Tendon-driven notched needle manipulation, guidance, and modeling in soft tissue under real-time ultrasound tracking

Abstract

Today, several medical diagnosis and therapeutic cancer interventions are performed using needles via percutaneous surgical procedures. The success of these procedures highly depends on the accurate placement of the needle tip at target positions. Improving targeting accuracy necessitates improvements in medical imaging and needle steering techniques. The former provides an improved vision on the target (i.e., cancerous tissue) and the needle, while the latter enables an enhanced interventional tool. In spite of considerable advancements in the medical imaging field, the structure of the needle itself has remained unchanged. In the past decade, research works have suggested passive or active navigation of the needle inside the tissue to improve targeting accuracy. In addition, to provide actuation and control for needle steering, an active needle has been introduced that’s actuated by internal tendons. This work is the culmination of studies involving the robot-assisted tracking system to (i) estimate the 3D shape of the active needle inside phantom tissue using 2D transverse ultrasound imaging, (ii) predict the 3D needle shape for real-time tracking, (iii) steer the active needle for patients with pubic arch interference, (iv) estimate tissue movement during an active needle insertion task, (v) model and control for bidirectional manipulation, (vi) perform a systematic 12-core transperineal prostate biopsy with minimal active needle insertions to avoid puncturing organs-at-risk, (vii) develop a mechanics-based model for needle-tissue interactions and (viii) autonomous control of the needle utilizing MRI-conditional parts.