Tremor attenuation for surgical robots using support vector machine with parameters optimization

Abstract

Reliability and accuracy of the minimally invasive surgery may be influenced by the surgeons hand tremor in the process of operation of the surgical robot. In this work, a tremor attenuation method based on machine learning is presented to track hand tremor effectively. This method is designed for filtering the interference signal with small scale sampling data and high dimension feature. First, a hybrid kernel is adopted in the proposed method to obtain a good generalization ability. Then, we introduce a optimization method to find t he optimal model parameters. Besides, experimental results demonstrated effectiveness of the proposed method in the case of limited samples.