ESTRO 2024 - Abstract Book

S4156

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

ESTRO 2024

For cord motion, there was no significant inter- and intrafractional motion (<0.5mm) on the spinal cord.

For patient motion, analysis from pre and post 3D MRI demonstrated that intrafractional motion was less than 1 mm and 1 degree regardless of target locations. The ranges derived from cine images remained below 2.5mm in most cases, where patient 2 exhibited more significant movement due to frequent swallowing.

Figure 1. Intrafraction motion analysis using real-time 2D cine MR images and 3D pre and post MR images

The range of Sup/Inf motion in the coronal (Cor) plane exhibited a discrepancy of up to 0.6 mm compared to that observed in the sagittal (Sag) plane, with the exception of patient 2. In this case, swallowing motion in the AP direction exaggerated and degraded tracking in the coronal plane. The outcome of motion tracking was significantly influenced by the selection of tracking structures. When the spinal cord was chosen as the tracking structure (a), tracking of sup/inf motion proved to be unreliable, primarily because of the absence of distinct boundaries in that direction. Conversely, when the single vertebrae and spinal cord(b) were selected together, the tracking process misidentified structures during image registration due to the structural similarities with nearby anatomical elements. Ultimately, the optimal choice of structure for tracking emerged as the adjacent three vertebral bodies including the spinal cord (c), which yielded tracking results similar to those obtained with a single vertebra in conjunction with the spinal cord.