ESTRO 38 Abstract book

S950 ESTRO 38

heterogeneous treatment dose, it is still advisable to minimize unnecessary and avoidable uncertainties by recalculating the imaging dose using the fractional couch shifts. EP-1759 Patient plan QA using EBT3 GafChromic film for the Unity MRI-Linac system J. Wolthaus 1 , S. Hackett 1 , B. Asselen- van 1 , W. Vries- de 1 , S. Woodings 1 , J. Kok 1 , P. Kroon 1 , B. Raaymakers 1 1 UMC Utrecht, Department of Radiation Oncology, Utrecht, The Netherlands Purpose or Objective In our department an Elekta Unity MRI-Linac was installed, accepted and commissioned prior to start clinical patient treatments in August 2018. Individual treatment plan QA on the Unity is performed for both pre-treatment and online adapted plans. The presented work describes our patient plan QA procedure of absolute film dosimetry using EBT3 film obtaining high resolution, high accuracy dose measurements to compare to TPS dose calculations. Material and Methods Plan QA measurements using EBT3 film were performed for more than 30 7-field IMRT plans (optimized to deliver 800 cGy in a single pelvic lymph node). An associate OD- to-Dose model has been determined in advance for the conversion of the pixel values of the film (optical density) to absorbed dose. For each box of films, an additional dose calibration was performed by cutting four small pieces of film (5x10 cm 2 , film orientation specified) and irradiating three of the films with defined doses (650, 800 and 1100 cGy) on the Unity system. The OD-to-Dose model is updated with the results of this measurement. On the Monaco TPS, the patient plan is calculated (2mm grid, MCVar 1%) on a 19x30x30 cm 3 RW3 slab phantom. Subsequently, the point location of the maximum dose is determined and the height defines the coronal film measurement plane. The 3D dose file is exported for TPS to film comparison. The RW3 phantom is positioned isocentrically on the couch. The film is placed at the determined height using slabs of various thicknesses (1– 10mm). A few droplets of water are spread below and above the film to prevent any (small) airgaps. The plan is delivered. Films are digitized the next day (Epson10000XL), to reduce the effect of post-exposure OD growth. Film images are corrected for scanner distortions and converted to absolute dose with the updated OD-to-Dose model, using in-house developed software. The same software is used for image registration of the 3D TPS dose file to the 2D film dose. Profiles can be extracted in both data sets to evaluate the similarity. A 2D local Gamma analysis (5%/2mm, stereotactic clinical criteria) in the film plane is performed. Results Currently, 34 patient plans have been analysed with this method: 12 prior to clinical introduction with varying tumour size and location and 22 clinical plans from 4 LN patients. The figure shows an example of the dose images, profiles and gamma distribution. All plans fulfilled the clinical criteria for an acceptable treatment plan and measured and calculated plans agreed well, the average pass rate was 99.7% using a 60% target dose threshold (mean gamma 0.3±0.2).

Conclusion A patient plan QA procedure using film was developed for the MRI-linac. No additional ion chamber measurements are required to normalise the film dose. The methodology can readily be extended to conventional linacs or ViewRay systems. The results showed excellent agreement between planned and delivered doses for the Elekta Unity system. EP-1760 Impact of cranial implants on proton dose distributions A. Sjögren 1 , K. Andersson 2 , L. Stolarczyk 2 , U. Granlund 1 , C. Vallhagen Dahlgren 2 1 Örebro University, Department of Medical Physics- Faculty of Medicine and Health, Örebro, Sweden ; 2 The Skandion Clinic, The Skandion Clinic, Uppsala, Sweden Purpose or Objective The presence of metallic implants in patients can cause considerable uncertainties in proton therapy treatment planning. The purpose of this study was to investigate the impact of cranial implants on proton dose distributions calculated with Eclipse v13.7 (Varian) using pencil beam scanning technique. Experiments were focused on verification of range calculations for plans based on CT series with and without metal artefact reduction. Material and Methods The four implants were investigated; a 0.3 mm and a 0.4 mm Ti mesh, a 0.5 mm burr-hole cover and a craniofix implant with a diameter of 16 mm. Each implant was attached to a bone equivalent plate (CIRS) and scanned on a solid water phantom with a Siemens SOMATOM Definition AS Open CT. The images were reconstructed with and without the metal artefact reduction algorithm iMAR. In addition, reference images without implants were acquired. A treatment plan with 35 mm WET range shifter, field size 8 x 8 cm, range 10 cm and modulation 8 cm was created in Eclipse 13.7 using PCS beam model and a calculation grid size of 1 mm. Measurements of depth dose distributions were performed in a water phantom (BluePhantom2, IBA) using Roos IC (PTW) to verify the calculations. In addition,a theoretical value for the proton range shift introduced by a titanium plate was calculated using the equation proposed by Moskvin et al: Δ(t M ,Z)=t M [ ⍴ M (1.192-0.158 ln (Z)-1] . [1] Where t M is the thickness of the titanium plate, ⍴ M is the plate density and Z the atomic number of titanium. Due to the structure of the craniofix implant, film measurements were performed instead. For this purpose Gafchromic EBT3 film, calibrated in a proton beam with the red channel method, was used to determine the dose at several different depths within a solid phantom setup.