SU-E-T-584: The Accuracy of the Respiratory Motion Tracking with Robotic Radiosurgery System

Abstract

Purpose: To evaluate the accuracy of the respiratory motion tracking system of the CyberKnife by Beam's eye view images from CCDcamera. Methods: The ball, 20.1 mm in diameter, in which gold marker was placed was used as a target. Using a dynamic motion phantom, it moved based on motion either of data sets from a sine curve or irregular patterns of respiration that came from two patients, both of which had 1 cm amplitude. Treatment planning was based on a static CT scan using standard parameters for CyberKnife treatment, and all of the beams in the plan aimed center of the ball. Beam's eye view images were recorded with a CCDcamera mounted on a linac head which was tracking the ball. The errors from displacement between the CCDcamera's central axis and center of the ball were measured by fifteen times per second, with self‐developed software. Results: In the case of the motion by the sine curve, errors were 2mm or less at 99% of the tracking time. In the case of the respiratory irregular motion, errors were 3mm or less in 99% of the tracking time. However, more than 5 mm errors were momentarily detected. Conclusions: We developed a method to measure statistically the error of the respiratory motion tracking. In this experiment, it was very small; however, larger errors were observed in irregular motion, comparing with regular motion with the same amplitude. The reason is the respiratory motion tracking system uses the algorithm based on prediction of the future target position using past pattern of motion. Thus, PTV margin may be set for each patient individually.