Room: Exhibit Hall
Purpose: Facilitate the comparison of measured and TPS derived beam profiles through the determination of detector response functions (DRFs) for multiple detectors on TomoTherapy units.
Methods: Transverse and longitudinal beam profiles were measured on a Radixact unit using five ion chambers (A12, CC13, A1SL, A26, and A16) and a diode detector. Individual DRFs were determined for each detector, jaw width (1, 1.8, 2.5 and 5 cm), and depth (15, 50, 100, 150, and 200 mm). The Gaussian shape parameter (Ïƒ) was varied and gamma was used as the optimization criteria to determine the DRF. The DRFs derived from each profile were evaluated for consistency with depth and jaw width for each detector. Gamma analysis was performed comparing measured profiles and TPS profiles convolved with the DRF derived from the 1 cm jaw width and 1.5 cm depth for each detector.
Results: Across all depths and jaw widths, the mean Â± std. dev of was 2.51 Â± 0.14, 2.25 Â± 0.16, 1.42 Â± 0.20, 1.16 Â± 0.18, 0.88 Â± 0.20, and 0.09 Â± 0.09 mm for the A12, CC13, A1SL, A26, A16, and diode respectively. Using a single DRF for all profiles resulted in passing gamma analysis for all profiles and depths. The Gaussian shape parameter was found to increase linearly with depth and jaw width for ion chamber profiles (RÂ² > 0.87). An empirical DRF was derived as a function of jaw width and profile depth, and has shown improved gamma results relative to the single kernel analysis.
Conclusion: TomoTherapy TPS beam profiles can be compared with measured profiles of ion chambers with varying dimensions through convolutions with detector-specific kernels. A single kernel was found suitable to achieve a passing gamma analysis, and empirical DRFs possess potential to improve performance.
Tomotherapy, Quality Assurance, Convolution