LAUSR

Abstract In prostate brachytherapy, radioactive seeds are implanted into the prostate for treatment of early-stage localized cancer. A major issue is seed displacement due to needle deflection, which is difficult to control as the needle is inserted manually. To address the problem and automate needle insertion, robotic systems, mathematical models for estimation and prediction and control algorithms have been developed. The method of choice for robotic steering of beveled-tip needles is predominantly intermittent axial needle rotation since this re-aligns the beveled tip and thus re-directs the needle. In this work, we present a method for needle steering to supplement axial needle rotation. A point force applied laterally to the needle near its point of insertion into tissue is used to displace the needle perpendicularly relative to its insertion axis. An advantage of this method is that the lateral force provides a continuous control input and thus continuous deflection control as opposed to axial needle rotation by 180 degrees in the 2D case of planar needle steering. Further, more control over deflection is possible as the lateral force provides direct shape change for the needle and improves the under-actuated nature of the needle. In order to predict and estimate needle deflection during insertion while applying both lateral force and axial rotation, a mechanics- and energy-based model for needle deflection is developed. Both single-layer and multi-layer tissue can be modeled if the tissue layer thickness is known. The accuracy of the model is validated experimentally. A comparison between the measured tip deflection and the model-based estimate shows only a small error. Moreover, control and sensitivity studies are carried out through insertion simulations using the model. The studies show the potentials and limitations for needle deflection reduction with various combinations of lateral force application and intermittent 180 degrees axial needle rotation.