ESTRO 38 Abstract book

S979 ESTRO 38

Monaco for the OARs, but were approximately 1.5% hotter than the Monaco for the PTVs (see fig. 2).

Gamma %<1 91% Mean Gamma 0.42 0.28 0.55 0.6 Gamma >1 % 1.09 0.72 1.46 1.5 100 91% -

84 fractions were analysed in the clinical phase. DRP and gamma results are shown in figures 1 and 2.

Conclusion It is possible to create a RayStation beam model for the Unity MR-Linac which is sensitive enough to detect gross errors in monitor unit calculations. With the prototype version of the MR-Linac beam model, the gynae and head and neck treatment sites had similar PTV DVH statistics to the prostate, however, additional sitesse will require validation before clinical implementation head and neck will require validation with this beam model. As the magnetic field is not modelled, care is required when interpreting doses near air interfaces. EP-1807 Use of SPC techniques to generate assessment criteria for transit dosimetry analysis A. Pyett 1 , C. Holmes 1 , C. Enever 1 , A. Fryer 2 , D. Paynter 2 , P. Rixham 2 , S. Weston 2 1 Leeds Cancer Centre, Radiotherapy Department, Leeds, United Kingdom ; 2 Leeds Cancer Centre, Medical Physics and Engineering, Leeds, United Kingdom Purpose or Objective When introducing a transit dosimetry programme the analysis criteria and the failure thresholds must be established. Conventional assessment of delivery using quasi-3D phantoms is usually performed using the γ parameter. A common threshold for gamma success is 95% of points with a γ< 1 using parameters of 3%/3mm. However any transit dosimetry analysis must allow for clinically insignificant variation in the patient setup and the patients anatomy, effects which do not affect phantom measurement. The aim of this work is to use measured data to establish clinically relevant parameters for the analysis of Lung SABR treatments Material and Methods At X lung SABR treatments are planned on Elekta Monaco and delivered using a single arc 6MV FFF VMAT technique on Elekta VersaHD linacs. The patient position is verified using CBCT before every fraction. A portal movie image measured during the treatment is used to reconstruct the delivered dose distribution on the 3D planning CT using the Elekta iViewDose software. iViewDose provides analysis of the delivered dose distribution in terms of Dose at the Dose Reference Point (DRP) Percentage of points that have a γ<1 Mean γ Near maximum γ During a pilot phase dose reconstructions were analysed using γ parameters of 3%/3mm. Mean values and confidence limits for all analysis parameters were calculated using SPC methodologies. The means and confidence limits were measured from the first twenty fractions to be analysed, ignoring the obvious outliers. Following clinical implementation the reconstructions were analysed against the criteria established in the pilot phase. Where fractions passed all the criteria no further action was taken. Where one or more of the criteria failed the fraction was investigated.

All of the fractions that failed at least one test were investigated. Almost all of the failures were true positives in that a reason could be identified for their failure:tumour motion, contour change, or oedema. One patient exhibited a failure which could not be explained due to anatomical changes. This was verified using a Delta4 system and the delivered distribution was found to be in good agreement with the planned distribution Conclusion Transit dosimetry can be an effective tool in determining deviations from intended treatment. Errors identified using transit dosimetry correlated well with anatomical changes identified using CBCT. No errors were identified due to treatment delivery errors. SPC techniques can be used to develop assessment criteria for transit dosimetry analysis. This should be performed on a site by site basis as different anatomical sites will yield different expected results and confidence levels depending upon the accuracy of the reconstruction and the normal anatomical variability of the site

Electronic Poster: Physics track: Treatment plan optimisation: algorithms

EP-1808 Impact of the optimization-convergence errors on lung IMRT-SBRT plans computed with the Eclipse TPS J.F. Calvo Ortega 1 , H. Marcelino 2 , S. Moragues Femenia 1 , C. Laosa-Bello 1 , J. Casals 1

Results Parameter

Target LCL UCL iViewDose Tolerance

% DRP Deviation 0.67 -3.7 5.1 4%