Fuzzy and Hybrid Prediction of Position Signal in Synchrony Respiratory Tracking System

Y. Sheng, S. Li, S. Sayeh, J. Wang, and H. Wang (USA)

Keywords

Modeling, adaptive, fuzzy, fuzzy learning, prediction

Abstract

Synchrony® Respiratory Tracking System is the Motion Tracking Subsystem of the CyberKnife® Robotic Radiosurgery System by Accuray Incorporated. It monitors the patient’s respiration and commands the manipulator to compensate for target motion while radiation is being delivered. Delay exists between the manipulator’s command and response. This delay will possibly cause unexpected or even dangerous oscillation of manipulator. Prediction is thereby needed to compensate the manipulator time lag for better tracking performance. In this paper, a fuzzy predictor and a hybrid predictor are proposed. Experimental results show that both of them generate better predictions than existing predictors, while huge performance improvement is obtained when the proposed hybrid predictor is used.

Important Links:



Go Back