Author: Wei Luo, Cameron Thayer-Freeman, Yifeng Zhao 👨🔬
Affiliation: University of Kentucky, Department of Radiation Medicine, University of Kentucky 🌍
Purpose: Real-time Kilovoltage (kV) imaging has emerged as a powerful clinical aid, providing high-precision localization of moving targets during treatment. However, the Megavoltage (MV) treatment beam may interfere with kV imaging simultaneously. This study investigates the impact of MV beam on kV and fluoroscopy image quality, assessing target visibility and image contrast in the case of static fields and clinical IMRT(VMAT) plans.
Methods: The study was conducted on a Varian TrueBeam using a Winston-Lutz cube phantom. The kV images were acquired while the MV beam was on using the triggered imaging technique. Contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) were used to assess image quality, with analysis being done in ImageJ. By closing the collimator, the effect of electronic noise from the machine was also assessed. Five photon energies (6MV-18MV, and 6MV-10MV FFF) were tested.
Results: The CNR and SNR of open field images decreased by up to ~7% compared with closed field images. CNR and SNR decreased by up to 20% when comparing closed fields to kV imaging with beam off, and up to 22% when comparing open fields to kV imaging. The effect of field size on image quality was within 10%. For IMRT plans, CNR/SNR lost over 20% with 40%, the largest loss in SNR for SBRT. Fluoroscopy for IMRT dramatically reduced image quality by over 90% in CNR.
Conclusion: Both MV scatter and electronic noise have significant impact on real-time kV/fluoroscopy image quality. Electronic noise was the dominant factor in evaluating image quality, with beam scatter contributing a comparatively smaller amount. The impact was not correlated with beam energies. Varying IMRT techniques can have different effects on image quality, with dose rate and field size being potential influencing factors. This was more pronounced in fluoroscopy. Further research is needed to improve real-time image quality.