Abstract Book

S208

ESTRO 37

OC-0407 Real-time dose verification of dynamic MLC tracking using a monolithic 2D silicon diode array M. Duncan 1 , M.K. Newall 1 , V. Caillet 2 , J.T. Booth 2 , M.L.F. Lerch 1 , V. Perevertaylo 3 , A.B. Rosenfeld 1 , M. Petasecca 1 1 University of Wollongong, Centre for Medical Radiation Physics, WOLLONGONG, Australia 2 Northern Sydney Cancer Centre, Radiotherapy, Sydney, Australia 3 Spa Bit, Spa Bit, Ukraine, Ukraine Purpose or Objective Patient motion management is important to avoid misalignment of the tumour and toxicities to healthy tissue during radiotherapy. One method of motion management is real time MLC tracking of the tumour. This technique has been recently applied clinically on a standard linac for both lung and prostate treatments. QA for these treatments is challenging and requires dose measurements with high spatial and temporal resolution. A dosimetry system for 2D dose verification is tested for SBRT treatments combined with MLC tracking of real patients breathing patterns. Material and Methods MP512 is a 52x52 mm 2 array of 512 individual ion implanted diodes with 2 mm centre-to-centre spacing. Each diode has an area of 0.5x0.5 mm 2 and all 512 channels can be read out in parallel by a custom designed multichannel electrometer. MP512 was placed in a solid water phantom and a CT scan was acquired. A two arc VMAT treatment plan was created in Eclipse on the CT geometry, the target was defined as a hypothetical 2cm diameter quasi-spherical volume aligned to the central pixel of the detector. A 3mm PTV expansion was created and 5Gy prescribed dose was optimised to this volume to be delivered in one fraction. Three patient lung motions were applied to the phantom using a moving platform and MLC tracking was implemented to compensate for the motion. Two types of tracking algorithms were used; predictive and non- predictive. 2D dose maps were reconstructed from MP512 and compared with EBT3 film and Treatment Planning System (TPS) dose maps through the use of a 2D gamma analysis. Dose profiles were extracted from MP512 for each motion modality and compared to the static reference case. This gave a measure of the efficacy of the tracking system and allowed identification of point- to-point dose errors. Results Fig.1 shows the comparison of the 2D dose maps extracted from EBT3 film, MP512 and TPS. Also shown are the dose profiles extracted from MP512 for each motion modality during one of the patient treatments. Gamma pass rates were above 95% for both 3%-3mm and 2%-2mm criteria when comparing the MP512 dose map to EBT3 film or TPS. When the treatment experienced motion the pass rate was as low as 55% (2%-2mm). Both tracking algorithms restored the pass rate to above 95%, indicating that the MLC tracking is effective in compensating for the motion. Motion distorted the dose profile and resulted in dose of ±150cGy compared to the reference. MLC tracking is able to reduce this error by more than half.

formalism relying on an MC-calculated chamber conversion factor ("CQ" dose).To position the EBT3 films during irradiation, they were placed in a custom PMMA holder. The dose perturbation due to the presence of the films and film holder was accounted for with an MC- calculated correction factor. Measurements were repeated with three sets of films, including a control film which was submerged but not irradiated. The netOD was calculated for each film in the region corresponding to the position of the ionization chamber. To account for the energy dependence of EBT3 film, a netOD-to-dose calibration curve was created for various beam qualities relevant to the INTRABEAM photon spectrum in water (HVL = 0.12 to 2.18 mm Al) using a Gulmay orthovoltage x-ray unit. The appropriate film calibration curve was found by interpolating to the expected INTRABEAM HVL at each depth in water. Results The EBT3 calibration curves for the lowest energy beams (HVL=0.12 and 0.816 mm Al) differed by up to 20% from one another, however there was good agreement (to 4%) between the higher energy beams (0.816 to 2.18 mm Al). This indicates that the energy dependence of EBT3 becomes significant for photon beam qualities below 0.816 mm Al (Eeff=21 keV).

In general, the EBT3 dose measurements agreed with the ionization chamber dose calculated with the Zeiss and CQ methods within uncertainties. However, the TARGIT dose was found to be significantly (30% to 60%) less than the EBT3 dose, and up to 80% less than the CQ dose. This result suggests that the TARGIT dose severely underestimates the physical dose to water.

Conclusion We have demonstrated with ionization chamber and EBT3 Gafchromic film measurements that the manufacturer reported TARGIT dose of the INTRABEAM system significantly underestimates the absorbed dose to water, with dose differences of 30% to 60% compared to EBT3, and up to 80% compared to CQ-calculated ionization chamber measurements.