Experimental depth dose curves of a 67.5 MeV proton beam for benchmarking and validation of Monte Carlo simulation

Abstract

Purpose: To measure depth dose curves for a 67.5 ± 0.1 MeV proton beam for benchmarking and validation of Monte Carlo simulation.

Methods: Depth dose curves were measured in 2 beam lines. Protons in the raw beam line traversed a Ta scattering foil, 0.1016 or 0.381 mm thick, a secondary emission monitor comprised of thin Al foils, and a thin Kapton exit window. The beam energy and peak width and the composition and density of material traversed by the beam were known with sufficient accuracy to permit benchmark quality measurements. Diodes for charged particle dosimetry from two different manufacturers were used to scan the depth dose curves with 0.003 mm depth reproducibility in a water tank placed 300 mm from the exit window. Depth in water was determined with an uncertainty of 0.15 mm, including the uncertainty in the water equivalent depth of the sensitive volume of the detector. Parallel-plate chambers were used to verify the accuracy of the shape of the Bragg peak and the peak-to-plateau ratio measured with the diodes. The uncertainty in the measured peak-to-plateau ratio was 4%. Depth dose curves were also measured with a diode for a Bragg curve and treatment beam spread out Bragg peak (SOBP) on the beam line used for eye treatment. The measurements were compared to Monte Carlo simulation done with geant4 using topas.

Results: The 80% dose at the distal side of the Bragg peak for the thinner foil was at 37.47 ± 0.11 mm (average of measurement with diodes from two different manufacturers), compared to the simulated value of 37.20 mm. The 80% dose for the thicker foil was at 35.08 ± 0.15 mm, compared to the simulated value of 34.90 mm. The measured peak-to-plateau ratio was within one standard deviation experimental uncertainty of the simulated result for the thinnest foil and two standard deviations for the thickest foil. It was necessary to include the collimation in the simulation, which had a more pronounced effect on the peak-to-plateau ratio for the thicker foil. The treatment beam, being unfocussed, had a broader Bragg peak than the raw beam. A 1.3 ± 0.1 MeV FWHM peak width in the energy distribution was used in the simulation to match the Bragg peak width. An additional 1.3-2.24 mm of water in the water column was required over the nominal values to match the measured depth penetration.

Conclusions: The proton Bragg curve measured for the 0.1016 mm thick Ta foil provided the most accurate benchmark, having a low contribution of proton scatter from upstream of the water tank. The accuracy was 0.15% in measured beam energy and 0.3% in measured depth penetration at the Bragg peak. The depth of the distal edge of the Bragg peak in the simulation fell short of measurement, suggesting that the mean ionization potential of water is 2-5 eV higher than the 78 eV used in the stopping power calculation for the simulation. The eye treatment beam line depth dose curves provide validation of Monte Carlo simulation of a Bragg curve and SOBP with 4%/2 mm accuracy.

FIG. 1.

Raw beam line used for benchmark measurements showing the Ta scattering foil (TA), the four sequentially placed collimating elements of the carbon collimator (C1), beam plug (PL), beam pipe (P), and second collimator (C2), the secondary emission monitor (S) enclosed in an evacuated box (B), the exit window (K) with the third collimator (C3), where the beam passes out of vacuum into air, and the Mylar window (MY) with water phantom (WP). See Table I in the Appendix for detailed geometry. A few simulated proton tracks are shown.

FIG. 2.

Crocker Lab beam line used for eye treatment, showing the wire chambers (WC1 and WC2), secondary emission monitor (SEM), exit window (E), collimators (C1, C2, and C3), ion chambers (IC1 and IC2), range modulator wheel (P), water column (H2O), mirror (M), patient shield (PS), patient assembly (PA), Mylar window (MY), and water phantom (WP). See Table III in the Appendix for detailed geometry.

FIG. 3.

Setup for positioning diode at the surface of the Mylar wall.

FIG. 4.

Percent depth dose curves including dose to water and dose to a Bragg–Gray cavity in water, simulated with topas for a 67.5 MeV proton beam with a 0.4 MeV FWHM peak transported through the raw beam line and incident on the thinner Ta foil. Dose to water compared to dose to Bragg–Gray cavities of silicon and air in water from the same simulation, normalized at the Bragg peak to show the effect of the SPR on the peak-to-plateau ratio. Inset shows the SPR calculated from these curves as the ratio of the water dose to the Bragg–Gray cavity dose (solid lines), compared to the SPR calculated with gamos (squares).

FIG. 5.

Bragg curves measured with PTW diode and Markus and Roos parallel-plate chambers (points) on the raw beam line to establish the accuracy of the diode for benchmark measurements of the shape of the Bragg curve. The dose scale is magnified on the left hand side plot to better show the plateau region, with points joined by straight lines. The depth scale is magnified on the right hand side plot to better show the peak region. The ion chamber measurements were shifted to match the diode measurement at the distal edge of the Bragg peak. The ion chamber measurements on the left are normalized at 10 mm depth to match the diode measurement at that depth and the curves on the right are normalized to 100% at the Bragg peak, using the analytical function fit to the points. Depth accuracy of the diode was improved in subsequent measurements.

FIG. 6.

The raw beam Bragg curves measured with PTW and EFD Si diodes (points). The inset is a magnified view of the Bragg peak, with depth accuracy of ±0.15 mm shown on the points closest to 50% of the maximum dose. Monte Carlo Bragg curves (lines) for 67.5, 0.4 MeV FWHM, 5 × 10 mm² protons simulated with topas. Additional simulations are shown without collimation for both foils (dotted lines) and for zero spot size for the thicker foil (dashed line).

FIG. 7.

The effect of changing the production cut in the beam plug on the simulation of the raw beam Bragg curve. Normalized to 100% at the Bragg peak to show the effect of the production cut on the peak-to-plateau ratio.

FIG. 8.

Proton depth ionization curves for eye treatment beam line measured with the PTW diode: Bragg peak and 20 mm SOBP with a nominal 10 mm water column. Curves are normalized to 100% at the Bragg peak to show the effect of the peak width on the peak-to-plateau ratio. Simulations had a thicker water column than the nominal thickness. Bragg curve simulated with different widths of the peak in the energy distribution. SOBP simulated with the measured angles subtended by the RMW blades and a 1° wider angle used for the base plate (both 1.3 MeV peak FWHM).

FIG. 9.

Photograph of range modulator wheel with schematic of a plate in one of the sectors.