S2391
Physics - Quality assurance and auditing
ESTRO 2026
stable independent phase calculation in deep hyperthermia therapy based on geometrical planning.
location. The aim was to develop a simplified independent phase calculation tool based on patient CT information. Material/Methods The human body is modeled as a fat-muscle mixture core with a fat layer around. Phases of each antenna are determined by calculating electric field pathlenghts differences to the tumor center location. The phase shifts are calculated relative to the phase of the electrical field released from the top antenna using python and compared to those obtained by EasyPlan. Further an estimate of the resulting specific absorption rate (SAR) difference between Easyplan and our model was performed at the focus point location to confine that it is not exceeding 5%. The SAR difference estimation at the focus point was using the superposition of the shifted electric field amplitude difference according equation 1 assuming all 4 antennas emitting the same power. Results The independent phase calculation was performed for 87 patients. The mean patient size was 37.43 cm and 21.46 cm for width and height, respectively. The mean phase differences between the independent calculation and EasyPlan were 2.5° (left), 5.4° (bottom), and 2.9° (right).The SAR difference estimation between Easyplan and the independent phase calculation is shown in figure 1. A summed euclidic phase difference of 14° is leading to a 2% relative difference in SAR at the focus point. Taking a 5% difference in SAR for independent calculation as tolerance the euclidic phase difference shall not exceed 23°. Using this tolerance only one independent phase calculation fails from 87 performed. Investigation of this case showed a very asymmetric distribution of fat and tissue, reducing the accuracy achieved for the proposed method.
Digital Poster 4099
Introducing new surface QA-metrics for SGRT Lilian Lorgeou 1 , Jad Farah 2 , Leone Aubignac 1 , Igor Bessières 1 , Mathieu Gonod 1 1 Medical physics, Centre G.F. Leclerc, Dijon, France. 2 Sales and clinical applications, VisionRT, London, United Kingdom Purpose/Objective: AAPM[1] and ESTRO[2] provide recommendations for SGRT acceptance, commissioning, and QA, mainly focusing on system performance and motion detection. However, these protocols do not include any assessment of surface quality. This study introduces two new surface QA metrics—surface similarity and surface roughness—. The metrics were tested on two camera generations (Gen5 HD and Horizon) and on AlignRT and MapRT systems from Vision RT Ltd. Material/Methods: A Python-based tool was developed to compute two QA metrics:Surface similarity: Calculated as the percentage of points lying within 1 mm of the reference surface.Surface roughness (noise index): Defined as the standard deviation of residuals (filtered using the MAD[3] method) relative to the mean plane of each surface point.First, the tool was used to calibrate the HU thresholdfor CT external contouring, optimizing similarity with the ideal model, a 20 × 20 × 20 cm ³ cube. This cube was scanned on a Siemens go.sim and imported into Eclipse v17.The virtual cube model was compared with CT-derived contours obtained from HU thresholds between − 800 and − 200 and smoothing levels from 1 to 15.Next, AlignRT camera repeatability was evaluated from 70 acquisitions by computing the average mean distance to the reference and the surface similarity. Finally, surface roughness of MapRT cameras was analyzed using a torso phantom. Results: For the calibration step, the highest similarity (90%) was achieved with an HU threshold of − 530 and smoothing of 1.The table summarizes average results for repeatability tests through the average mean distance and the surface similarity: Horizon 1Horizon 2Horizon 3Gen 5 HDAverage mean distance (mm)0.080.070.100.04Surface similarity (%)78.762.683.484.9A marked roughness difference was observed between cameras imaging the phantom’s right side (< 0.1 mm) and left side (> 0.1 mm). This increased noise was also visible in patient surface images. VisionRT confirmed the issue and
Conclusion This study presents a novel independent method for independent calculating the phases in hyperthermia treatment. The results demonstrate that the independent phase calculation is in good agreement with the existing planning system, offering a fast and
Made with FlippingBook - Share PDF online