Abstract Book

S916

ESTRO 37

advantage of such system in Cyberknife ® QA as compared with conventional radiochromic film QA procedure. Material and Methods The QA system consists of a set of PMMA block, a fluorescent screen, a CCD camera and a light-tight box. The characteristics of fluorescent screen (Kodak, Lanex regular screen) were quantified on a linac. PMMA block was placed behind the fluorescent screen to provide enough backscattering. Light signal would be reflected by a mirror below the PMMA block to the CCD camera. Dose linearity was examined and ensured in the range of interest. A median filter was applied to all captured images to remove radiation induced noise. An open field, 115mm x 100mm shaped by MLC at 800mm SAD and 15mm depth was measured and compared with the commissioning data. The pixel size of the image was 0.1mm X 0.1mm. MLC leakage test was also performed in the same settings for both X1 and X2 bank. The captured transmission signal was averaged along each leaf. Transmission % was calculated by dividing an averaged signal of the open field. A PDD measurement on 60 mm fixed cone collimator was also performed by rotating the gantry by 90°. A deconvolution kernel was applied to resolve the influence of Cherenkov emission or any other unwanted light scattering. A comparison with the commissioning data was done subsequently. Results Measured crossline and inline profile were comparable with the commissioning data which were measured by diode, as shown in Fig. 1(a)&(b). Gamma index (2%/2mm) has a 100% passing rate in both profiles. The minimum detectable dose in this system was about 2 mGy from the dose linearity measurement. In MLC leakage test, the required MU was reduced from 50000 in EBT3 film measurement recommended by Accuray to 5000 in current settings and the SNR was as large as 30. Required time was reduced from approximately 110 minutes to 35 minutes, including setup time. The captured MLC transmission signal was plotted in Fig. 2 (a)&(b) for X1 bank and X2 bank, respectively. Both MLC banks have a radiation transmission below 0.5%. Measured PDD data for fixed cone agreed well with the commissioned data with a Gamma index of (2%/2mm) 100% passing rate, as indicated in Fig.1 (c).

Electronic Poster: Physics track: Basic dosimetry and phantom and detector development

EP-1714 Accuracy And Reproducibility Of Radiochromic Film Dosimetry C. Laosa 1 , J.F. Calvo-Ortega 1 , S. Moragues 1 , J. Casals 1 1 hospital Quironsalud, Radioterapia, 08035, Spain Purpose or Objective To investigate the accuracy and reproducibility of radiochromic film dosimetry using a multichannel software. Material and Methods Three GAFChromic EBT3 films (20.3 x 25.4 cm 2 ) of a same lot were used. Twenty seven 2x2 cm 2 fragments were obtained from each film. For each film, packages of three fragments were irradiated to nine dose levels (0, 0.5, 1, 2, 3, 5, 7, 8 and 9 Gy), such as the three fragments received the same dose. Irradiations in three different days within a month were performed, one film per day. Only one film lot calibration was used. Twenty hours after exposure, the 27 fragments from each film were deployed on the bed scan (Epson V750 Pro), such as three fragments with a same dose were placed at the center and the lateral sides alongside the scanner lamp direction. Distance between the two lateral fragments was 20.3 cm. Scanning protocol: warm up 30 minutes, 5 scans, first one was discarded; 2 mm thick transparent compression glass (20.3 x 25.4 cm 2 ), film portrait orientation, transmission mode, 78-bit RGB format; and processing tools off. Film dosimetry was performed using the Radiochromic.com software (version 2.7), using its multichannel CHIP (Channel independent perturbations) algorithm, in terms of pixel values and applying the lateral correction option included in Radiochromic.com v 2.7. A “response correction” feature is included in Radiochromic.com to eliminate interscan variability and other perturbations. Three analyses were done to derive the dose of the fragments: 1) response correction was not applied (R1); 2) response correction was used using just the unexposed fragments (R2), and 3) response correction was used using the unexposed and the 9 Gy fragments (R3). Relative dose difference between the dose obtained from each fragment (“measured” dose) and the delivered dose (“true” dose) was calculated. Results Relative dose differences of 8.0% (SD= 6.1%); 2.6% (SD= 2.1%) and 1.2% (SD= 2.0%) were obtained when the response corrections R1, R2 and R3 were applied, respectively. An action level (“accuracy”) of 5% was derived for the response correction R3. Reproducibility of the measured doses in the three measurement sessions was 2.2%, 1.1 % and 0.8%, for cases R1, R2 and R3, respectively. Conclusion Accuracy of 5% and reproducibility of 0.8% were estimated for EBT3 combined with Radiochromic.com v 2.7 when the response correction R3 was used. EP-1715 Performance of fluorescent screen based quality assurance system for Cyberknife M6 T.L. Chiu 1 , B. Yang 1 , C.W. Cheung 1 , W.W. Lam 1 , K.Y. Cheung 1 , S.K. Yu 1 1 Hong Kong Sanatorium & Hospital, Medical Physics & Research Department, Happy Valley, Hong Kong SAR China Purpose or Objective To study the performance of using a fluorescent screen based system for QA testing of Accuray Cyberknife ® M6, especially on InCise2 ® MLC. To investigate the potential