ESTRO 37 Abstract book

S940

ESTRO 37

Conclusion The PRIMO default initial beam parameters for 6 MV photon beams from Varian Clinac 2100 linacs allows obtaining dose distributions in a water phantom which agree within 3% with a database of experimental dosimetric data from a large series of linacs. The findings of this work represent a first step in the validation of the Monte Carlo software PRIMO for independent verification of radiotherapy plans computed by a commercial treatment planning system. EP-1752 15 years of independent peer review results of beam output at more than 2000 Institutions Worldwide R. Howell 1 , S. Smith 1 , J. Palmer 1 1 ut Md Anderson Cancer Center, Radiation Physics Outreach, Houston, Usa Purpose or Objective Independent peer review is an important tool for improving quality of the clinical physics program. While there have been many reports of results of such programs, they have largely focused on institutions that participate in clinical trials. The purpose of this study was to report independent peer review results from a broad spectrum of institutions, including both academic and non-academic centers. Material and Methods We analyzed the results from independent peer review of beam output for different types of radiation therapy beams, i.e. Photons (2 – 25 MV), electrons (2 – 20 MeV), and orthovoltage (1.9 mm AL – 3 mm Cu). Specifically, calculated summary statistics for the ratio between dose measured by independent peer review and dose reported by the institution. The analysis included data from over 2000 institutions in the United States and more than 150 from other countries. All beams monitored over the past 15 years (2001 – 2016) were included in the analysis. In total data for 155, 237 instances of individual beam output checks were analyzed. Figure 1: Map showing the distribution of institutions where beam output was monitored. Results The mean ratio between measured and stated doses for all beams, photon, electron, and orthovoltage beams were 0.999±0.018, 1.000±0.016, 0.999±0.019, and 0.995±0.033, respectively. While the mean values for each beam type were very close to one, > 5% of the beams monitored were more than more than ± 3% from a 1.000. Often discrepancies were found to indicative of incorrectly calibrated beams, misinterpreting the irradiation instructions, or errors in completing the irradiation form. In many instances communication with individual institutions led to identifying and correcting specific issues; specific examples will be included in the presentation. Figure 2: Results of independent peer review of external beam radiation therapy beam output for more than 155, 237 instances (Figure 1) Conclusion For a large sample of academic and non-academic institutions located throughout the world, the majority of beams monitored were found to be well within ± 3 of the stated dose. However, there were many instances where we identified serious calibration related issues that were subsequently corrected. Dissemination of such data can help prevent future similar errors from occurring.

EP-1753 Sensing ability of EPID-based in vivo dosimetry for VMAT M. Tanooka 1 , K. Tarutani 2 , H. Doi 2 , H. Suzuki 2 , Y. Takada 2 , M. Fujiwara 2 , Y. Toda 3 , H. Fujimoto 3 , M. Miyashita 3 , A. Okumura 3 , K. Kagawa 3 , N. Kamikonya 2 , K. Yamakado 2 1 Takarazuka City Hospital, Department of Radiotherapy, Takarazuka, Japan 2 Hyogo College of Medicine, Department of Radiology, Nishinomiya, Japan 3 Japan Organization of Occupational Health and Safty Kansai Rousai Hospital, Department of Radiotherapy, Nishinomiya, Japan Purpose or Objective Electronic Portal Imaging Device (EPID) is used to detect X-rays that are emitted from patient during treatment. In a commercialized system, the first image is used as a reference to be compared with the second and subsequent images which are acquired in the treatment course. Since the verification method uses integrated data, the detectability of errors decreases as the amount of the integrated data increases. The aim of this study was to investigate sensing ability of EPID-based in vivo dosimetry by dividing and evaluating treatment plan. Material and Methods We used TrueBeam (Varian Medical Systems, Palo Alto, CA) as a treatment machine and aSi-1200 as an EPID. An ExacTrac (BrainLAB AG, Feldkirchen, Germany) pelvis verification phantom was also used. One arc prostate treatment plan was used and divided into 12 sections every 30 degrees. Then, we placed the phantom on the couch and acquired images using X-rays that passed through the phantom. Furthermore, we generated four different patterns of errors in one section of them. One was generated by changing the opening of the MLC into 3 mm, 2 mm and 1 mm. The other three were generated by shifting the position of the phantom in x-, y-, and z- directions by 1 cm. The difference between the reference and error patterns was evaluated in both the integrated and divided data using gamma analyses. Results Figure 1 shows gamma analysis tolerance maps using an acceptance criterion of 0.5% dose difference and 0.5 mm DTA. Table 1 shows comparisons of gamma-index pass rates of the integrated and divided data. As for MLC error, the consistency of the integrated dosimetry data was 99.8% in pass rate of gamma analysis, it decreased to 93.2% in comparison of divided data. As for phantom shift error, the consistency of the integrated dosimetry data was 100%. It changed into 99.9%, 98.7%, and 92.0% in x-, y-, and z-directions when divided dosimetry data were compared. Errors were reduced in integrated data, and detectability of errors was improved by dividing the data. However, the x-direction error was hard to be detected. The reason was considered to be due to differences in sensing depending on the number of divisions and moving distances.