Purpose: To assess the linear accelerator beam delivery accuracy by means of gamma index using treatment log files to take into consideration all sources of error during delivery. A secondary system, PTW VERIQA, that performs an independent Monte Carlo 3D dose calculation was used.
Methods: Twenty patients, selected at random, with treatment sites of head and neck, pelvis, abdomen, or lung, were exported to the 2nd MU verification software. The log files from the delivery of their plans were collected and using a home-grown application, were converted to Dicom plans. They were then imported into the Monte Carlo 2nd verification system which calculated the dose distribution. Two comparisons were then made: a) physician approved plan vs secondary verification plan and b) physician approved plan vs logfile based plan. The gamma index for these plans were compared against the gamma index for the patient specific QA using the Octavius phantom and analyzed using PTW Verisoft software.
Results: The differences of gamma index between a0 and b0 scenarios was within ±5%. The standard deviations between the gamma index calculated in VERIQA using the Dicom plan from the log files, the Dicom plan from Pinnacle, and the patient specific QA ran on the OCTAVIUS phantom was less than 10% for all but two patients.
Conclusion: Dicom plans generated from treatment delivery log files can be used in VERIQA to assess the total gamma index passing or failing rates while also accounting for any errors that may have occurred during delivery. By comparing the gamma index for the total treatment volume as well as various OARs in VERIQA, you can assess the difference between what was generated in the treatment planning software and what was delivered.
Funding Support, Disclosures, and Conflict of Interest: Supported by PTW