Purpose: Motion management is a critical component of radiotherapy. We previously proposed a new scheme of using kV scattered x-ray photons for marker-less real-time tumor tracking in lung cancer radiotherapy. This study performs Monte Carlo (MC) studies and experimental validations to characterize its performance.
Methods: A thin slice of x-ray beam (120kVp, 200mA, 50ms) was generated from the x-ray tube on a Varian TrueBeam linear accelerator and the slice was configured to pass the iso-center and the superior-inferior axis. The beam was directed to the moving target around the iso-center in a CIRS lung phantom. We measured the outgoing Compton-scattered photons using a photon-counting detector with a slat collimator. We set up a MC simulation system using an in-house developed GPU-based MC package for kV photon simulations and GATE/Geant4 for MV photon simulations modeling the image formation process. We also investigated effects of scattered photons from the therapeutic MV beam, the impact of scattered photons from bed and walls, and the imaging performance under several beam angles typically used in real treatments. We compared results in simulations and measurements.
Results: The tumor can be clearly identified on the scattered images in both simulation and experimental results. MC simulations and experimental measurements agreed within 9 %. The scattered photons from the therapeutic MV beam generated a relatively uniform background signal without obvious anatomy structure; hence, they are expected not to affect tumor tracking. MC simulations showed that the amplitude of scattered photons from bed and walls was < 3% of the imaging signal of interest. The image quality does not significantly depend on kV beam angles.
Conclusion: MC simulations and experimental measurements confirmed the feasibility of using scattered x-ray imaging as a method for real-time image lung tumor tracking.
Not Applicable / None Entered.
Not Applicable / None Entered.