Percutaneous intervention is a common type of minimally invasive procedure used to treat
and diagnose many disorders. Typically, the needle used is rigid and straight and therefore
requires a straight-line path to the target. Problems though arise in complex scenarios,
where there is a narrow region the needle must travel through while avoiding critical tissue.
As the tissue deforms, the needle may need to be reinserted to achieve the correct path,
increasing the risk of tissue trauma. Needle steering aims to solve this, as well as enable
procedures that would otherwise not be possible. Within this context, this thesis describes
the development of less intrusive control schemes for a novel biologically inspired needle
steering system codenamed STING.
This thesis presents several research contributions to needle steering, which advance the
STING needle towards clinical use. A characterisation of the steering behaviour in three
dimensions was performed, which demonstrates the ability of the multi-segment design
to follow curvilinear paths through planar movements in full three dimensions. A planar
on-line path planning method and constrained closed-loop controller were proposed that
allow the needle to be used in dynamic, multi-target environments. Experimental validation
through one and two moving targets scenarios demonstrated the STING’s multi-targeting
capabilities, and a reduction in the placement error compared to the existing literature.
Finally, a cyclic actuation control scheme was developed, with the aim of reducing tissue
motion arising due to the insertion process. in vitro validation of the scheme within gelatine
tissue phantoms demonstrated no increase in error over standard actuation methods, but
a significant reduction in tissue deformation. A reduction of the outside diameter of the
STING needle from 12mm to 2.5mm was also achieved within this work, culminating in
a clinically viable prototype. The thesis concludes with a summary of the research and
suggestions for future work.