Purpose: To build on previous experiments and improve reproducibility of electron FLASH delivery on a conventional linear accelerator, a pulse-gating circuit was constructed and tested with several dosimeters.
Methods: FLASH mode was achieved by operating a Varian iX linear accelerator with 15MV photon operating parameters, with the target and flattening filter retracted and a 12MeV electron foil in the beam’s path. Pulse gating was achieved by an Arduino-based circuit interfaced with the linac gating module and a scintillator-photodiode sensor that was in the beam’s path to count pulses. Dosimeters were placed at a source-to-surface distance of 47-47.5cm, with 0-5mm of build up and 2-3cm backscatter. Measurements were performed in open fields and on cut-outs ranging from 1x2cm2 to 3x3cm2. Doses were measured using a 0.01cc volume ion chamber, optically stimulated luminescence dosimeters (OSLDs), Gafchromic MD film and a novel plastic scintillation detector and Hyperscint spectral analysis. Film profiles and PDD were used to confirm dosimetry measurements were performed in a region receiving >90% of maximum dose.
Results: A dose rate of 225 ± 11Gy/s was achieved in FLASH mode. The pulse-gating method delivered the specified pulse count in 89% of irradiations and was within 2 extra pulses for 11% of the beams, as confirmed by scintillator. Corrected ion chamber, OSLD and scintillator doses showed excellent agreement for a range of 0-60Gy, while film showed saturation effects past 30Gy. The dose per pulse varied from 1.1Gy/pulse (7 pulses delivered) to 1.3Gy/pulse (40 pulses delivered). Oscilloscope and scintillator readings indicated consistently lower amplitude of the first pulse compared with subsequent pulses.
Conclusion: A prototype pulse-gating method for eFLASH deliveries on a conventional linear accelerator was implemented. Further efforts will aim to reduce circuitry delays resulting in extra pulse delivery and compensation for pulse height for more consistent dose delivery.