3-D Path-Following Control for Steerable Needles With Fiber Bragg Gratings in Multi-Core Fibers

Abstract

Steerable needles have the potential for accurate needle tip placement even when the optimal path to a target tissue is curvilinear, thanks to their ability to steer, which is an essential function to avoid piercing through vital anatomical features. Autonomous path-following controllers for steerable needles have already been studied, however they remain challenging, especially because of the complexities associated to needle localization. In this context, the advent of fiber Bragg Grating (FBG)-inscribed multi-core fibers (MCFs) holds promise to overcome these difficulties. Objective: In this study, a closed-loop, 3-D path-following controller for steerable needles is presented. Methods: The control loop is closed via the feedback from FBG-inscribed MCFs embedded within the needle. The nonlinear guidance law, which is a wellknown approach for path-following control of aerial vehicles, is used as the basis for the guidance method. To handle needle-tissue interactions, we propose using Active Disturbance Rejection Control (ADRC) because of its robustness within hard-to-model environments. We investigate both linear and nonlinear ADRC, and validate the approach with a Programmable Bevel-tip Steerable Needle (PBN) in both phantom tissue and ex vivo brain, with some of the experiments involving moving targets. Results: The mean, standard deviation, and maximum absolute position errors are observed to be 1.79 mm, 1.04 mm, and 5.84 mm, respectively, for 3-D, 120 mm deep, path-following experiments. Conclusion: MCFs with FBGs are a promising technology for autonomous steerable needle navigation, as demonstrated here on PBNs. Significance: FBGs in MCFs can be used to provide effective feedback in pathfollowing controllers for steerable needles.

Review of robotic systems for thoracoabdominal puncture interventional surgery

“… 103 A three-axial force sensor was proposed to measure the pulling force, the steering force, and the lifting force in a flexible endoscopic surgical robot with FBGs embedded in the sensing structure. 104 Notably, the higher the degree of polynomial fitting, the greater the accuracy required for the position information. Otherwise, any deviation of position information could lead to significant changes in the reconstruction results.…”