Modeling And Path Planning For Inserting A Surgery Needle Into Soft Tissue

Needle insertion into soft tissue is an essential component of many clinical procedures and has attracted considerable attention in recent years. Many applications such as biopsies, injections, neurosurgery, and cancer treatment require the use of minimal invasive percutaneous procedures, in which precise needle tip placement to a clinical target is necessary. When inserting a needle into soft tissue, the complex tissue environment, tissue deformation, and needle deflection will result in the needle deviating from its intended path and not reaching its target position. This thesis aims to study an optimized path planner to enable steerable needles reach their target position in soft tissue environment. The research is carried out in the following perspectives.First, the investigation of needle-tissue interaction aims to predict needle deflection during insertion is presented. A prediction model that can estimate needle deflection is necessary for surgery planning before the real procedure. By using the Rayleigh-Ritz method, this thesis develops a prediction model that can successfully predict the deflection of a needle undergoing single or multiple bends during insertion. Through incorporation with the tissue model and needle geometry, the prediction model can also account for the needle deflection when the radius of the curve is not constant.Second, this study addresses the path-planning problem in inserting a bevel-tip needle in static environment (rigid tissue). Given the bevel-tip and flexibility of the needle, the trajectory of the needle tip can be approximated as a curve of constant radius. This study is the first to regulate the curved path within two parallel lines and determine the optimal distance between the two parallel lines such that the generated moving path of the needle has the shortest length with the least number of needle rotations. Both2D and3D path-planning problems have been defined and solved according to the presence of an obstacle in the trajectory path. Third, the influences of the tissue properties to the insertion of a needle into soft tissue are discussed. Given the tissue deformation and nonhomogenous tissue properties, the target position and the radius of the curve path vary with the insertion of the needle, which may result in the failure of the insertion task. Therefore, this thesis develops a dynamic path planner that can replan the path to adapt to changes in the target position and curve radius. A mass-spring model for modeling soft tissue deformation is adopted to simulate the dynamic environment. An experimental setup for steering a thin and flexible needle into phantom tissue (silica gel) with vision feedback is established to demonstrate the performance of the proposed path planner associated with tissue modeling.In summary, this thesis makes an important contribution in developing a mechanics-based prediction model and an optimal path planner for steering a flexible needle into soft tissue while considering nonhomogenous tissue properties and needle-tissue interactions.