Room: Exhibit Hall
Purpose: The purpose of this study was to test the clinical viability of creating a RapidPlan (RP) model on prostate or prostate fossa including seminal vesicles if applicable and comparing treatment planning results to the Washington University â€“ St Louis (WUSTL) prostate model in Varian library as well as the clinically approved individually optimized plans (IOP).
Methods: RP model was generated and trained using 45 clinically treated RapidArc volumetric modulated arc therapy (VMAT) plans. Model generated dose objectives were used as optimization parameters with manual modifications to the planning target volume (PTV) and Normal-Tissue-Objective (NTO). Model OARs included the rectum, bladder, left and right femoral head. The applicability of the model was tested against the IOP and the WUSTL model optimized plans for 10 cases. One round of optimization was performed without planner intervention. All plans were normalized to PTV coverage V100%=95%. Differences between plans were compared for PTV max dose, Gradient Index (CGI), doses to V75Gy, V70Gy, V65Gy for bladder and rectum and total MU.
Results: RP, IOP, and WUSTL values for PTV max dose for all < 110% and significantly similar (p<0.05 for all). Bladder doses for all plans were similar (p<0.05 for all). Rectum doses for IOP were superior to RP and WUSTL values. RP rectum values were lower than the WUSTL model (p<0.05). MU values varied between 488 to 688 and varied between models. In addition, the plans generated with RP model were expedited with total optimization and dose calculation time of 12Â±3 min.
Conclusion: This work demonstrates the feasibility of creating in-house prostate model to generate clinically acceptable treatment plans of high quality.
Not Applicable / None Entered.
TH- External beam- photons: VMAT dose optimization algorithms