Abstract

Objectives: Fiducial markers have been used for prostate patient treatment localization with special features such as easy to be distinguished and efficient improvement in workflow. However, there are small migrations of the implant positions, and this displacements may due to different reasons such as prostate internal structure change, external organ at risk pressure, interaction process between markers and tissues and so on. Even though the displacement fell into clinical target margin, it is valuable to understand the characteristics of these migrations and their potential clinical effects. In this study, a coordinate system transfer method was used to evaluate these fiducial marker system variations during the course of the treatment.

Methods: An arbitrary SBRT patient with 5 fractions treatment in 800cGy per fraction was selected for this study. Firstly, the initial marker positions from the simulation CT image set were abstracted with flood-filled technique, then, the positions of these fiducial markers were retrieved from the fractionated CBCT image sets based on bony registrations. These different positions of the fiducial markers in a CBCT image set formed different coordinate systems. From the different fractions of these coordinate systems, which defined by the four implanted fiducial markers, the translation and rotation of level were computed based on singular value decomposition method.

Results: For 5 CBCT image sets, the average element values of rotation matrix were 0.99, -0.02, -0.05, -0.01, 0.97, -0.11, 0.49, 0.48, and -0.08, at the order of ij, which i and j ranged from 1, to 3.
The average element values of translation vector were -0.60, 1.95, and 0.38 with the unit in centimeter. And their corresponding standard deviation were 1.18cm, 0.69cm and 0.54cm.

Conclusion(s): Through a coordinate system transfer matrix computation method, a simple clinical variation during prostate SBRT procedure was evaluated. This approach provided a comprehensive understanding of the localization variation on prostate SBRT and could be extended to other single isocenter multi-lesions-based radiotherapy. And finally, with appropriate setting, this method could be used to execute the quality assurance for deformal image registration accuracy when high precision dosimetry needs to be understood.