ESTRO 36 Abstract Book

S779 ESTRO 36 _______________________________________________________________________________________________

on the kV CBCTs were compared to corresponding dose distributions derived from the ArcCHECK measurements. Secondly, the PerFRACTION3D reconstructed dose distributions were compared to corresponding RayStation dose calculations (RaySearch) using the kV CBCTs and the original treatment plan data. The comparison was in either case carried out by the use of 3D gamma analysis as well as by comparing traditional dose volume histogram statistics. Following the testing, a clinical deployment into the clinical routine was carried out in order to perform daily dose monitoring of delivered patient dose. Results A 3D gamma analysis (2%/2mm) showed gamma passing rates >98% when comparing ArcCheck measurements/RayStation plans to the doses reconstructed in PerFRACTION3D for all plans studied. Preliminary results from comparing daily PerFRACTION3D dose reconstructions to planned dose distributions in patients treated for cancer in the head and neck region showed similar agreement. Conclusion The dosimetric accuracy of the presented method for reconstructing the delivered 3D dose distribution daily CBCT appears very promising. Further studies are however required to study the sensitivity to dosimetric changes from changes in patient anatomy and set-up errors not accounted for by the CBCT based set-up verification procedure. EP-1459 Relative Signal Ratios using an unshielded silicon detector: data from 30 centers C. Talamonti 1 , S. Russo 2 , M.D. Falco 3 , A. Bartoli 4 , E. Cagni 5 , L. Strigari 6 , M. Esposito 2 , F.R. Giglioli 7 , C. Fiandra 7 , C. Marino 8 , S. Clemente 9 , M. Stasi 10 , E. Villaggi 11 , P. Mancosu 12 1 University of Florence, Dip Scienze Biomediche Sperimantali e Cliniche, Firenze, Italy 2 Azienda Sanitaria Firenze, Fisica Sanitaria, Florence, Italy 3 Policlinico Chieti, Fisica Sanitaria, Chieti, Italy 4 Istituto Nazionale Fisica Nucleare, Sezione di Firenze, Florence, Italy 5 ArciOspedale Santa Maria Nuova, Fisica Sanitaria, Reggio Emilia, Italy 6 IFO, Fisica Sanitaria, Roma, Italy 7 Ospedale Molinette, Fiisica Sanitaria, Torino, Italy 8 Humanitas, Fisica Sanitaria, Catania, Italy 9 IRCCS CROB, Fisica Sanitaria, Potenza, Italy 10 Ospedale ordine Mauriziano, Fisica Sanitaria, Torino, Italy 11 AUSL Piacenza, Fisica Sanitaria, Piacenza, Italy 12 Humanitas, Fisica sanitaria, Milano, Italy Purpose or Objective The aim of this study was to provide Relative Signal Ratio (RSR) values of three Linac models (Elekta, Elekta Synergy BM and Varian) using measurements performed in a multicenter Italian study. An eventual mathematical description of the RSR curve was proposed in order to calculate RSR for arbitrary field sizes with high accuracy. Material and Methods Thirty centres with different LINACs joined this project. All measurements were performed using the new IBA unshielded silicon diode RAZOR and the Stealth flat ionization chamber fixed on the gantry as reference. The RAZOR was positioned at 10cm depth in water and SSD=90cm. RSRs were calculated for square field size ranging 0.6-5.0cm. Data were normalized to the 3x3cm 2 field size. In-plane and cross-plane profile were measured to correct RSR for the effective field size (EFS).The error assigned to the experimental points was estimated taking into account the statistical dispersion of the repeated measurements and the error introduced on RSR by the

detector positioning uncertainty. Collected data were clustered by linac model and each group was fitted using the function proposed by Sauer (Med Phys2007). Moreover the obtained curves were compared with the one published by Sauer and Wilbert (SWF) calculated from the fit to the mean values of four solid state detector data series of an ELEKTA linac. Results The experimental data (blue points) with the fit function (black line) and the SWF (pink dotted line) are shown in figure 1 for each linac model. In all cases, It is evident how the proposed analytical functions fit perfectly to the data for all field sizes and for all the three linac models investigated. Deviations from the SWF are shown in the figure 2. The largest difference of the RSR value respect to the SWF ones are evident below 2cm of effective field size. This is mainly due to the different geometry of the head of the linacs and therefore small changes in the energy spectrum

Conclusion In general, the adopted methodology removes much of the ambiguity in reporting and interpreting small field dosimetric quantities and facilitates a clear dosimetric comparison across a population of linacs. Moreover it minimizes the influence of measurement uncertainties and it allows accurate determination of values for non- measured field sizes. EP-1460 Detection of forced errors in VMAT plans using EPID and Epiqa dosimetric system D. Granero 1 , L. Brualla 1 , A. Vicedo 1 , M. Trinitat 1 , J. Rosello 1 1 Eresa-Hospital General Universitario, Department of Radiation Physics, Valencia, Spain