Beam Properties of An In-Room Proton Therapy Accelerator
C Bloch*, P Hill, E Klein, Washington University School of Medicine, SAINT LOUIS, MOSU-E-T-471 Sunday 3:00:00 PM - 6:00:00 PM Room: Exhibit Hall
Purpose: To quantify basic beam properties of the first gantry-mounted proton therapy accelerator, situated in the treatment room and delivering beam to the patient without any intermediate bending magnets.
Methods: Monte Carlo simulations for prototype beamlines of the Mevion S250TM proton therapy system were performed using MCNPX. For each configuration, open profiles in air were determined at 3 different positions for 3 range settings (216 profiles). Similarly, half-beam-blocked profiles were determined for the same settings (216 profiles). Additionally, beam axis fluence profiles in air were determined for 3 range settings (72 profiles). Finally, Bragg peak depth-dose curves in water were determined for 3 range settings (72 profiles). Using these 576 profiles, several beamline parameters were determined for each proton range and configuration, including virtual SAD, effective source size, and effective SAD. Additionally, Bortfeld's analytic approximation of the Bragg curve was fit to each depth-dose curve to determine R0, s and e. The 72 values determined for each parameter were compared within each configuration and across all configurations.
Results: Within any configuration, the change in range was too small (=2.5 cm) to have a large effect on any of the extracted parameters. Comparing parameters among configurations revealed smooth trends. These parameters were then fit as a function of range.
Conclusions: The simplicity of this system results in beams with greater similarity among configurations as compared to other proton therapy systems. Because a beam with fixed initial energy is degraded only by tissue-equivalent materials, range straggling and beam divergence are nearly constant across all beamline configurations. Beam parameters change smoothly as functions of proton range, with no major discontinuities between configurations. This suggests that the entire machine may be fully and accurately characterized in a treatment planning system by fewer measurements than would normally be required.