Room: Exhibit Hall
Purpose: To determine the optimal smoothed compensator smearing value using a clinical treatment planning system for a given setup margin and proton range.
Methods: We created a defined geometry in a computerized water phantom with different heterogeneities by overriding the structure to different Hounsfield Units. The Plan Uncertainty Tool within Eclipse version 13.7 was utilized to analyze the dose coverage following lateral shifts along the direction of the density gradient. The shifts were made in such a way that the denser material would overlap and a distal cold spot is created within the target if the smear value is insufficient. We calculated the minimum dose of the shifted plan at the distal end of the target in order to determine the sufficient smear value. According to our clinical practice we post-processed all calculated compensators using the manual editing tool to reduce hotspots in the nominal plan.
Results: We found that the required smear values correlate well with the setup margin and the minimum dose increases with increased smearing value. However, the increase is not strictly monotonic and there are steps with 2mm increments. For deep targets almost full coverage is achieved independently of the smear value. However, for shallow targets a smear value of the setup margin (3mm) plus 1mm is needed to achieve at least a 70% minimum dose. Similar results were found for high-HU and low-HU overrides. Beam divergence also has a negligible effect on distal target coverage.
Conclusion: Longer proton ranges require lower compensator smear values compared to shorter ones. This is in contradiction to data published in the literature. To have sufficient target coverage for all ranges a smear value of the setup margin plus 1mm is needed.