Room: Exhibit Hall
Purpose: RadCalc (Lifeline Software, version-6.2) is routinely used in our clinic to verify computational accuracy of clinical HDR plans generated by Oncentra (Elekta, version-4.5.2) TPS. This study started with a failed clinical plan in which dose to some relevant points deviated over 15%.
Methods: HDR plans were generated on CT using Oncentra, then performed a second check using RadCalc. Numerous plans were created using two dwell positions in central channel and a single dwell position in either one or all of the peripheral channels for cylinders of 25mm, 30mm, and 35mm diameter, having 6, 8, and 8 peripheral channels, respectively. Considered calculation points were: cylinder tip, tip+5mm, tipâ€™ (5mm lateral to tip), and tipâ€™+5mm.
Results: Dose computation difference between TPS and RadCalc tended to increase as the single dwell time in periphery was increased as well as when the cylinder size increased. Dose differences at tip+5, tip and tipâ€™ for 35mm cylinder with 20s dwell time in each of the peripheral channels were 28.4%, 13.5% and 10.4%, respectively. When a second dwell position (t=0s) was added to each peripheral channel, dose differences at tip+5, tip, and tipâ€™ were 16.8%, 8.34%, and 4.52%, respectively. Dose difference at tip+5, with two dwell positions (t=20s and 0s) in a single peripheral channel, decreased from 2.9%, 1.8%, and 6.6% to -0.1%, 1.3%, and 3.9% for 25mm, 30mm and 35mm cylinders, respectively. Dose difference at tip+5, tip, and tipâ€™ decreased to 7.0%, 3.6%, 1.5%, when dwell times in peripheral channels were changed from 20s and 0s to 10s and 10s.
Conclusion: One dwell position is not sufficient for Radcalc to properly calculate the angle alpha and beta used in TG-43 dose calculates. At least two dwell positions in each channel are needed to define the source orientation. Larger distributed dwell time improves dose computation accuracy.