Room: Stars at Night Ballroom 4
Purpose: To present a proof of concept for treatment verification of the Elekta Unity MR-linac adapted workflow (Elekta AB, Stockholm, Sweden) using in vivo 3D EPID dosimetry.
Methods: The Elekta Unity MR-Linac allows for simultaneous irradiation, MR imaging and EPID acquisition of images during treatment. The EPID-based dose reconstruction algorithm for conventional linacs was adapted to account for the increased amount of scatter received by the panel and the extra attenuation by the MRI. The algorithm parameters were empirically determined by fitting EPID images to a set of square field measurements performed under a variety of different phantom setups and gantry angle values. Two 9-field IMRT treatments were considered in this study (one prostate and one rectum). Portal images acquired during the delivery of the first three daily adapted plans were used to reconstruct in vivo patient dose distributions. Additionally, in air portal images of the original plan were acquired to reconstruct virtual patient dose distributions that can be used for pre-treatment verification. The EPID-reconstructed dose distributions were compared to their corresponding planned dose distributions (original and adapted) by performing Î³-analysis (3% global/3mm, 50% isodose) using Î³-mean and Î³-pass rate values. The differences in dose value at the center of the PTV (%Î”DPTV_C) were also considered.
Results: Î³-pass rate values were higher than 95% in the pre-treatment verification of the original prostate and rectum plans, respectively. The average Î³-pass rate and %Î”DPTV_C values in the in vivo verification of the six adaptations were 92.0Â±1.0%(1SD) and 2.0Â±1.9%(1SD), respectively.
Conclusion: The presented results provide the first experimental evidence that 3D EPID dosimetry is feasible with the Elekta Unity MR-linac for both pre-treatment and in vivo verification. This study paves the way for the development of an automated EPID-based dosimetric verification solution for the adapted workflow in the Elekta unity MR-linac.
Funding Support, Disclosures, and Conflict of Interest: Support for this research was provided, in part, by Elekta AB, Stockholm, Sweden.