Proton Lung Planning Based On Beam Specific PTV
S Flampouri*, B Hoppe, Z Li, University of Florida, Jacksonville, FLTH-C-144-7 Thursday 10:30AM - 12:30PM Room: 144
To investigate the optimal use of beam specific PTV (bsPTV) in passively scattered proton lung treatments.
An algorithm is developed and implemented to calculate bsPTV for lung cancer targets. The uncertainties in proton treatments, including setup errors, machine delivery variability, breathing motion, proton range uncertainties and their combinations are taken into account. The effect of breathing motion on the proton range is calculated from 4DCT. Margins can be added for translational and rotational setup errors and breathing motion changes during the course of treatment as well as for their effect on beam range of each treatment field. It is notable that the range uncertainties are calculated not as a function of the total range but taking into account the individual voxel HU uncertainty along a proton beamlet. bsPTVs are calculated for beams delivered on actual lung proton treatments. Plans based on bsPTVs (bsPlan)are generated and compared to conventional proton plans (origPlan). A range of TPS parameters and normalization methods are investigated.
All bsPlans increase the treated volume compared to origPlans. When the origPlans and bsPlans are created to compensate for similar uncertainties and are normalized the same way, dose distributions produced are similar (Gamma index>1 to 7-12% of voxels for a 3 field plan). When the bsPlans are normalized per field the maximum dose as well as the irradiated volume increase.
A tool to calculate beam specific PTVs for proton treatments is developed to include the uncertainties involved. The volumes created can be used for beam angle optimization (smaller bsPTV corresponds to smaller uncertainty). bsPTVs can be the plan targets without the need of extra margins or they can be used for conventional proton plan evaluation (compared to produced dose distribution).
Add this talk to vcal | ical | Contact Email: