Encrypted login | home

Program Information

Dark Current Effect On CyberKnife Beam Dosimetry

no image available
H Kim

H Kim*, A Chang , Soonchunhyang University Hospital, Seoul, Korea


SU-F-T-554 (Sunday, July 31, 2016) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: All RF linear accelerators produce dark current to varying degrees when an accelerating voltage and RF input is applied in the absence of electron gun injection. This study is to evaluate how dark current from the linear accelerator of CyberKnife affect the dose in the reference dosimetry.

Methods: The G4 CyberKnife system with 6MV photon beam was used in this study. Using the ion chamber and the diode detector, the dose was measured in water with varying time delay between acquiring charges and staring beam-on after applying high-voltage into the linear accelerator. The dose was measured after the time delay with over the range of 0 to 120 seconds in the accelerating high-voltage mode without beam-on, applying 0, 10, 50, 100, and 200 MUs. For the measurements, the collimator of 60 mm was used and the detectors were placed at the depths of 10 cm with the source-to-surface distance of 80 cm.

Results: The dark current was constant over time regardless of MU. The dose due to the dark current increased over time linearly with the R-squared value of 0.9983 up to 4.4 cGy for the time 120 seconds. In the dose rate setting of 720 MU/min, the relative dose when applying the accelerating voltage without beam-on was increased over time up to 0.6% but it was less than the leakage radiation resulted from the accelerated head. As the reference dosimetry condition, when 100 MU was delivered after 10 seconds time delay, the relative dose increased by 0.7% but 6.7% for the low MU (10 MU).

Conclusion: In the dosimetry using CyberKnife system, the constant dark current affected to the dose. Although the time delay in the accelerating high-voltage mode without beam-on is within 10 seconds, the dose less than 100 cGy can be overestimated more than 1%.

Contact Email: