摘要：Abstract Purpose Due to inter-fraction variation in applicator position, organ displacement and deformation, doses to targets and normal tissues may not be accurately represented by adding the doses from external beam radiation therapy (EBRT) and intracavitary brachytherapy (ICBT) using rigid image registration. Deformable image registration permits organ and applicators to be spatially matched in 3D, enabling more accurate tracking of the accumulated volumetric dose to the target as well as organs at risk (OAR). This study assesses the dosimetric impact of using deformable image registration to determine the cumulative EBRT and ICBT doses to the rectum and bladder. Methods and materials Data from 20 patients with stage IB1-IVA cervical cancer were analyzed. Nine of the patients were treated with ICBT and EBRT which included a nodal or parametrium boost while eleven were treated with ICBT and EBRT with no boost. Dose summation was performed in two stages. For the first stage, only the ICBT fractional doses were added using both “parameter adding” and deformable registration techniques. In the second stage, the ICBT and EBRT doses were combined using “parameter adding” in two ways. Partial “parameter adding” considers the cumulative ICBT dose from deformable registration as one parameter while full “parameter adding” uses fractional ICBT parameters. The cumulative minimum doses to 2cc (D2cc) of the rectum and bladder were compared between deformable registration and “parameter adding” techniques. Results Dose summation of ICBT fractions only using deformable registration yielded D2cc values that were (10.1 ± 9.5)% lower for the rectum and (7.2 ± 6.3)% lower for the bladder compared to “parameter adding”. When ICBT and EBRT doses were summed deformably, the group without EBRT boost had D2cc that were (0.0 ± 4.6)% and (−1.2 ± 2.9)% lower for the rectum and bladder respectively compared to partial “parameter adding”. With EBRT boost, the differences were (−2.9 ± 4.0)% and (−3.2 ± 3.3)% for the rectum and bladder respectively. For full “parameter adding”, the differences from deformable sum were (2.7 ± 5.0)%, (2.6 ± 5.0)% without EBRT boost and (0.6 ± 4.8)%, (−1.5 ± 3.7)% with EBRT boost. Conclusion Comparison of deformable dose summation with the technique of “parameter adding” suggests that “parameter adding” can be used as a good approximation of D2cc when adding ICBT and EBRT doses with or without boost. With EBRT boosts, deformable dose summation may more accurately represent dose to normal critical structures but these differences remain small compared to “parameter adding”.