Sliding-based Switching Control For Image-guided Needle Steering In Soft Tissue

This paper represents a sliding based controller for steering beveled-tip needles to stationary locations in soft tissue in a 2D environment by performing appropriate switches of bevel orientation by 180 degree axial rotations while the needle is being inserted. Assuming the rotation velocity to be large enough with respect to insertion velocity, the out of plane motions of the needle can be neglected. The proposed controller has a non-model-based structure, which is fed by the needle tip deflection error and its derivative obtained from ultrasound images and outputs the switching pattern as the control law. To analyze the stability and convergence of the error, the kinematic unicycle model for beveled-tip needle motion in soft tissue is employed and the constraints on switching parameters are derived. The performance of the controller in the sense of targeting error and number of switches is verified using an experimental setup to insert the needle into gelatin phantom tissue and ex-vivo biological tissue.