ESTRO 2025 - Abstract Book

S3270

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2025

tumour matching. Pure tumour shifts were calculated by subtracting intrafractional patient motion (quantified from bone matching between pre- and post-CBCTs) from the observed tumour displacements. The translational tumour shifts from both fractions were tested for statistical significance using t-tests. Using accelerator logfiles, temporal analysis was performed by recording the time points from patient positioning on the couch to each CBCT acquisition. Vertical tumour positions from both fractions were plotted against time and fitted linearly providing a slope (Fig. 1c), resulting in a single parameter describing the average tumour motion across both fractions for each patient. These slopes were then analysed against patient-specific parameters. Pearson correlation was used to test for correlation with continuous parameters (tumour size, respiratory motion, lung volume), Kruskal-Wallis test compared four distinct tumour positions (right/left, upper/lower lobe), and Mann Whitney U tests were used between left/right and upper/lower lobe tumour positions. Results: Analysis of translational tumour shifts (Fig. 1a) revealed no significant displacement in lateral and longitudinal directions (p=0.30 and p=0.15 respectively). However, a systematic vertical tumour displacement was observed in the dorsal direction (µ=-0.12 cm, p<0.001). A patient with a large vertical tumor shift can be seen in Fig. 1b as an example. No significant correlations or differences were found between the vertical tumour motion and patient specific parameters (Table 1).