ESTRO 38 Abstract book

S37 ESTRO 38

shown that the posterior prostate often drops between treatment planning and treatment.

Conclusion Novel IVD systems can be used to provide feedback on a reconstructed BT dose distribution. The observed needle shifts confirm a precise reconstruction of the needles on MR images and only limited discrepancies in the position of treatment needles (1SD < 2mm). The dosimetric impact of the needle movements relative to the dosimeter was small except in one case where the source entered the bladder wall. Overall needle migration has previously been observed during prostate HDR-BT [1] and has a larger dosimetric impact than the needle shifts relative to the dosimeter when the bladder wall case is excluded. Needle migration cannot be observed with the ST method presented here as the dosimeter needle will typically migrate together with the entire needle implant. [1] Buus S et al. Needle migration and dosimetric impact in high-dose-rate brachytherapy for prostate cancer evaluated by repeated MRI, Brachytherapy 17(2018). OC-0078 Error detection thresholds for routine real time in vivo dosimetry in HDR prostate brachytherapy J. Mason 1 , A. Henry 2 , P. Bownes 1 1 Leeds Cancer Centre, Medical Physics and Engineering, Leeds, United Kingdom; 2 Leeds Cancer Centre, Clinical Oncology, Leeds, United Kingdom Purpose or Objective Real time in vivo dosimetry (IVD) for HDR prostate brachytherapy may provide independent verification of correct dose delivery and is part of our routine clinical workflow since 2014. This study evaluates the use of error detection thresholds in IVD. Material and Methods IVD is implemented using a microMOSFET placed near the centre of the prostate in an additional needle. Error threshold calculation methodology was determined from the first 40 cases. IVD measurements were then performed for a further 72 patients treated with 15Gy or 19Gy single fraction treatments, applying the error thresholds to per- needle and per-plan measurements. Needle insertion and treatment planning use real-time trans-rectal ultrasound (TRUS). Thresholds are determined individually for each patient using an uncertainty analysis; source-MOSFET distance uncertainty is the dominant component. A measurement is outside the error detection threshold if it is outside the (k=2) uncertainty range. Additionally, per-needle measurements are only flagged as outside the threshold if absolute difference between measured and predicted reading is >20mV (~0.2Gy). Plans/needles outside the thresholds were reviewed. Results Table one summarises differences between measured and predicted reading per-plan and per-needle, and the uncertainty based thresholds. Figure 1 shows a measurement for one patient with two needles outside the threshold (highlighted). Of 3 plans outside the threshold, 1 was attributed to uncertainty in identifying the MOSFET position. 2 were due to uncertainty calculation limitation in the case where one needle is very heavily weighted so that the overall plan measurement fails even though each per-needle measurement passes. These false errors can be addressed by improvements in measurement technique and developing the method for aggregating per-needle uncertainties at the plan level. Of 16 needles outside the error threshold, 2 were due to uncertainty in the reconstructed needle position of <2mm, 4 were due to uncertainty in identifying the MOSFET position. One needle reconstruction error when corrected caused the urethral D10 to increase by 3.5% which would have exceeded the dose constraint. For the remaining 10 needles no obvious error could be determined, however 8 of these were located posteriorly and previous work has

Conclusion IVD in HDR prostate brachytherapy using a microMOSFET provides a high level of confidence that we are correctly delivering the planned dose to our patients, also allowing detailed analysis of dosimetric effects due to needle movement and reconstruction errors. In common with other IVD methods, it is difficult to completely avoid declaring false errors when defining error detection thresholds for real-time per needle IVD. The results of this analysis can be used to inform decisions on when to interrupt treatment if errors are detected during real- time IVD. OC-0079 Expanding Calibration Service for LDR Brachytherapy Seeds by Photon Fluence Determination T. Schneider 1 1 Physikalisch-Technische Bundesanstalt, Department 6.3 "Radiation Protection Dosimetry, Brunswick, Germany Purpose or Objective The emitted radiation field and the dosimetric properties of LDR brachytherapy photon sources depends on the nuclide and the design of the seeds. BRAPHYQS and GEC- ESTRO encourage promotion of an efficient solution in Europe to monitor and assure seed design constancy, preferably like the CLA2004 - methodology controlled by AAPM. In Europe, Aspects of the Quality Assurance (QA) in the production of seeds, including the constancy of the sources are not covered by the calibration of the seeds but by the process of CE-certification. Thereby the company’s QA system must be outlined, and the effectiveness and permanent improvement must be demonstrated. Low energy photon spectrometry is a very sensitive tool to monitor the constancy of the radiation field of the seeds. However, establishing and maintaining a spectrometry unit requires considerable effort and such services are scarce within the brachytherapy community. The national metrology institute of Germany has decided to expand its calibration service for LDR seeds to include determination of photon fluence spectra at specific polar and azimuthal angles around the source from 2020 on. The inclusion of spectrometry into regular calibration service requires highly automated spectra measurement and data evaluation. Material and Methods A commercially available detector system from Canberra (HPGe) is used controlled by the Genie 2000™ software