Program Information

Material Interface Behavior of the Eclipse Modulated Spot Scanning Proton Algorithm

D Mundy*, C Beltran, M Herman, Mayo Clinic, Rochester, MN

SU-C-500-7 Sunday 1:00PM - 1:55PM Room: 500 Ballroom

Purpose: Evaluate the accuracy of the Eclipse proton dose calculation at high and low density material interfaces.

Methods: A phantom designed to test the Eclipse algorithm accuracy at material interfaces was developed. The phantom is a 40-cm water cube with two slab heterogeneities (dense bone and air). The materials abut at the central-axis and dose from a single 144.7-MeV beam spot along and spanning the interface was calculated. Depth and thickness of the slabs were varied to examine effects on beam termination and penetration of the bone. The dose distributions were compared using a variety of metrics.

Results: The most notable observed difference was the presence of three Bragg peaks in Eclipse versus two in Geant4. The proximal peak corresponds to the range associated with the presence of the bone and the distal to that of the air. Eclipse calculates an intermediate peak of substantial intensity. The depth of this peak scales linearly with the thickness of the slabs, but is independent of the slab depth. Intensity of the peak decreases with increasing thickness, but is constant with depth. The energy loss associated with this peak causes a decrease (~45%) in the intensity of the distal peak. The distal peak locations agree to within 2-mm, while the proximal peaks are deeper in Eclipse by ~5-mm for all shallow slabs and 1.6-mm for the deep slab. Intensity of the proximal peaks differed by 2-5% with the greatest difference observed with the deep slab.

Conclusions: Large discrepancies were identified in the Eclipse spot scanning proton dose calculation algorithm relative to Monte Carlo calculations in the presence of a material interface. These differences are much greater than those observed in a homogenous phantom. Further investigation is necessary to determine the impact of these discrepancies on patient dose calculations under clinical conditions.

Contact Email: