ESTRO meets Asia 2024 - Abstract Book

S326

Physics – Motion management and adaptive radiotherapy

ESTRO meets Asia 2024

Figure 1. Monte Carlo simulation geometry including RANDO phantom.

The Monte Carlo model was validated against measurements performed in a simple slab phantom geometry and with an anthropomorphic RANDO phantom. Whole body 3D models of the RANDO phantom and eight volunteers were acquired using a VECTRA WB360 whole body photogrammetry system. Participants were imaged in the Stanford TSET treatment positions, with and without simulated positioning errors. In-house Python code was used to convert the exported models to synthetic CT images for dose calculation. Dose was automatically summed across beam pairs and mirrored treatment positions (e.g., left anterior oblique and right posterior oblique), but dose to specific points of interest across non-mirrored treatment positions needed to be added manually.

Results:

There was agreement within statistical uncertainty between the phantom Monte Carlo calculations and measurements, except within 25 cm of the floor, where the lack of modelling of floor scatter in the simulation geometry resulted in deviations consistent with those previously reported by Nevelsky et al. [1]. Example dose calculations can be seen in Figure 2.

Figure 2. Example dose calculations, with B-factor scale as defined by AAPM Report 23 [2].

Regions of high and low doses were seen where expected, including within fingers and ears, and on left and right sides of participants in anteroposterior poses, respectively. Variations in homogeneity were observed with body shape and posture, and the impact of positioning errors, such as incorrect rotation, could be seen. Simulated surface doses were consistent with expected skin dose B-factors defined by AAPM Report 23 [2]. Dose calculations in the participant models were consistent with previously reported TSET in-vivo dose measurements [3], except at shoulder and outer elbow (where there are high dose gradients due to beam occlusion), and the lower leg, ankle and top of foot (due to lack of floor scatter in MC simulations).

Conclusion: