Program Information
Flattening Filter for Sliding Window IMRT Treatment Revisited
C Ma*, T Long , M Chen , X Gu , S Jiang , W Lu, The University of Texas Southwestern Medical Center, Dallas, TX
Presentations
SU-I-GPD-T-503 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To study the characteristics of optimal flattening filter shape for sliding window IMRT that minimizes beam-on-time. We challenge the convention of using flattening filter in IMRT treatment. A mathematical model is formulated for determining optimal flattening filter or flattening filter free (FFF) in sliding window IMRT treatment.
Methods: Conventional sliding window IMRT delivery utilizes a flattening filter, which may increase delivery time due to the reduction in beam profile. The effects of field size, beam profile, fluence profile, and flattening filter thickness have on delivery time are modeled and studied. We assumed conic flattening filter and beam profile shapes, where the relative peak height can be adjusted. Given a desired fluence profile, the resulting linear programming model yields an optimal filter effect. Single-leaf pair fluence maps were analyzed and solved in closed form to gain insights for arbitrary clinical fluence profiles. Clinical cases were studied as well.
Results: Mathematical analysis of a single leaf pair model showed that for not-so-large IMRT fluence (field size<25cm) with sliding window delivery, FFF delivery was optimal. The model was applied to several different sizes of clinical fluence maps. In 700 IMRT prostate fluence maps with field size about 10 cm and 17 IMRT head and neck fluence maps, it was found that FFF was the best choice to minimize the total beam-on-time. FFF boosted the treatment efficiency to about 50% on average for these fluence maps. As the size of the fields increases, the efficiency of FFF delivery decreases.
Conclusion: The conventional practice of using a flattening filter for sliding window IMRT should be re-evaluated. By using FFF delivery, beam-on-time can be decreased significantly. This study lays the foundation for more complex filter shape optimization.
Contact Email: