Feedback Control for Steering Needles Through 3D Deformable Tissue Using Helical Paths.

Area of Science:

• Robotics
• Medical Devices
• Control Systems

Background:

• Bevel-tip steerable needles offer potential for enhanced accuracy in minimally invasive procedures.
• Current 3D needle steering lacks demonstrated robustness against tissue deformation and uncertainty.
• Advanced planning algorithms have yet to fully address these challenges.

Purpose of the Study:

• To present a feedback controller for 3D steerable needles capable of precise trajectory control.
• To demonstrate the controller's efficacy in correcting for perturbations and tissue deformation.
• To achieve high targeting accuracy despite system under-actuation and non-local controllability.

Main Methods:

• Development of a model predictive control (MPC) framework for needle steering.
• Optimization of needle twist rate to minimize predicted helical trajectory distance to target.
• Implementation of fast branch and bound techniques for real-time execution.
• Evaluation using simulated perturbations (imaging noise, needle deflection, curvature errors) and a 3D finite element simulator with deformable tissue.

Main Results:

• The controller successfully steers needles along 3D helical paths, adapting helix radius for perturbation correction.
• High accuracy was achieved for targets distant from the insertion point, even in under-actuated systems.
• In simulations with deformable tissue, the controller reduced targeting error by up to 88% compared to open-loop control.
• Real-time execution at kilohertz rates was enabled by fast branch and bound techniques.

Conclusions:

• The presented feedback controller significantly improves targeting accuracy for 3D steerable needles in the presence of tissue deformation and uncertainty.
• The MPC framework offers a robust solution for real-time needle guidance in complex medical environments.
• This technology holds promise for advancing the capabilities of minimally invasive surgical procedures.