Open Source Tool to Compare MR-Linac Logfiles with Monaco RT Plans
M Kusters1*,J Kaas1, P de Graaf1,S Biggs2 (1) Radboudumc, Nijmegen, NL (2) Radiotherapy AI, AUS
Presentations
PO-GePV-T-323 (Sunday, 7/10/2022) [Eastern Time (GMT-4)]
ePoster Forums
Purpose: In this study we adapted the MeterSetMap comparison of the PyMedPhys GUI 0.38.0 to compare Elekta TRF log files of the MR-Linac Unity with the RT plan sent by Monaco to Mosaiq during the online workflow. We tested this tool for 25 different IMRT patient plan treatments for prostate, oligo and rectum. The source code for this software is available online at the PyMedPhys website.
Methods: The software decodes the temporal tracking files and automatically analyses them by comparing them offline to the planned treatment described in the DICOM RT plan associated with the planned beams. Logs and RT plan were analyzed by creating pseudo fluence maps from recorded and planned values of MLC locations, jaw locations, and monitor units per unit time. Fluence maps were then utilized to calculate a 2D gamma index with a 2%–0.5mm criteria for each fraction.
Results: The gamma pass rate was found to be larger than 96.4% for all patient cases. Based on the log files no delivery errors were found. Additional IMRT delivery QA with the Delta4 phantom also passed the dose criteria.
Conclusion: Elekta TRF Log files for evaluation of linac integrity and patient QA can be utilized to allow for reliable analysis of system accuracy and performance. The tool can be used to pick up gross data transfer errors. Additional patient-specific IMRT QA and daily Linac QA will still be necessary. Deficiencies in a plan’s agreement between treatment planning system and the Linac due to deviations such as small field output factors or the MLC model will not be picked up by this tool. Furthermore, calculations of log files alone this uses leaf and jaw positions as reported by the Linac which may not agree with their position in reality due to issues such as a hardware or calibration fault.
Keywords
MR, Linear Accelerator, Quality Control
Taxonomy
Not Applicable / None Entered.
Contact Email