Purpose: Two-dimensional detector arrays are commonly used to verify IMRT dose delivery. An area of interest is the use of machine log files for machine and patient-specific quality assurance. We aim to develop an application that uses a true Monte Carlo model to calculate dose based on machine parameters extracted from log files from patient plan delivery.
Methods: A model of the Elekta VersaHD linear accelerator was developed using the EGSnrc codes. This model was validated to be within 1%/1 mm accuracy of measured PDDs and profiles. A tool was created in MATLAB that extracts information from machine log files from delivered beams and “injects” it into the patient’s DICOM RP file, replacing the initial parameters used to calculate the “planned” dose. Plans were created from log files from 10 patients’ IMRT QA deliveries. These plans were converted to the inputs required by BEAMnrc/DOSXYZnrc using a MATLAB script developed in-house. Doses were calculated using Source 21 in DOSXYZnrc. Results were compared to measurements made with the PTW Octavius phantom and evaluated in Verisoft with a gamma criterion of 3%/2mm.
Results: The 3D gamma passing rates for log file-based plans calculated in DOSXYZnrc ranged from 95.7% to 99.9%, and all statistical uncertainties were below 3%. In 9 out of 10 patients, doses calculated in DOSXYZ showed better agreement with measured values than the doses calculated in Pinnacle.
Conclusion: Using plans with parameters from log files, we can obtain a more accurate representation of the dose that was delivered to the patient and determine the differences from the dose that was initially planned. The “DICOM Injection” tool can provide greater insight on the impact of machine errors in delivered plan quality. When used with Monte Carlo, users have the potential to see an accurate representation of the dose delivered to the patient.
Funding Support, Disclosures, and Conflict of Interest: The project described was supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant TL1 TR002647. This project was also supported by the Cancer Prevention and Research Institute of Texas (CPRIT) Research Training Award (RP170345).
Monte Carlo, Quality Assurance