SU‐E‐J‐38: Fast Patient Localization Method in Total Marrow Irradiation with Helical Tomotherapy

Abstract

Purpose: Reducing imaging and treatment times is a crucial issue in total marrow irradiation (TMI) with helical tomotherapy. The purpose of this study was to develop the fast patient localization method for TMI patients with helical tomotherapy. Methods: The topographic images were acquired by getting the megavoltage CT (MVCT) after running calibration procedures with the couch traversing through the gantry at static orthogonal beam angles. Geometric and detector response corrections were performed to generate a MV topogram (MVtopo). We also generated kilovoltage topogram (kVtopo) from the projection data of 3D CT images datasets to reproduce the same geometry as MVtopo. Imaging dose and the optimal image acquisition parameters including jaw size, couch velocity, and dose rate were investigated. Geometric uncertainty and image quality of MVtopo were analyzed. In‐house image registration software was developed and verified with a rando phantom. Finally, MVtopo was acquired for a TMI patient. Results: The longitudinal geometric uncertainty was 0 mm, 10.2 mm, and 23.6 mm, in the jaw size of 0.3 cm, 1 cm, and 2.5 cm, respectively. No jaw size dependency in the lateral geometric uncertainty was observed. Couch velocity did not have an effect on the geometric uncertainty. The doses of MVCT and MVtopo with the jaw size of 3 mm and the couch velocity of 0.8cm/sec were 1.5cGy, and 1.7 cGy, respectively. Contrast‐Noise‐ratio, noise, and MTF were improved by using lower dose rate (55MU/min) and smaller jaw size (3 mm). The image registration accuracy with our in‐house software was within 1 mm. In a clinical TMI case, MVtopo with couch speed of 1.0cm/s reduced imaging time by 80% compared with MVCT. The couch shifts derived from KVCT with MVCT registration, and kVtopo with MVtopo registrations agreed within 4 mm. Conclusion: The MVtopo could be useful for TMI treatment with helical tomotherapy. This work was supported by the National Institute of Health grants (1R01CA154491‐01).This work was also supported by PHS Cancer Center Support Grant P30CA77398.