ESTRO 37 Abstract book

ESTRO 37

S463

SNC600c) with and without density inhomogeneities. Prior to each irradiation, the phantom was CT scanned encompassing the corresponding insert and detector(s) for treatment planning and dose calculation purposes. Two plans were performed in the Monaco treatment planning system (TPS); a C-shape target case and a multiple brain metastasis case. Although all detectors were irradiated using the same plan, the prescription dose was varied according to the calibration or linearity range of each dosimetric system. Dose deliveries were performed by a Versa HD linac using 6 MV FFF beams. Calculated dose distributions were exported and compared with measurements.

dynamically changing proton beam delivery times. A camera triggering technique is critical to ensure that all the proton beam deliveries are captured by the camera which are typically missed when it enters the readout phase. The triggering system design uses open-sourced hardware such as Arduino microcontroller and integrated circuits. Results We previously validated the detector’s performance for proton therapy by studying its imaging speed (11 ms), spatial resolution (0.2 mm), individual beam tracking ability in 3D, and conducting rapid beam range measurements for 94 discrete energies with submillimeter precision (0.073± 0.03 mm). The microcontroller utilizes synchrotron-generated START/STOP signals to simultaneously initiate/terminate the imaging cycles for all cameras as illustrated in figure 2. Additionally, the dose monitor pulse sequence from the synchrotron is also actively monitored to identify proton deliveries. A beam delivery to a certain location, for e.g. beam #1 in figure 2, can consist of multiple proton ‘spots’. This technique counts these spots and verifies their number for any location with the treatment planning system. This helps in implementing a location- based imaging technique where beams delivered to different locations are imaged in separate camera frames. This approach can potentially lower the ambient noise contribution to the image by matching the frame and beam delivery times, use spot-counting to identify beams that introduce errors into a treatment plan, and significantly reduce the size of imaging dataset. Conclusion The features of the existing detector and recent camera synchronization work makes our 3D detector particularly promising as a clinical tool for scanned proton beam quality assurance, machine commissioning studies, and for precise verification of patient treatment plans.

Figure: (a) The developed phantom accommodating the insert for TLD dosimetry. (b) The phantom with inserts for (top to bottom) ion chamber measurements with and without inhomogeneities, PRESAGE and film dosimetry. Results All ion chamber measurements were in close agreement (within 1.5%) with TPS predictions. TLD dosimetry is time consuming, involving increased dose uncertainties (of approximately 6% for absolute dose). Films provided 2D dose distributions in good agreement (93% passing rate with 3%/2mm criteria) with calculations but related uncertainties are dose-level dependent. Although capable to register 3D dose distributions providing the unique advantage of DVH verification, PRESAGE relative dose results suffer from optical-CT related artifacts and increased registration uncertainty. Conclusion Phantom and methods developed allow for the combination of different detector types exploiting the advantages of each dosimetric system. PO-0878 Development status of a real-time 3D scintillation detector system for proton dosimetry C. Darne 1 , F. Alsanea 1 , D. Robertson 2 , S. Beddar 1 1 The University of Texas MD Anderson Cancer Center, Radiation Physics, Houston, USA 2 Mayo Clinic Arizona, Radiation Oncology, Scottsdale, USA Purpose or Objective We have developed a three-dimensional (3D) scintillation detector using an organic liquid scintillator (LS), which can image and characterize the dose delivered by external radiation therapy sources. In this work, we describe the developmental status of this detector for use with scanning pencil proton beams. Material and Methods The detector system consist of a large volume liquid scintillator contained inside an acrylic tank (20x20x20 cm) viewed by 3 high-speed scientific complementary metal-oxide-semiconductor cameras (5.5 megapixel). The cameras capture the spatial locations and intensity variations of the scintillation light signals produced by the scanning pencil proton beams in 3D (Figure 1). The correction factors for several optical artefacts caused by propagation of the scintillation light within the LS were also determined. We have recently been working on automating synchronization of camera acquisitions to the

Figure 1. A schematic of the 3D liquid scintillator detector. It shows the 3 camera orientations with respect to the tank and the pencil beam direction.