Program Information
An Automated Method for the Verification and Quality Assurance of Treatment Planning Systems
S Oh*, D Lee , B Leong , Y Wu , S Kim , T Kim , Virginia Commonwealth University, Richmond, VA
Presentations
SU-I-GPD-T-481 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: During commissioning of a LINAC, it is important to verify the consistency of beam data measurements against reference data from the treatment planning system (TPS). The manual verification of dose consistency during TPS commissioning and quality assurance (QA) is tedious and error prone. We propose an automated tool to improve the accuracy and efficiency of this process.
Methods: The proposed procedure consists of 3 processes; data preparation, data formatting, and data comparison. First, TPS plans are manually generated over beam scan geometry. The rest is automatically performed using an in-house MATLAB program . Percent depth dose (PDD), crossline- and inline-profiles are extracted from the 3D dose matrix of the TPS. TPS and Varian provided TrueBeam data are used as a reference. Then, the reference and corresponding scanned beam data are individually compared by computing dose and distance deviations. Graphical profile outcomes are stored for dosimetric comparisons by the physicist depending on the passed distance-to-agreement (DTA) level.
Results: DTAs of PDD and profiles were accurately and efficiently computed across two references and a measurement. Out of 215495 comparison points from 806 scans, 94.58%, 5.07% and 0.34% satisfied 1%/1mm, 2%/2mm, and 3%/3mm DTA criteria, respectively. Physicists can evaluate a concise report summary that includes graphical and tabular results, and can focus on the review of required cases.
Conclusion: An auto-comparison procedure was successfully implemented on the commissioning of a new TrueBeam with the TPS plans, detecting errors in missing beam scans and wrong beam scan conditions. This is an evident that this can be a valuable tool for the verification and QA of treatment planning systems. Furthermore, this approach can be utilized for each data format and across multiple data formats among institutions.
Contact Email: