Room: AAPM ePoster Library
Purpose: develop a framework for patient-specific reconstruction of dose delivery incorporating in-treatment measurements of respiratory motion.
Methods: reconstruction framework is based on a previously developed and validated 4D Monte Carlo dose calculation method (4DdefDOSXYZnrc/EGSnrc), which accounts for patient motion and deformation due to respiration. The method uses as inputs treatment delivery log files, deformation vectors obtained from exhale to inhale CT registration and a respiratory motion trace. This work represents the first application of this method to a cohort of patients undergoing radiotherapy for locally advanced lung cancer.
Patient surface motion during VMAT treatment delivery was recorded using RADPOS system for 3 patients. Measurements were acquired for 1-3 treatment fractions per patient. 4D dose reconstruction was performed and compared for each fraction with a static dose calculation of the planned dose on the exhale CT image as well as the Monaco dose calculated on the average CT. The reconstructed 4D dose distributions along with the static planned dose calculation were then imported into CERR (Computational Environment for Radiological Research) for comparison.
Results: results from the first patient (tumour in right upper lobe) showed a maximum variation in the GTV D98% of 0.5% for the 4D dose calculations over 3 treatment fractions. There was also a decrease of 8.7% in the GTV D98% for one of the 4D dose calculations as compared to the planned dose from the TPS (average scan). An increase in the GTV D2% by 1 cGy was noted in two fractions. The OARs showed no variation between 4D dose calculations for the 3 treatment fractions.
Conclusion: framework has been developed to reconstruct the dose to moving anatomy using Monte Carlo dose calculations. Preliminary results indicate that significant variation between the planned and doses delivered to the target volume in each fraction can occur.
Funding Support, Disclosures, and Conflict of Interest: This research was supported by the Natural Sciences and Engineering Research Council of Canada.