Measurement of Time Delays in Gated Radiotherapy for Realistic Breath Motions
B Chugh*, W Smith, University of Calgary, Tom Baker Cancer Centre, Calgary, ABSU-E-T-120 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: The accuracy of treatment delivery in gated radiotherapy depends on time delays between target entry into the gated region and beam-on, as well as target exit from the gated region and beam-off. This study measures time delays, extending past work with simple sinusoidal motion to more realistic breath motions.
Methods: The setup consisted of two platforms moving synchronously, programmed with realistic breathing motion. The vertical platform supported an infrared marker block, tracked by the Varian Truebeam Real-time Position Management system. The horizontal platform supported an x-ray sensitive film with a copper sheet for build-up. A stationary cone mounted above the film defined a spot, translated into streaks by the moving platform. These film streaks were measured to determine the treatment-beam delivery time, based on the speed of the platform. Time delays were determined from differences between the duration of the gating window and treatment-beam delivery time. To discriminate beam-on and beam-off time delays, either the start or the end of the gating window coincided with a stable point of motion. Time delays were measured for different beam energies and motion patterns.
Results: At 6MV beam energy, beam-on time delay was significantly longer (p=0.01) for realistic (mean±SEM of 172±8 ms) than for sinusoidal (mean±SEM of 133±3 ms) motion and beam-off time delay was significantly longer (p=0.05) for realistic (mean±SEM of 83±6 ms) than for sinusoidal (mean±SEM of 64±4 ms) motion. 15 MV results were consistent for beam-on and beam-off delays. The observed late time delays can result in treatment inefficiency (beam-on delays) or, more importantly, to geographical miss (beam-off delays).
Conclusion: The shape of breath motions may be a significant factor determining the time delay. These preliminary findings suggest that the use of sinusoidal motion may underestimate time delays. Further work is needed to determine if sinusoidal motion is sufficient.