Novel approach for determining beam latency using the center of mass of the radiation field measured with a diode detector array.

PURPOSE

This study presents a novel approach to measure beam-on/off latency of a surface-guided radiotherapy (SGRT) system by employing a setup with a commercial diode detector array to simultaneously capture motion profiles and monitor real-time radiation delivery on a single timestamp, without the need to synchronize different measurements devices.

METHODS

Beam-on/off latencies were assessed for an AlignRT SGRT system integrated with an Elekta Versa HD and a Varian TrueBeam linear accelerator (linac). The SGRT system triggered the beam by tracking a commercial modular phantom setup comprising of an anthropomorphic head add-on surrogate and a centrally located diode detector array, with the setup mounted on a Dynamic Motion Platform (DMP). Leveraging the diode array's high sampling frequency of 50 ms, motion was simultaneously tracked by mapping the dose distribution's center of mass (CoM) to the radiation delivery timeline during the delivery of a 2×2 cm2 static field. The region of interest (ROI) for CoM calculation was confined to dose levels exceeding a 10% threshold. Latencies were evaluated for 6 MV flattened (FF) and flattening filter-free (FFF) beams at various dose rates (DR) using three DMP motion patterns (sinusoidal, sharkfin, and sawtooth). Amplitude-based gating was implemented at the SGRT system with a 5 mm position tolerance.

RESULTS

For the Versa HD, beam-on latency averaged 1650 ± 110 ms, remaining stable for DRs of 400-1200 MU/min for the 6 MV FFF beam. For the 6 MV FF beam, latency increased exponentially to approximately 2000 ms when the DR dropped from 200 to 50 MU/min. Beam-off latency averaged 230 ± 90 ms and was independent of DR for both 6 MV and 6 MV FFF beams. For the TrueBeam linac, beam-on latency averaged 280 ± 90 ms, showing no DR dependence in either beam configuration, while beam-off latency averaged 360 ± 90 ms, slightly shorter for 6 MV FFF compared to 6 MV FF.

CONCLUSIONS

The CoM method has been effectively implemented to quantify beam-on/off latencies by integrating SGRT-based motion tracking with real-time radiation dose measurements from a diode detector array, all synchronized to a single timestamp. The results revealed latency variations across different linac vendors and underscored the influence of dose rate and beam energy. With this approach, beam-on/off latencies can be efficiently assessed, providing crucial guidance for defining target volume margins, enhancing both precision and safety in high-dose stereotactic radiosurgery (SRS).