An approach for estimating dosimetric uncertainties in deformable dose accumulation in pencil beam scanning proton therapy for lung cancer

Abstract Deformable image registration (DIR) is an important component for dose accumulation and associated clinical outcome evaluation in radiotherapy. However, the resulting deformation vector field (DVF) subject to unavoidable discrepancies when different algorithms are applied, leading dosimetric uncertainties of accumulated dose. We propose here approach proton therapy estimate as a consequence modeled or estimated DVF uncertainties. A patient-specific uncertainty model was built on first treatment fraction, by correlating magnitude differences five DIR results at each voxel any single reference DIR. In following fractions, only needs be geometric were this model. The then derived considering uncertainty, gradient fractional recalculated distribution direction factor from applied fraction. This respectively compared DIRs individually warping. validated seven NSCLC patients, with nine repeated CTs. proposed model-based method able achieve conservative voxel-to-voxel comparison within ±5% prescribed ‘reference’ 77% voxels body 66%–98% investigated structures. induced lung tumor treatments.