טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentNeubach Zipi
SubjectUltrasound-Guided Robot for Flexible Needle Steering
DepartmentDepartment of Mechanical Engineering
Supervisor Professor Moshe Shoham
Full Thesis textFull thesis text - English Version


Abstract

Numerous clinical procedures involve needle insertion for diagnostic or therapeutic purposes. Such procedures include e.g. biopsies, regional anesthesia, blood sampling, prostate brachytherapy, deep brain stimulation, and ablation. All procedures require navigation and accurate placement of the needle tip at an organ, vessel or a lesion. The validity of a diagnosis or the success of a therapeutic treatment is highly dependent on the needle insertion accuracy.

Almost all needle insertion procedures are currently performed free-hand and the accuracy of these procedures depends on the physician expertise.? As procedures become more specific, accurate and minimally invasive, there is a need for a system that performs automatic and controlled needle steering for ensuring precise positioning of needle tip at the target.

This research describes the development of a robotic system for flexible needle steering inside soft tissue under real-time ultrasound imaging. The flexible needle is modeled as a linear beam supported by virtual springs. After the user selects a target and obstacles a trajectory is planned to avoid the obstacles and reach the target. Based on the virtual spring model, the inverse kinematics problem is solved to calculate the needle base movements required for the tip to follow the desired path. At each step of the robot the needle tip position is detected from the ultrasound image and its deviation from the planned path is taken into account in the next step to minimize the tracking error. Since the theoretical model depends on the tissue stiffness, a novel method is presented to estimate the tissue stiffness based on the tissue motion caused by the needle presence. Simulations and experiments results of needle insertion in materials with different stiffness properties showed high correlation between tissue stiffness and tissue motion. This method can also be used to estimate the target stiffness and indicate on pathological tissues.

The system performance was verified experimentally. A robot was used to maneuver the flexible needle base. The needle was inserted into a phantom with several different stiffness layers. The closed-loop experiment with the updated tissue stiffness showed tracking error of , which points on a much higher accuracy than free hand needle insertion.