ESTRO 2025 - Abstract Book

S3157

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

Brainlab Exactrac Dynamic is a system that facilitates patient positioning and monitoring during treatments using a 4D thermal camera and a stereoscopic X-ray system. This study aims to determine the appropriate threshold of Hounsfield Units (HU) and smoothing parameters to generate external contours (BODY) in Eclipse (Varian) for patients, thereby optimizing the SGRT system compared to the X-ray system. Material/Methods: The anthropomorphic phantom Cranial Verification Phantom is positioned using the Cranial 4Pi Immobilization Mask. A simulation CT study is conducted on this setup using a General Electric LightSpeed RT system. Based on these images, 26 BODY contours were generated in the Eclipse Planning System by varying the Hounsfield Units (HU) between -150HU and -960HU and adjusting the smoothing parameters between 1 and 4. Each BODY was subjected to positioning three times on a Varian TrueBeam linac, first with SGRT, then with an X-ray image both taken with Exactrac Dynamic. The displacements suggested by this second technique, indicate the accuracy of the SGRT device. The measured displacement vectors were divided by the daily calibration factor to standardize the measurements across different days. To analyze the Hounsfield Unit (HU) threshold, a cubic fit was performed, and a confidence interval was calculated to provide a range of optimal HU thresholds. An ANOVA test was conducted to assess the significance of the smoothing parameters. Results: The representation of normalized displacements against the HU threshold with a 95% confidence interval, as well as normalized displacements against the smoothing parameters, is shown in Figure 1. Minimum displacement vector obtained with the cubic fit was 0.88±0.03mm . Through cubic fitting, it was determined that the optimal HU threshold is -828±9HU; however, with a 95% confidence interval, the range of -970HU to -665HU can also be utilized. For the smoothing parameters, although differences are observed, the ANOVA test indicates that these differences are not statistically significant (Figure 2).