Session Item

GPUMCD is a fast dose calculation algorithm used for treatment planning on the Unity MR-Linac. Treatments for the MR-Linac must be calculated quickly and accurately, especially for two important MR-Linac aspects: off-axis positions, and angular transmission through the cryostat, couch and MR-coils. Therefore, the aim of this research is to quantify the largest system related errors for GPUMCD calculations over the range of clinically-relevant field configurations and gantry angles. 

The error was investigated comparing dosimetric outcomes from measurements and calculations. Dose profiles (crossline, inline and PDD) were measured and calculated for 11 different (off)-axis positions (up to 15 cm) and varying field sizes ranging from 2x2 cm² to 10x10 cm². A microdiamond detector in combination with the PTW Beamscan MR watertank was used to measure relative PDD dose profiles and relative crossline and inline dose profiles at a depth of 10 cm. Absolute doses for all fields were measured using a microdiamond and Semiflex 3D detector, normalised to an expected dose in a 10x10 cm² field. The calculations were performed using a dose grid of 1 mm with a statistical uncertainty of 0.2% (fields ≤ 3 cm²) or 0.5% (fields > 3 cm²) per field.
The transmission was investigated at different gantry angles by measuring and calculating the relative dose of a 10x10 cm² field, normalized to the reading at gantry 90°, with 100 MU every 5 degrees. The measurements were performed using the cryostat characterization tool from Elekta with a Farmer chamber at isocenter in a 56 mm diameter build-up cap. The calculations were performed using a dose grid of 2 mm with a statistical uncertainty per field of 0.5%.

Differences in absolute dose between measurements and calculations ranging from -1.8% to 2.1% were demonstrated (absolute difference ranging from -0.8 cGy to 1.0 cGy per 100 MU). For the relative crossline profiles the dose difference in the central region ranged from -0.5 ± 0.7% (mean ± standard deviation) to 1.0 ± 1.6% (see Fig. 1). For the relative inline profiles the dose difference ranged from -0.6 ± 0.6% to 0.8 ± 1.3%. For the PDD profiles the dose difference ranged from -0.1 ± 0.4% to 0.9 ± 0.5%. Positions further from the central axis or different field sizes did not cause larger dosimetric errors.

The measured angular transmission generally showed good agreement with the calculations with transmission differences ranging from -1.1% to 2.0% (see Fig. 2). Only, relatively large deviations of up to 2% were observed for beams passing through the edges of the table (120 and 240 degrees). 

GPUMCD is a suitable dose calculation algorithm illustrating similar accuracy of off-axis fields and central axis fields. The largest error is caused by the couch transmission and the differences in absolute dose, showing errors up to 2.1%. The relative dose profiles are showing a lower dosimetric error, with a maximum dose difference of 1% between measurements and calculations.