Dose verification for rotating multileaf collimator IMRT

A new method of delivering IMRT has been proposed in which the entire MLC is rotated between each segment. Current linacs were not designed for IMRT delivery with collimator rotation and extensive quality assurance testing must be done before it can be used clinically. This work describes the different areas of testing that need to be considered, namely (1) the commissioning and QA of the Dynamic Beam Delivery (DBD) Toolbox (Varian Medical Systems, Palo Alto) required for collimator rotation control on a Varian CL21EX linac, (2) accurate fluence modeling of rotated apertures, and (3) dosimetric verification of full IMRT treatments delivered with a rotating MLC (RMLC). The DBD toolbox was tested and the collimator rotation angle was found to be accurate and reproducible to within 0.5 degrees. Fluence distributions of varying complexity were generated using the RMLC algorithm and the accuracy of the fluence modeling was validated using film based verification methods. IMRT treatment plans for a prostate, nasopharynx, and c‐shape target were generated with the RMLC segmentation algorithm and were delivered to a phantom. Measured and calculated dose distributions were compared using dose difference, distance‐to‐agreement, gamma factor maps and two‐dimensional profiles. The level of agreement was comparable to clinically accepted plans. Our results show that we can accurately control collimator rotation and precisely model fluence distributions generated from rotating MLC apertures. We also show that the RMLC technique is capable of delivering 3D clinical dose distributions accurately and reproducibly. This work was supported in part by Varian Medical Systems. © 2005, American Association of Physicists in Medicine. All rights reserved.