Room: ePoster Forums
Purpose: To assess performance and differences in dose distributions calculated with four algorithms in spine SBRT
Methods: 16 were calculated in Elements Spine SRS1.5 (Brainlab) with Monte Carlo dose to medium (MC_Dm, 2mm-2%), taken as gold standard. Clinical protocol included 36Gy, three fractions, 90% coverage of PTV. Plans recalculated with invariable configuration using Pencil Beam (PB) and Monte Carlo dose to water (MC_Dw); exported to Eclipse 15.1 (Varian) and computed with identical conditions utilizing AAA, Acuros XB (AXB_Dw and AXB_Dm). Plans compared via 3D-Î³ analysis (1mm-1%). D95%, D98%, Mean and D2% of the PTV; and Dmax, D0.35cc, D1.2cc to spinal cord (SC) were evaluated
Results: MC_Dw values were in average 3.73% for PTV and 0.89% for SC compared to MC_Dm; similarly, for AXB_Dw vs AXB_Dm, 3.34% for PTV and 1.26% for SC. 3D-Î³ analysis considering voxels with at least 30% of prescribed dose showed 96.04% agreement with AXB, 92.63% with AAA, and 95.43% with PB (rising threshold to 85% of the dose, 84.43% with AXB, 77.49% with AAA and 83.05% with PB). Average differences of AXB_Dm vs MC_Dm were 0.27% for PTV and -13.87% for SC. Comparing AAA with MC_Dm, an average difference -0.97% for PTV and -10.31% for SC; whereas analysis vs MC_Dw yielded -4.52% for PTV and -11.09% for SC. Comparing AAA with AXB_Dm, -1.23% for PTV and 4.28% for SC, whilst assessments vs AXB_Dw yielded -4.41% for PTV and 2.99% for SC. Analysis of PB produced comparable results
Conclusion: Dw overestimates doses with respect to Dm. Agreement between AXB_Dm, AAA and PB vs MC_Dm to the PTV is shown. However, greater differences found for SC, mainly closer to interface (-6.24% for PB, -10.31% for AAA and -13.87% for AXB_Dm). Type A and B Dw is better approximated by type C Dm, without applying stopping powers-based conversion factors
Not Applicable / None Entered.