ESTRO 2021 Abstract Book
S1452
ESTRO 2021
stoichiometric method proposed by Schneider et al. in 1996. Plastic inserts of accurately known and certified electronic density and chemical composition from a CIRS phantom were scanned located at the center of a solid water phantom. The calibration predicts the response of the CT for photoelectric, coherent and Compton scattering in terms of its chemical composition. Next, knowing the chemical composition of biological samples, the predicted HU units using the CT calibration were calculated, as well as the SPR for those samples, using the Bragg additivity rule for the ionization potential. This relationship between HU and SPR was eventually fitted to a three-linear curve. For the biologic validation, a series of 14 samples made of pork and cow tissues including fat, cortical bone, soft bone, brain, blood, round, loin, liver, kidney, skin, and ground meat were packed in plastic boxes of 5 x 5 x 8 cm 3 . Those boxes were inserted on a water equivalent phantom and scanned in the CT to get the HU and eventually obtain the SPR with the CT calibration curves. Additionally, the samples packed in the boxes were used to measure the SPR using a proton beam. They were placed on the side of the water tank and an IDD was measured using a Bragg peak chamber, with the sample located between the beam and the tank. The IDD was repeated with the same energy without the sample. The relationships between the R80 metrics for both IDD was used to determine the SPR. Results Agreement between the SPR determined by CT calibration and by measurement was found to be similar for the three CT. We disregarded the experimental values for the samples prepared with soft bone and with ground meat because deficiencies in the sample preparation (air gaps). We present the data for the differences between predicted and measured SPR for the Drive in the following table:
Tissue sample
Diff. (%)
Fat
-1.16
Skin
-0.2811
Brain
0.9722
Kidney
0.0032
Pork loan
0.092
Blood
0.943
Ground round
-0.3655
gGround liver
0.0918
Liver
-0.0455
Round
-0.0975
Loan
0.1366
Bone (core)
-0.474
For our clinical practice and CT calibration curve introduced in our TPS, we tweaked the table for the HU corresponding to fat to further reduce the uncertainty in this area.
Conclusion Our CT calibration was experimentally validated ensuring the reliability of its determined SPR values for dose calculation in nthe presence of biological tissue inhomogeneities.
PO-1728 Optimal camera setup for SGRT system on a ProteusOne proton therapy system, a simulation study L. Delombaerde 1 , T. Depuydt 1,2 1 KU Leuven, Department of Oncology, Leuven, Belgium; 2 University Hospitals Leuven, Department of Radiation Oncology, Leuven, Belgium Purpose or Objective Conventional surface guidance systems (SGRT) usually consist of up to 3 cameras and are ceiling mounted. They can aid in patient setup and intra fraction monitoring during proton therapy treatments but require a direct line-of-sight, which can be hindered by the gantry or nozzle in pencil beam scanning proton therapy. Adding additional camera’s to the SGRT system was investigated in a simulation environment to assess the improved surface visibility compared to the standard configuration. Materials and Methods A simulation environment was created in MATLAB consisting of geometric models of the ProteusOne gantry and kV-imaging equipment (IBA, Belgium), the uncovered face of ten patients immobilized in open-face thermoplastic masks as a surface mesh and the SGRT camera positions, as shown in figure 1. Two SGRT setups are compared (a) the installed three camera configuration (AlignRT, VisionRT) and (b) the proposed five- camera system consisting of two existing ceiling mounted cameras and 3 small cameras attached to the kV- imaging equipment. The visibility of a specific face of the patients’ surface mesh is determined by tracing a direct line from each of the three corner point vertices to the camera position. Obstruction by either the gantry, nozzle, imager or self for one or more vertices results in the face being deemed not-visible. Additionally, a minimal grazing angle of 25° of the incident ray to the face is required. The total visibility of the surface is tracked and expressed in percent relative to the complete surface. The symmetry of the partial surfaces is expressed as the relative center-of-mass (CoM) shift between the complete surface and the partial surface. Clinically used gantry, nozzle and couch combinations were simulated (gantry: 0° - 25° - 50° - 75°,
Made with FlippingBook Learn more on our blog