ESTRO 2025 - Abstract Book

S3505

Physics - Optimisation, algorithms and applications for ion beam treatment planning

ESTRO 2025

Results: PGT allowed a clear distinction in the number of detected events between the covered and uncovered parts of the treatment field for all three RS scenarios (see Fig. 2 for the 5 mm plate). The strength of the PGT signal correlated with the magnitude of the RS. The prediction of the introduced range shift based on 11-fold data provides more precise results (RMSE=0.9 mm) than based on single-fold data (RMSE=2.7 mm), while the accuracy is comparable (Fig. 2).

Conclusion: The applied PGT approach is able to detect range deviations of a few millimeters in an anthropomorphic phantom. Applying the model to 11-fold data yielded more precise results than applying it to single-fold data, signifying the importance of sufficient statistics. Although this study tested for non-clinical range shifts, the applied approach marks an important step toward RS detection in clinically realistic scenarios. While anatomical variations and tissue heterogeneities introduce greater complexity, the higher oxygen levels in the human body as compared to the phantom are expected to increase the number of detected events.

Keywords: prompt gamma-ray timing, proton range verification