validation in Radiotherapy Geant4 has to reproduce precise dose distributions in human body. which requires correct simulation for the interactions between various types of beams (X-ray, proton, heavy ions) and materials along beam line reliable descriptions of ≫ electromagnetic processes ≫ hadronic/nuclear processes ≫ nuclear decay processes in the relevant energy regions and particle types. These are non-trivial issues! Physics validation is one of the most critical aspects in the project.

4. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 4 Hadrontherapy Facilities in Japan The Energy Research Center Wakasa Bay (Tsuruga: 200 MeV) Hyogo Ion Beam Medical Center (Nishi-Harima: 320 MeV/u) Shizuoka Cancer Center (Mishima: 230 MeV) NIRS (Chiba: 90 MeV, 400MeV/u) NCC East Hospital (Kashiwa: 235 MeV) U. of Tsukuba PMRC (Tsukuba: 250 MeV) Ion beam Proton beam Jpn (world) # Proton beam facilities:5 (23) # Ion beam facilities: 2 (4)

5. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 5 HIMAC at NIRS Operation since 1994 Treatment beam: 12 C Over 2,000 patients have been treated Experiment Areas RFQ Linac 800 KeV/u Alvarez Linac 6 MeV/u Synchrotron 800 MeV/u Treatment Rooms Ion Source ~65 m

6. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 6 Hadron (proton/carbon) Beam Hadron beams allow conformation of dose distribution better than photons and electrons; Ref. http://www.nirs.go.jp/tiryo/himac/himac2.htm proton carbon Ｘ -ray  -ray neutron Relative Dose (%) 50 100 5.0 10.0 15.0 0.0 Depth - Human Body (cm) A sharp peak of energy deposition at the end of the range (Bragg peak) The sharp fall-off of the Bragg peak for carbon beam A small range straggling Carbon produces a longer tail after the Bragg peak.

7. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 7 Conformation of Irradiation Field Patient body Wobbler magnets Y X Ridge Filter Scatterer Range Shifter Collimator Compensator (Bolus) Target volume (tumor) Bragg peak Spread-out Bragg peak (SOBP) Depth dose Beam Ridge Filter B y = A y sin(  t) B x = A x sin(  t+  /2) Spiral beam divergence to create a uniform irradiation field

8. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 8 Experimental Setup Treatment position (isocenter) Vacuum window Water target Acrylic vessel Test beam line of HIMAC(NIRS) Secondary emission monitor Wobber magnets XY Scatterer (lead) Dose Monitor (ionization Chamber) Collimator Ridge filter (aluminum) Range shifter (unused) Multi-leaf Collimator (open) Collimator Beam profile Monitor (ionization Chamber) Beam 12 C Beam Energy 290, 400 MeV/u

9. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 9 Water target / Scored region Dose distribution in a water target was measured using the horizontal arrayed dosimeters voxel size of each element is 2 x 2 x 1 mm. scanning along the depth direction 400 mm 2 mm 1 mm 2 mm Water target Beam ( 12 C) Scored region

10. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 10 Physics List Generic Ions elastic scattering Binary light ion cascade or JQMD ≫ cross section : Tripathi / Shen radioactive decay ionization / multiple scattering Hadron elastic scattering L(H)EP+Binary cascade ionization / multiple scattering electron/gamma standard EM

11. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 11 Bragg Peak Simulation (Binary Cascade) 290MeV/u 40oMeV/u Overall profile of Bragg peak seems to be well reproduced, but… We found a small bump just before the peak… What is this!?

12. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 12 Bragg Peak – more in detail BCJQMD Secondaries of 11 C produce the bump of BC. JQMD shows no bump. Production rates of 11 C (one neutron stripped off) and 11 B (one proton stripped off) are different between Binary Cascade and JQMD. Production rate of 11 C in BC is over created.

13. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 13 Comparison between Experiment and Simulation (290 MeV/u) Bragg Peak SOBP (Spread-Out Bragg Peak) w/ Ridge Filter offset=-0.8mm offset=-1mm tends to underestimate the tail effect coming from beam fragments

14. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 14 Comparison between Experiment and Simulation (400 MeV/u) offset=-1.2mm offset=-2.8mm Bragg Peak SOBP w/ Ridge Filter tends to underestimate the tail effect coming from beam fragments slight inconsistency in offset values?

15. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 15 Tail Effect – more in detail Bragg Peak 290MeV/u40oMeV/u SOBP Tail effect is underestimated by 10-20%. Binary Cascade

16. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 16 Summary A joint project among Geant4 developers and medical physicists in Japan is on-going. Physics validation in medical application (particle therapy) is a critical issue. A new test beam line in HIMAC was constructed, and experimental data was obtained. It is a good chance to validate Geant4 ion physics. Geometry of the test beam line was implemented in Geant4, and comparisons with simulation were carried out. We tried the Binary Cascade model and the JQMD model for describing ion interactions. Overall profile of the Bragg peaks are well reproduced by Geant4 simulation. … but, we found a problem with the Binary Cascade model in our problem domain. We hope that it will be improved. The tail effect coming from ion fragments is not fully reproduced. Geant4 tends to underestimate the effect. There are some space to be improved.

17. Koichi MurakamiGeant4 Physics Verification and Validation (17-19/Jul./2006) 17 Acknowledgements T.Sasaki, K.Amako, G.Iwai (KEK) T.Aso (TNCMT) A.Kimura (Ashikaga Univ.) T. Koi (SLAC) M.Komori, T.Kanai, N.Kanematsu, Y.Kobayashi, S.Yonai (NIRS), Y.Kusano, T.Nakajima, O.Takahashi (AEC) M.Tashiro (Gunma Univ.) Y.Ihara, H.Koikegami (IHI) supported by