Study on physiological tremor recognition algorithm in the vascular interventional surgical robot

Abstract

With the development of surgical robotic technology, more and more requirements on safety property were raised upon the surgical robots. For master-slave control surgical robots, it's physiological hand tremor that influences the accuracy and success rate of the robot-assisted surgery. Focusing on the physiological tremor recognition and cancelling, this paper proposes a moving-window-least-square-support-vector-machine-based recognition algorithm and adaptive filter method to rectify the wrong operations caused by physiological tremor. The performance assessment was shown with the indicator of accuracy, which implies that the MWLSSVMAF reduces accuracy error of the tremor signal. The comparison between recognition results and surface electromyographic signal is conducted for assessing the classification accuracy rate. Also, some experiments on correction effects are carried out. The results indicate that the method proposed by our research possesses better classifying accuracy rate of 83% and that the secure property requests of vascular interventional surgical robot have been improved obviously.