Optimisation of path planning for minimally invasive interventions on prostate using MR-robot: Application to on-live pets

Abstract

The robot-assisted biopsy system under Magnetic Resonance Imaging (MRI) guidance is a promising candidate for prostate cancer diagnosis in terms of improving the reliability and accuracy of reaching micro-lesions of perineum organ tissues. This has been explored for the last decade. A pre-requisite for effective interventional diagnosis is an accurate intervention planning that is difficult to achieve manually. This is due to multiple factors mainly, the fact that the insertion needles trajectories should avoid organs at risk and limited manoeuvrability, under clinician control. Thus, Biopsy-robot is a feasible solution to improve targeting accuracy. However, the planned trajectories should be consistent with the needle-driven robot's kinematics capabilities. To eliminate these problems, we present an integrated concept composed of MRI-robot and Brachytherapy (BT) path planning algorithm, allowing to minimise the number of entry points of BT needles, from a physical grid to reach all the treatable volume of the prostate gland. This means minimising the number of inserted needles from the perineum, compared to a conventional straight insertion of multiple needles. Indeed, the proposed method is able to set-up safe percutaneous routes defined by a set of oblique straight-line paths linking a minimum number of feasible entry points and desired finite target points, defined within the prostate volume. By adopting such a procedure, the aim is to contribute in minimising tissue damage and patient trauma, and risk of infection, and shortening recovery time. A combinatorial approach is proposed for the minimisation of the insertion point. The development, implementation, and experimental evaluation of the needle path planning method for the 5 Degree of Freedom (DoF) robot guide is presented and performed within a 1.5 Tesla MRI to evaluate how the proposed approach is efficient and reliable on-live pets.