Room: ePoster Forums
Purpose: (a) To evaluate the localization accuracy in cranial radiosurgery using the single-isocenter-multiple-lesion technique; (b) to quantify the targeting error due to intra-fractional patient motion
Methods: Treatment plans were developed on a Rando cranial phantom and delivered on a linac under CBCT guidance following our clinical workflow. The cranial phantom contains three radio-opaque markers of about 7 mm in diameter. These spherical markers were contoured out as PTVs which can also be identified on the portal images and were used to evaluate the targeting accuracy. The distance between the isocenter and the PTV ranges from 0 to 8 cm. At the treatment, the phantom was first immobilized in the Encompass mask and driven to the initial couch positions followed by localization CBCT imaging. The bony anatomy of the phantom allows 3D registration of the CBCT to the planning CT, which provides 6D couch corrections. A verification CBCT was taken before commencing treatment. The threshold for additional couch correction is 1mm/1deg. The deviation between the centers of the treatment field and target in the portal images is defined as the localization accuracy. Angular errors were intentionally introduced to simulate intra-fractional patient motion to assess its effect on targeting accuracy.
Results: The localization accuracy was within 0.4 mm in all cases, independent of the iso-target distance. Intra-fractional patient motion, as simulated by an angular error up to 1 degree, has negligible effect on targeting accuracy when the isocenter is centered on PTV; however, targeting error increases with iso-target distance. With a 1 degree error, the targeting error exceeds 1 mm, a typical PTV margin in radiosurgery, at a iso-PTV distance of 5 cm.
Conclusion: The study demonstrates a high localization accuracy in stereotactic radiosurgery using the single-iso-multiple-lesion technique. However, robust intra-fraction monitoring of patient motion is critical to maintain the targeting accuracy.
Not Applicable / None Entered.
Not Applicable / None Entered.