ESTRO 2025 - Abstract Book

S3299

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

ESTRO 2025

creating and characterising the performance of the first open-source couch tracking system to enable real-time adaptive radiotherapy.

Material/Methods: The performance of a custom-built couch tracking system was experimentally investigated using a robotic motion phantom and surface imaging as shown in Figure 1. The couch tracking system software uses the real-time superior-inferior measured target motion signal to achieve a counteracting couch response. The couch is a CAD designed polymethyl methacrylate base and mount with acrylic ball bearings. The couch is driven by a Firgelli linear actuator motor controlled by a Raspberry Pi. To characterise performance, a robot was mounted on the couch, and programmed to reproduce the superior-inferior motion from eight patient-measured lung and prostate tumour motion traces from an international real-time adaptive radiotherapy collaboration. 5 Motion was measured using an Intel RealSense surface imaging camera. Measurements were performed with and without couch tracking to quantify the geometric targeting benefit of couch tracking. The latency was quantified by simultaneously measuring the target and couch impulse response with surface imaging.

Results: The couch tracking system successfully reduced the geometric targeting error for both respiratory and non respiratory patient motion traces. For the lung tumour motion traces, the mean and standard deviation geometric error with and without couch tracking were 0.1±1.9 mm and 0.4±3.5 mm,, respectively. For the prostate tumour motion traces, the mean and standard deviation geometric error with and without couch tracking were 0.1±0.7 mm and 0.2±1.7 mm,, respectively. An example of the couch tracking and no couch tracking results for selected lung and prostate motion traces is shown in Figure 2. Measured latency was 300±100ms.