Issues In Closed-Loop Needle Steering

Percutaneous needle insertion is amongst the most prevalent clinical procedures. The effectiveness of needle base interventions heavily relies on the needle targeting accuracy. However, during insertion the needle interacts with the surrounding tissue and deflects away from its intended trajectory. To overcome this problem, a significant research effort has been made towards developing robotic systems to automatically steer bevel-tipped needles percutaneously, which turns out to be a comprehensive and challenging control problem. A flexible needle inserted in soft tissue is an under-actuated system with nonholonomic constraints. Closed loop feedback control of needle in tissue is challenging due to measurement errors, unmodelled dynamics created by tissue heterogeneity, and motion of both anatomical obstacles and targets within the tissue. In this paper, we review recent progress made in each of the complementary components that constitute a closed-loop needle steering system, including modelling needle-tissue interaction, sensing needle deflection, controlling needle trajectory, and hardware implementation.