Using ECG in Motion Prediction for Radiosurgery of the Beating Heart

A new, emerging application for robotic radiosurgery is treating atrial fibrillation by irradiating the beating heart [1, 2]. For this method, the CyberKnife’s motion compensation system [3] is used to deliver highly focused radiation to the beating heart to create ablation lines around the pulmonary veins. To improve the system’s targeting accuracy, motion prediction is needed to compensate for inevitable latencies. The current generation of the CyberKnife sports a latency of approximately 120 ms. It has been shown that live tracking of the ablation site using 3D ultrasound is indeed possible [4, 5]. We have recently studied the applicability of motion prediction algorithms used in tracking respiratory motion [6]. A new idea to improve prediction quality is to make use of an additional surrogate signal: the ECG. Using a custom-built synchronisation board, a biosignal amplifier and a high-speed IR tracking camera, we have synchronously recorded the ECG and cardiac apex beat motion trace of a healthy male volunteer.