Portal dosimetry prediction using Pinnacle ® TPS for clinical quality assurance applications.
PO-1375
Abstract
Portal dosimetry prediction using Pinnacle ® TPS for clinical quality assurance applications.
Authors: Fawzi|, Maher(1)*[maher.fawzi@curie.fr];Cerbah|, Yacine(1);Eid|, Doris(1);Alabdoaburas Mohamad|, Mohamad(1);Sandjon|, Modestin(1);Belshi|, Rezart(1);
(1)Institut Curie Ensemble Hospitalier, of Radiation Oncology, St Cloud, France;
Show Affiliations
Hide Affiliations
Purpose or Objective
The aim of this work is to create a model in Pinnacle dedicated to portal pre-treatment dose verification, for image prediction in clinical radiotherapy routine.
Amorphous silicon Electronical Portal Image Devices (EPID) are widely used for patient positioning. Several work have been done to develop method to use EPIDs for dose map calculations. This study concern Varian Truebeam (EPID a-Si 1200, 6MV, 6MV FFF) and Novalis Tx (EPID a-Si 1000, 6MV and 6MV SRS) for Volumetric Modulated Arc Therapy and stereotactic treatment technics.
Material and Methods
The method was developed for the energies and treatment units mentioned below. The EPID was calibrated according to the manufacturer procedure then portal images were acquired for a variety of rectangular fields, and complex field, using the portal device. The approach is to extract in-line and crossline profiles, output factors and implement them in Pinnacle physics tool to create a portal dosimetry machine that models the response of portal imager device.
The model of portal dosimetry machine (PinnEPID) was based on the parameters of the clinical machine model. Slight adjustment was done in physics tool to best fit between measured and calculated profiles and output factors.
Acquired portal images from simple complex and clinical treatment plans were performed, to evaluate the new model, using local gamma index criteria, (distance to agreement and dose difference).
Results
A portal dose prediction method based on Pinnacle® was developed without modifying the commissioned parameters of the model used in clinic. For jaws-only collimated fields, a good accordance between the measured and calculated profiles in central-axis and off-axis areas were obtained, even for the small square fields sizes ( 1cm², 2cm², 3cm²). Processed-measured output factors ratio were between 0, 97-1,021. The passing rates were respectively 99.6% and 98.2% for routine local gamma index of 3%- 3mm and 2%- 2mm, with 10% of maximum dose threshold. For modulated clinical treatment plans ( Prostate, Pelvis, ENT, Lung, Breast) the mean local gamma index results yield respectively 98.7% and 97.3% of points passing for 3%- 3mm and 2,5mm- 2,5% criteria, with 10% of maximum dose threshold. Pass rates were similar for PTW-OCTAVIUS phantom matrix measurements for local gamma 3%- 3mm. Furthermore, the comparison of the predicted images calculated with Pinnacle® PinnEPID and EclipseTM portal dosimetry showed a very good agreement.
Conclusion
This study showed that the use of machine treatment parameters is sufficient with small adjustment to create efficient and accurate predicted images for pre-treatment patient Quality Assurance.