ESTRO 37 Abstract book

S976

ESTRO 37

Results The DIR and thus dose warping highly depended on the defined deformation volume. The warped dose in the exhale phase represented rib and tumor doses well while the entrance dose in the lung was clearly wrong due to insufficient modelling of the sliding pleura motion by the DIR (Fig 1B). Although the SS method used a wrong entrance path through the ribs the target doses agreed well with the original inhale doses (Fig 1D+F). The Dice index between warped and SS dose in the exhale phase was 0.77 and 0.71 for the 50% and 95% isodoses, respectively. The DVH parameters of the full plan agreed well for LN, but had discrepancies for T (Fig 2).

Material and Methods The measurements are performed with the StereoPHAN phantom (Sun Nuclear). The phantom consists of a cylindrical housing (PMMA), with a cubic insert for radiochromic film (EBT3). The following steps of the SRT chain are included in the end-to-end test: CT acquisition, isocentre determination in the treatment planning system, phantom positioning on the treatment couch, Cone Beam CT (CBCT) imaging, image registration in XVI, the automatic couch shift, and dose delivery. The MR/CT registration error, and intrafraction motion are not included. The CT scan (Siemens, Sensation open) of StereoPHAN was acquired with our stereotactic brain protocol (slice thickness 1.5mm) with the film in coronal orientation. The phantom was aligned at the lasers and positioned parallel to the table with a digital spirit level. A stereotactic brain VMAT plan with 2 arcs, 6MV FFF, table rotation 0° and 90° (Philips, Pinnacle 9.10 and P16), was copied to the CT data set of the phantom and the dose was recalculated with a dose grid of 2 mm and control points at 2 degrees. On the linac (Elekta, Versa HD), the phantom was aligned at the lasers and positioned parallel to the table. A CBCT was acquired and the film insert was registered to the planning CT with a grey value registration. The setup error was corrected for with the automatic table shift. Scans and table shifts were repeated until the setup error was smaller than 0.05cm. These setup errors cannot be corrected with the automatic table shift. Therefore the deviation of the isodose lines in the film were corrected for the residual setup error in XVI up to 0.1mm. The phantom with EBT3 film was irradiated with the stereotactic plan. The film was scanned and converted to a dose map, which was registered to the planned dose with a grey value registration algorithm in Verisoft (PTW, Freiburg). To gain statistics of the accuracy of our treatment chain, the end-to-end test was repeated 10 times and the mean and standard deviation (SD) were computed. Results In figure 1, the deviation (AB, GT) of the isodose lines in the film compared to the plan are depicted for 10 measurements. The mean deviation is -0.1mm in the AB direction and 0.0mm in the GT direction. The SD is 0.2mm in both directions.

Conclusion The SS dose reconstruction agreed with the warped dose for the single spot plan. For the full plan, the agreement was good for the LN, but only fair for the T probably due to volume changes and lung deformations near the tumor. Future work will investigate the SS method for larger tumors and smaller motion, where it is likely to perform better. The SS method is applicable when a 4DCT is unavailable, e.g. for respiratory motion exceeding the 4DCT motion or target drift during deep-inspiration breath-hold. EP-1813 End-to-end test with StereoPHAN to determine the accuracy of a stereotactic treatment chain C. Panneman 1 , J.J. Kaas 1 , A.M. Olszewska 1 , T.A. Van de Water 2 , A.M. Van Mourik 1 , F.W. Wittkämper 1 1 Netherlands Cancer Institute, Radiotherapy department, Amsterdam, The Netherlands 2 Radiotherapeutisch Instituut Friesland, Radiotherapy department, Leeuwarden, The Netherlands Purpose or Objective In stereotactic radiation therapy (SRT), a high dose in one or few fractions is prescribed to a small volume. To guarantee that the tumor receives the prescribed dose, a PTV margin is added to account for geometric uncertainties in the treatment chain (e.g., machine related uncertainties, setup errors, intrafraction motion). The goal of this study is to determine the total geometric inaccuracy in the brain stereotaxy treatment chain by means of an end-to-end test for a linac. The results will be related to current use of the PTV margin.

Conclusion The inaccuracies in the treatment chain are very small in relation to our current PTV margin of 2mm. However the