Geometric characterization of an electromagnetic surface tracking system in a radiation therapy environment

Background

Despite advances in image-guided radiation therapy (IGRT), real-time, soft-tissue-based, volumetric motion monitoring remains unsolved. Integrated MRI+Linac systems are a solution, but are costly and complex. X-ray and optical photogrammetry-based systems have their limitations. Surrogate-based motion models, which use external signals to estimate internal motion, offer an alternative. We explore the feasibility of an electromagnetic (EM) fiducial-based device integrated with a surrogate-based motion model for real-time in-room volumetric motion monitoring.

Purpose

To assess the feasibility of an EM-tracking system in the linac room, with an eventual goal of integrating it into an MRI-compatible system for real-time volumetric motion monitoring.

Methods

We empirically assessed the impact of gantry rotation and the radiation beam on EM-tracking accuracy using a sinusoidal motion trajectory (2 cm peak-to-peak, 5 s per cycle) programmed into a 2D motion platform. Four EM-tracking sensors were affixed to the platform, and their recorded trajectories were compared to the programmed motion under various conditions, including static and dynamic gantry positions, with and without radiation beams, and during CBCT acquisition.

Results

The EM-tracking system faithfully reproduced the programmed sinusoidal motion during treatment beam (MV) and CBCT acquisition (kV + gantry rotation). With the beam off and static gantry and static motion platform at 0°, the average point-wise tracking difference was < 0.5 mm compared to gantry angles of 90°, 180°, and 270°. Similarly, with a moving platform, the sensors achieved a < 1 mm difference at the same angles. Additionally, the gantry's clockwise and anticlockwise rotations caused a < 0.5 mm difference on average at all angles during beam-off.

Conclusion

Preliminary results show the EM-tracking system operates with sub-millimeter accuracy in the linac room, with minimal effects from the radiation beam, gantry motion, or CBCT acquisition, supporting its feasibility for real-time volumetric motion monitoring during IGRT.