Force Signal Based Robot-assisted Needle Insertion

In a robot-assisted needle puncture surgery, the interactive force at the needle tip and soft tissue is very complicated and difficult to measure, thereby becoming a challenging problem to control a robot to implement a precise puncture surgery. It is thus very significant to carry out research on this topic. Herein, with an optical fiber sensors, we measured the interaction force at the tip of a needle in real time, and use the force information to identify the tissue types and to control the robot. The work done is given below.First, the insertion force at the tip of the needle is measured using an optical fiber force sensor, and the characteristics of interactive forces are analyzed from the measured force signals. Moreover, the Mallat algorithm is employed to do the wavelet transform of the force signals measured, for the purpose of tissue types and boundaries identification.Secondly, in terms of Simone model, we established the correlation of insertion force and the displacement of the needle tip, and analyze the potential factors degraded the insertion accuracy. Using the force feedback and the force pattern at the soft tissue boundaries, a robot-assisted needle insertion scheme is proposed to realize the insertion force control based on a PID control law.Finally, on a robot-assisted insertion setup, we conducted insertion experiments with the phantoms from the swine liver and belly tissues. The experimental results show that soft tissue types and boundaries can be identified effectively with the wavelet transform analysis method presented, and verified the feasibility to reduce the insertion force by a force feedback control scheme in the robot-assisted needle insertion.