Room: Exhibit Hall
Purpose: MRIdian allows us to treat moving target with more accuracy using Cine-MR during treatment. The beams automatically pause when the target moves beyond the boundary, and the beam automatically resumes when the target moves back into the boundary. Information of Cine-MR is stored in MRIdian system every fraction of treatment. Our purpose is to assess quality of treatment by the MR-IGRT system in terms of the number of beam-pause, treatment efficiency and tumor movements from the stored Cine-MR information.
Methods: Real-time Cine-MR images with four frames/second during treatments were analyzed using an in-house software for five patients with upper-abdominal tumors by expiratory breath-hold irradiation. A boundary margin was 3 mm in this study. The number of beam-pause was evaluated by analysis of beam status displayed on Cine-MR images. Total beam-on time greater than 10 seconds (T(ex)) were defined as total duration of expiratory breath-hold, and a ratio of T(ex) to whole beam-on time (T(w)) was calculated to assess treatment efficiency for each fraction. Tumor positions were analyzed from the center of mass of three tracking points, which indicates coordinate of the tumor tracked by deformed image registration techniques.
Results: The mean of T(w) and T(ex) in each fraction for all patients was 12.3 (7.6 â€“ 26.7) and 6.7 (2.2 â€“ 25.8) minutes, respectively. The mean of the number of beam-pause was 8.7 /minute (2.0 â€“ 18.7 /minute), the mean treatment efficiency was 52.4% (18.9 â€“ 97.2%), and they have a strong correlation (R = 0.87). The tumor position displacements (1.96 SD) during beam-on ranged from 0.7 to 2.4 mm in the anteroposterior, and from 1.1 to 4.6 mm in the craniocaudal.
Conclusion: Cine-MR analysis provides helpful information to assess treatment efficiency and intra-fractional tumor movements. The finding is that treatment efficiency was strongly correlated with the number of beam-pause.