(Sunday, 7/14/2019)
Room: ePoster Forums
Purpose: To evaluate the effect of incorrect or incomplete beam data on the accuracy of dose calculations in a model-based treatment planning system.
Methods: In the Philips Pinnacle treatment planning system (TPS), the existing commissioning beam data for 6 MV photons were altered by: 1) Eliminating a large sub-set of the data, leaving only minimum data required by the system (incomplete beam data) and 2) Shifting the cross-profiles for the complete dataset by 5 mm to simulate error in data collection (incorrect beam data). The two sets of data were then used for beam modeling and the model was compared to the existing, complete, beam data model. The “incomplete beam data� and “incorrect beam data� machines were then commissioned and used to compute dose calculations for a variety of plans to be delivered and measured in phantom.
Results: Beam modeling results show that inadequate data has little discernible effect on computed profiles. Incorrect beam data results in an increase in the appearance of horns in the central portion of the computed cross-profiles and widens the penumbra in the computed in-profiles. This is more critical for small fields. The measured doses from plans created by the two sets of data show no difference for static fields but there is a decrease in the gamma pass rates of modulated beams containing small sub-fields.
Conclusion: The minimum data is sufficient to commission a machine in this TPS but incorrect data results in discrepancies in the modeled profiles, especially in the central portion of the profiles and penumbra. This affects the dose calculation accuracy of small fields the most. The 5 mm shift employed here is obviously exaggerated and improbable to happen during beam data collection but smaller shifts, which are more probable to happen, may affect beam model as well.
Keywords
Taxonomy
Not Applicable / None Entered.
Contact Email
