Exhibit Hall | Forum 2
Purpose: As an analogy to kV imaging and treatment coordinate coincidence in AAPM TG142, MRI-MV isocenter coincidence check in MRgRT needs to be performed. However, the current isocenter coincidence checking procedure in 0.35T MRgRT doesn’t consider variations of MRI isocenter and distortion at various gantry angles. In this study, we proposed in-situ QA tools for quantifying MRI-MV isocenter coincidence and distortion in 0.35T MRgRT.
Methods: In-situ QA tools included the modified QUASAR™ MRID3D cylindrical phantom and software to quantify MRI-MV isocenter coincidence and distortion in position. The center block of the QUASAR™ MRID3D cylindrical phantom (Modus QA, London, Ontario, Canada) has been redesigned to accommodate MV isocenter measurements. The modified center block included two imaging surrogates in determining the MRI isocenter and correlating with 1502 machined fiducials for image distortion quantification. In addition, two cylindrical compartments house a circular film for in-plane star-shot measurements and a rectangular strip of film for a longitudinal MV position. Axial volumetric 3D images at various gantry positions (0~3300 with 300 intervals) were acquired using TrueFISP sequence on the phantom followed by star-shot irradiation.
Results: The MRI isocenter shift (mean: x=0.56 mm, y=0.04 mm, z=0.03 mm) and distortion of the in-situ QA tools (mean±std for distortion: x=0.24±0.23 mm, y=0.35±0.41 mm, z=0.20±0.26 mm) were in agreement with the results of the original phantom. MV isocenter variation of the in-situ QA tools was consistent with our routine star-shot measurement (Minimum tangent circle=0.31 mm, Maximum distance from the origin=0.63 mm, Long shift=0.98 mm).
Conclusion: We demonstrated the clinical feasibility of in-situ QA tools for MRI-MV isocenter coincidence and distortion in 0.35T MRgRT. The proposed approach streamlines MRI-MV isocenter coincidence and distortion checking procedure to characterize MRgRT system more accurately and efficiently.
Not Applicable / None Entered.