ESTRO 37 Abstract book

S934

ESTRO 37

Purpose or Objective The purpose of this work was to evaluate the feasibility of using the IBA Stealth chamber as a reference detector in beam data scanning for different field sizes. Material and Methods The Stealth chamber (Stealth) from IBA is designed for use as a reference detector for scanning. The detector is attached to the collimator head and covers the entire area up to the largest field opening (40x40 cm 2 ). One benefit of the Stealth is not having to reposition the reference detector for different field size measurements as must be done when using standard reference chambers. The Stealth is designed to be “beam invisible” with good reproducible reference signal due to a larger volume. A comparison test between the CC13 and the Stealth when used as a reference detector for measuring scanning data was performed using the IBA BluePhantom2. PDDs and profiles for various field sizes from 4x4 to 40x40 cm 2 were acquired for 6MV on a TrueBeam. Scanned data measurements were performed as follows: (1) CC13 used as reference with no Stealth, (2) CC13 used as reference with Stealth in collimator head but not used as reference, and (3) Stealth used as reference. The measurements were performed utilizing these two devices sequentially for each field size, thus avoiding the introduction of setup variations into the comparison. Most measurements were repeated on a second TrueBeam to verify the results. Results Measurements of PDD data resulted in good agreement up to 20x20 cm 2 . Table 1 shows PDD data comparison for different field sizes at various depths. Deviations between the Stealth and CC13 were observed for 30x30 and 40x40 cm 2 . Changes in the PDD curve at shallow depths (i.e., shift in buildup region) resulted in differences for the PDD values at depth for larger field sizes. Profiles measurements showed good agreement up to 20x20 cm 2 with some minor differences observed in the d max region for field sizes ≥20x20 cm 2 (Fig.1). The 40x40 cm 2 profiles show an irregular shape in the shoulder region when compared to those acquired using CC13 as a reference detector. This irregularity seems to be the result of the screws on the Stealth being mounted close to the edge of the 40x40 cm 2 field along the central axis. Profiles and PDD measurements performed with the Stealth in the beam but not actively used yielded the same results as those obtained with the Stealth as a reference detector indicating that any observed changes were the result of placing the Stealth in the field.

and localizing dwell positions will reduce the time required for routine applicator quality assurance relative to film-based measurement. Material and Methods The Raven detector system (LAP Laser, Germany) is an optical and radiosensitive tool designed for EBRT quality assurance consisting of a fluorescent screen imaged using a CCD camera with variable signal integration time. For HDR brachytherapy measurements, the CCD camera was set to integrate signal for 100 ms per image (10 fps). An algorithm was implemented in Matlab (Mathworks, USA) to automatically detect and localize source dwell positions within the resultant images by analyzing changes in signal intensity between consecutive frames. To demonstrate feasibility, a plastic interstitial applicator was attached to the surface of the fluorescent screen and connected to an Ir-192 Flexitron afterloader (Elekta, Sweden) with 24.0 mGy m 2 h -1 source strength. For comparison, a piece of EBT3 radiochromic film (Gafchromic, USA) was placed between the applicator and fluorescent screen. A treatment plan was delivered consisting of four dwell positions with 1.0 cm spacing and uniform 6.8 s dwell times. Following delivery, the dwell positions were automatically identified using the Raven images and custom software. The film was digitized using a flatbed scanner and registered to the Raven coordinate system using fiducial marks using 3D Slicer. The automatically-identified Raven-based dwell positions were compared to the manually-identified film-based dwell positions in terms of 2D Euclidean distance. Results Figure 1 displays the co-registered film and four Raven images with horizontal lines for visual reference, and Table 1 summarizes the manually and automatically identified dwell positions relative to film-based dwell position 1. The mean±SD 2D distance between the Raven and film-based dwell position centers was 0.4±0.1 mm.

Conclusion We have demonstrated the feasibility of automatically identifying brachytherapy source positions using a fluorescent screen-based optical detector. Systematic differences in dwell positions between film and Raven may be due to residual film registration error or bias in the automatic localization algorithm. We will present further characterization of device performance for brachytherapy applicator quality assurance. EP-1743 Comparison of IBA Stealth with CC13 for use as reference detector in beam data scanning G. Beyer 1 , N. Hindocha 2 , R. Paiva 1 , N. Abushena 2 , R. Patel 2 , D. Mateus 1 1 Medical Physics Services Intl Ltd, Medical Physics, Cork, Ireland 2 University College London Hospitals, Radiotherapy, London, United Kingdom