Axially rigid steerable needle with compliant active tip control

Abstract

Steerable instruments allow for precise access to deeply-seated targets while sparing sensitive tissues and avoiding anatomical structures. In this study we present a novel omnidirectional steerable instrument for prostate high-dose-rate (HDR) brachytherapy (BT). The instrument utilizes a needle with internal compliant mechanism, which enables distal tip steering through proximal instrument bending while retaining high axial and flexural rigidity. Finite element analysis evaluated the design and the prototype was validated in experiments involving tissue simulants and ex-vivo bovine tissue. Ultrasound (US) images were used to provide visualization and shape-reconstruction of the instrument during the insertions. In the experiments lateral tip steering up to 20 mm was found. Manually controlled active needle tip steering in inhomogeneous tissue simulants and ex-vivo tissue resulted in mean targeting errors of 1.4 mm and 2 mm in 3D position, respectively. The experiments show that steering response of the instrument is history-independent. The results indicate that the endpoint accuracy of the steerable instrument is similar to that of the conventional rigid HDR BT needle while adding the ability to steer along curved paths. Due to the design of the steerable needle sufficient axial and flexural rigidity is preserved to enable puncturing and path control within various heterogeneous tissues. The developed instrument has the potential to overcome problems currently unavoidable with conventional instruments, such as pubic arch interference in HDR BT, without major changes to the clinical workflow.

Fig 1. Schematic of the steerable needle assembly.

(A) Inner needle. (B) Outer catheter. The assembly of (A) and (B) forms the active steerable needle (A+B). The boxes show the transversal cross-section of each part.

Fig 2. Schematic of needle steering.

An applied force on the proximal end of the needle (F_proximal) results in deflection of the distal end of the needle (δ_di). The roller supports in the middle function as flexure bearings for needle deflection.

Fig 3. Needle steering with the manufactured prototype.

An applied force on the proximal end of the needle results in deflection of the distal end of the needle. The needle guide in the middle functions as flexure bearings for needle deflection.

Fig 4. Pseudo-rigid body model of the inner needle.

The inner needle is modelled with six pseudo-rigid-body links and six torsional springs. The horizontal segments and roller supports indicate the needle guiding. Downward movement of the proximal end results in axial pushing of the bottom segment and axial pulling of the upper segment resulting in upward movement of the distal tip.

Fig 5. The finite element model of the active steerable needle in normal and bent condition.

The boxes show the configuration of the inner needle during bending. The arrows indicate the direction of the shift of the segments. Stress and bending apply for the computational model with X_pr = 120 mm and θ_pr = 70°.

Fig 6. The relationship between proximal input angle and relative translation between segments for different proximal needle lengths.

The red line indicates the distal output. Markers on the distal output line relate to the vertically aligned markers on the input lines. Maximum stress and deflection is described for all proximal inputs.

Fig 7. Set-up and pipeline for shape reconstruction of the fixed-bent needle steering experiment.

The Robotic Ultrasound System (RobUSt) is used for shape reconstruction of the steerable instrument. For each insertion, RobUSt performs a volumetric scan of the tissue simulant, recording US images along with corresponding transducer pose data, expressed in the global reference frame at the base of the robotic manipulator. The silhouette of the instrument is segmented from each frame. All segmented data are assembled in a point cloud. The point cloud is processed using an iterative shape reconstruction algorithm, described in detail in Suligoj et al. [47]. The iterative shape algorithm first generates a series of points along the instrument (red dots). These points are used to fit a third order polynomial (green) describing the continuous shape of the instrument.

Fig 8. Lateral deflection of the steerable needle over insertion length in the fixed-bent needle steering experiment.

The mean error ± σ in lateral direction for two different proximal bending angles and two initial straight insertion depths. The curves characterized by steering after 40 mm are superimposed on the curves of steering from <0,0,0> to determine what the influence of the initial depth is on needle steering.

Fig 9. Set-up of active needle steering experiment in prostate tissue simulant.

Proximal needle manipulations in Z-direction allow for steering at the distal tip. Insertion of the needle in the prostate tissue simulant and movement of the US transducer is performed bimanually. The camera visualizes the back surface of the tissue simulant in transversal plane. The prostate, pubic arch, urethra and rectum are illustrated in the figure in sagittal plane to demonstrate the clinical applicability of the steerable needle for prostate brachytherapy. The pubic arch and urethra are circumvented to reach occluded prostate tissue.

Fig 10. Set-up of active needle steering experiment in ex-vivo tissue.

The trajectory of the steerable needle (red) over a length of 90 mm towards a target (black) located laterally from the neutral axis. Control of the steerable needle was performed in 3D. Segmentation of the needle and target from the US images are used for analysis.