ESTRO 2025 - Abstract Book

S3225

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

ESTRO 2025

1332

Digital Poster Efficacy of non-invasive mechanical ventilation to reduce organ motion for abdominal radiotherapy Jorrit Visser 1 , Johannes K Veldman 1 , Michael J Parkes 1 , Markus F Stevens 2 , Joost G. van den Aardweg 3 , Eva Versteijne 1 , Arjan Bel 1 , Irma W.E.M. van Dijk 1 , Zdenko van Kesteren 1 1 Radiotherapy, Amsterdam UMC, University of Amsterdam, Amsterdam, Netherlands. 2 Anaesthesiology, Amsterdam UMC, University of Amsterdam, Amsterdam, Netherlands. 3 Pulmonology, Amsterdam UMC, University of Amsterdam, Amsterdam, Netherlands Purpose/Objective: Respiratory motion is a challenge in abdominal radiotherapy. Motion mitigation strategies include gating or tracking, abdominal compression and breathing control, imposed either voluntarily or mechanically. Breath-holding reduces such motion, but shows considerable residual motion (1). With non-invasive mechanical ventilation (NIMV), a regular breathing pattern (regularized breathing, RB) can be imposed at a much higher frequency compared to free breathing (FB), typically at 30 or 60 breathing cycles per minute (brpm). These higher frequencies result in smaller tidal volumes that reduce motion (2). With deep inspiration breath-hold the lungs are more inflated than with FB, offering some dosimetric benefit. A more favorable patient anatomy is also achieved by adding positive end-expiratory pressure (PEEP) to RB, which further increases lung volume. Here, respiratory motion of abdominal organs was compared between FB and two NIMV strategies. Material/Methods: Following local medical ethics committee approval, ten healthy volunteers (5F/5M; median age 49 years, range 20 67 years) practiced three breathing strategies twice, on different days: free breathing (FB), RB at a frequency of 30 brpm (RB30) and RB60 with PEEP (RB60P). Alternating single-slice coronal and sagittal MR images were acquired with a frequency of 2.6 Hz during a period of 350 s using a balanced SSFP acquisition. On a single image at mid ventilation the liver, spleen and kidneys were delineated. Other images were deformably registered to this reference image (3), resulting in a deformation vector field (DVF) for each image. For each voxel within a delineated organ the motion over time in the cranial-caudal (CC), anterior-posterior (AP, only sagittal images), and right-left (RL, only coronal images) directions was determined using the DVFs after excluding 5% outliers. Organ motion was defined as the median voxel motion within the organ. Organ motion was compared between breathing strategies using the Wilcoxon signed rank test. Results: Figure 1 shows boxplots of the organ CC motion for the different breathing strategies. For all organs, the CC and AP motion was significantly smaller during RB30 and RB60P (Table 1). Compared to FB, the organ CC motion was reduced on average by 43% and 66% for RB30 and RB60P, respectively.