Click here to


Are you sure ?

Yes, do it No, cancel

Automated, Predictive, and Optimized Average Dose for Organs at Risk in Proton Beam Therapy

S Petro*, B Morris, K Lai, L Rigsby, S Hedrick, Provision Center for Proton Therapy, Knoxville, TN


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Puropse: The goal in treatment planning is to deliver full prescription dose to the target tumor volume while sparing the OARs. Due to tumor volume variation and OAR proximity, the achievable level of sparing is difficult to determine. Many tools exist for predictive x-ray treatment planning, but few exist for proton therapy. To address this need in our clinic, we wrote a script to determine the achievable average OAR dose for pencil beam scanning(PBS) proton therapy.

Methods: Scripting and testing was performed in RayStation v6. A retrospective examination was performed to determine typical dose falloff from 100% of the prescription, which concluded that falloff was approximately 5% per millimeter. This approximation was used to develop a script to determine the estimated average dose for OARs surrounding the target. This script creates ten uniform expansions around the target from 2mm to 2cm to simulate falloff to 0%. The intersection of these expansions with the OARs is determined. The sum volume of those intersections is multiplied by the falloff (10% of prescription). Then to determine the approximate average dose, this sum is divided by total volume of the structure. Ultimately this script can visually show the disparity from an “optimal” plan, allowing the user to create peak EUD objectives.

Results: A secondary retrospective examination was conducted to determine accuracy of the EUD prediction script. The predicted average was compared to the actual average on past plans to determine script efficacy. Initial findings revealed that the script was predicting lower estimated average doses than actual achieved dose due to OARs not being pushed as hard as possible during optimization. When pushed to the limit, our prediction script estimated actual achievable dose within 3%.

Conclusions: Utilizing our devised script enables efficient optimization of PBS proton plans, producing optimal OAR sparing and tumor volume coverage.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email