Needle steering mechanics and design cases

Abstract

Needle interventions play an important role both during the diagnosis and treatment of liver cancer. However, due to intermediate anatomical structures, such as the ribs and lungs, deep seated lesions are not always directly accessible. In addition, instrument-tissue interaction forces may cause needles to deflect during insertion. This leads to placement errors and possibly faulty diagnostic or therapeutic results. In literature, discussed methods to increase the reachability of deep seated lesions and decrease the chance on placement errors, include improvements of the medical imaging quality and of the initial needle-target alignment. In addition, the option to steer needles is actively being investigated. Needle steering involves the planning and timely modifying of instrument-tissue interaction forces in order to control the deflections in tissue. Currently investigated steering methods employ needle base manipulations, bevel-tip needles, pre-curved stylets, active cannulas, programmable bevel-tip needles, and articulated-tip needles. The technique proposed in this work employs an actively articulated needle tip. The aim of this research is to enhance our understanding of where needle-tissue interaction forces originate and how they can be effectively modified to steer needles. This is done by means of force measurements and device functionality evaluations during needle insertions in tissue simulants. The influence of tip shape on the formation of bending forces during needle insertion was studied in a fundamental and macroscopic experiment (Chapter 3). It was found that articulated bevel-tip needles are more efficient in building up bending force than matched conical-tip needles. However, increasing the tip articulation angle has a larger positive effect on bending force. Furthermore, it was found that the resultant force orientation depends on the insertion force and that the size of this vector rotation varies per tip shape. In general, the radial (bending) force component increases faster than the axial (insertion) force component. The study of these relations is relevant for the accurate estimation of tip-loads in mechanics-based needle steering models. To reach predefined targets, a teleoperation platform was developed (Chapter 4). The angle of an articulated, conical-tip needle was controlled in a closed-loop system. On-line feedback on the tip position was obtained through 3-D shape reconstructions, using fiber Bragg grating (FBG) based strain measurements. A simple PI-controller demonstrated the needle's nimble maneuverability by continuously amending the tip angle and navigation path. An advantage of articulated-tip needles is that they do not require axial rotations to change the steering plane. Optimal paths may in the future be defined with respect to the clinical task, the limitation of tissue damage, and (when applicable) the abilities of a human operator. Human operation of steerable needles is discussed by means of experimental results in manual and shared control steering tasks. In the implemented shared control setting (Chapter 5), a path planner determined a single-curved path to the target, in which the needle curvature and tissue straining conditions were minimized. The controller estimated the error between the actual and planned path and informed the human operator by means of low intensity force guidance. The ability of users to interact with the teleoperation platform and the acting kinematic needle steering constraints, was found to vary considerably. This stresses the need for studying the effective use of communication channels, e.g. by evaluating the weights users assign to the presented feedback. In the end, shared control may teach users how to cope with the acting needle steering constraints, and guide them in complicated steering tasks. Manual needle steering tasks were performed by means of a novel, tip-articulated and hand-held instrument (Chapter 6). Targets in five principal steering directions were successfully reached under visual feedback. An average targeting accuracy of 0.5 ± 1.1 mm is reported for 100 mm insertions. This shows that active manual needle steering allows for an effective compensation of the variability among insertion paths. This dissertation discusses important remaining challenges in the bridging of technical and clinical work fields and the realization of an operational steerable needle. The tip-tissue force measurements have provided insights in the ways current needle designs and mechanics-based navigation models can be improved. The tip-articulated needles show clear advantages for control systems, and allow for a manual approach in needle steering. Finally, the shared control of steerable needles was studied and may be of use to guide practitioners in case of a complex navigation task.