Evaluation of Inherent Dose-Uncertainty for VMAT Using a Dose-Uncertainty Model
T Inoue*, S Sugimoto, Y Horikawa, C Kurokawa, K Fukata, A Isobe, K Ito, K Sasai, Juntendo University Graduate School of Medicine, Bunkyo-ku, TokyoSU-E-T-23 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
To investigate the dose-uncertainty, which can occur during treatment planning and dose delivery in volumetric modulated arc therapy (VMAT), by generalizing the dose-uncertainty model.
QA plans for prostate cancer patient treated with VMAT were created for a water phantom (40x40x20 cm3) using Pinnacle3 radiation treatment planning (RTP) system (Philips Radiation Oncology Systems, WI). The plans were exported as DICOM RT dose files from RTP, and the files were processed using an in-house program to obtain the dose-uncertainty map. Inherent uncertainty of dose calculation (IU) was evaluated in this study. The IU considered was originated from the error of dose in a high-dose gradient region caused by finite size of the calculation grid and the finite size of the detector system during RTP commissioning. The IU was calculated for each control point and was summed up to obtain total IU for each VMAT plan.
Dose-uncertainty distribution on the isocenter plane in the axial slice was investigated. Comparing with a dose-uncertainty distribution of a prostate IMRT plan, which was evaluated by other group, dose-uncertainty distribution of VMAT plan spread out around the target, because beams entered from all directions over 360 degrees. The maximum IU, which was around 10 % of the prescribed dose, appeared the boundary between the target and the OARs (bladder and rectum).
In this study, dose distribution with the inherent dose-uncertainty was evaluated using the dose-uncertainty model generalized for VMAT, and applied to VMAT plans for prostate cancer patients. IU of VMAT spread out around the target if compared with that of IMRT plan. The maximum dose-uncertainty was found in the boundary between the target and OARs and was about 10 % of prescribed dose.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by JSPS KAKENHI Grant Number 23791449