ESTRO 37 Abstract book

S958

ESTRO 37

Material and Methods PDDs, profiles and output factors were measured in a PTW BEAMSCAN water tank, using a variety of chambers and were completed from a 28 cm x 28 cm field size, the maximum on the Halcyon system, down to an off axis 1 cm x 1 cm field. Output factors were also measured off axis to ensure the accuracy of the Eclipse calculations. All measurements were compared to Varian’s published data, along with the data generated by Eclipse, which was calculated on a homogenous water phantom. Results All measured PDDs agreed with Eclipse values to within 1.2%, output factors within 2%. There was a trend in Eclipse of underestimating the dose in the build-up region, with agreement improving with depth. Eclipse was also found to overestimate small field output factors compared to those measured, a trend which disappeared with increasing field size. Conclusion Point dose measurements, Varian portal dosimetry, and PTW Octavius4D were all used successfully to conduct pre-treatment patient specific QA on a variety of plans delivered using a Varian Halcyon linear accelerator. EP-1784 Real-time dosimetry for IORT procedures P. Stevens 1 , L. Verbraeken 2 , A. Plesu 2 , E. D'Agostino 2 , D. Verellen 1 , R. Weytjens 1 1 GZA- Ziekenhuizen - St. Augustinus, Radiation Physics, Wilrijk, Belgium 2 DoseVue, Radiation Physics, Turnhout, Belgium Purpose or Objective Intraoperative Radiation Therapy (IORT) is a boosting technique where the tumor site is irradiated during surgery. IORT is demonstrated to be an effective boost strategy in breast cancer with excellent local control rates and low healthy tissue toxicity. It also improves the quality of life of the patient limiting the time spent for radiation treatments and the treatment costs. The main limitation of IORT lies in its uncertainties. A large source of uncertainty is the deposited dose, with possible consequences on the outcome of the treatment. In-vivo dosimetry offers the possibility to solve the issue in dose uncertainty by measuring the dose during or immediately after the treatment. Unfortunately, the currently used dosimeters are not designed for IORT, where aspects as minimal invasiveness, capacity to cope with large instantaneous dose rates and real-time capability are fundamental. We are testing a device based on optical fibers that will serve to monitor the amount of radiation dose deposited at given positions in the patient, in real-time. Optical fiber dosimetry is in general hampered by the so-called stem effect: when an optical fiber is exposed to radiation, a light signal is produced along its length. This spurious signal overlaps with the signal generated in the sensitive part of the fiber (dosimetric sensor) and needs to be effectively suppressed, to obtain reliable dosimetric results. Material and Methods A rare-earth doped material was used to build a sensor on top of a PMMA optical fiber. The fiber has an external diameter of 1.3mm and the sensor has a length below 1mm. Preliminary clinical measurements were obtained in a water phantom, using a 6MeV Mobetron electrons accelerator, with a dose rate of 1000MU/min. The dose measuring device was optimized in a way that the stem signal originating from the fiber itself was below 2% of the dosimetric signal. Results Our optical fiber based system, tested under clinical conditions, showed a satisfactory sensitivity, as well as a robust suppression of the stem effect.

Our preliminary results show a good agreement with reference data in terms of relative depth dose profiles (PDD). Figure 1 shows the result of such a PDD measurement, performed on the Mobetron linear accelerator.

Conclusion Rare-earth based optical fiber dosimetry, thanks to their high light yield and favorable spectral properties, offer a true alternative to perform real-time optical fiber dosimetry, with effective suppression of the stem effect. These devices are characterized by minimal invasiveness, capability to work under high instantaneous dose rates and real-time performances. We are currently completing the development of a device that will allow real-time dose monitoring at multiple positions. We believe that our device will eventually increase the quality and safety of IORT treatments. EP-1785 Impact of CT artifacts of planar phantoms on VMAT pre-clinical dosimetry using the Acuros algorithm P. Colombo 1 , G. Pallazzi 1 , L. Perna 1 , S. Broggi 1 , G. Cattaneo 1 , C. Fiorino 1 , P. Mangili 1 , R. Parisi 1 , R. Calandrino 1 1 IRCCS San Raffaele Scientific Institute, Medical Physics, Milano, Italy Purpose or Objective The recent upgrade of the Treatment Planning System (TPS) in our Institute led to investigate the accuracy of the implemented calculation algorithm (Acuros 13.5). As a part of the TPS commissioning program, we performed pre-clinical verifications of VMAT plans comparing two different methods; the purpose was to evaluate the dependency of the calculation accuracy on the CT imaging artifacts of the phantom used for measurements. Material and Methods Based on the delivery reliability from the previous experience, we compared measured planar dose against calculated dose distributions using the g function without dose threshold and 3%-3mm pass/fail criteria (4%-3mm for complex case). The measured and calculated dose maps were obtained respectively with the Mapcheck2+Mapphan and the CT-scan of this system. However, the presence of different materials on the detector plane creates artifacts in the CT images, reducing the reliability of calculation. This effect was analysed in two ways. First, we manually assigned the Hunsfield Unit (HU) of the water to artifacts and air (phant-CT mod ). Second, we acquired the MVCT of the phantom (on an Helical Tomotherapy unit available in the Department), as MV-Xray beams reduce artifacts (phant- MV); dose distributions were calculated considering the previously calibrated HU curve of the MV scanner. The dosimetry impact of phant-MV against phant-CT mod was investigated by comparing the g values of 30 pre-clinical verifications of VMAT plans delivered with 6MV XRays on a Varian Clinac-2300ix: 6 breast, 12 pelvis (prostate, rectum), 5 thorax-abdominal (lung, pancreas, stomach) 7 head and neck. Statistical differences were analyzed using a t-test. Results The results in terms of g values and p-values are summarized in table 1. The differences of g-3%/3mm between the two methods (mean 2.2% range 0-9.4%) are