Modified COAST Guidewire Robot

• Modified COAST is a teleoperated, two-tube robotic guidewire featuring an active tip steering mechanism for precise navigation in vascular interventions.
• It employs a laser-micromachined outer tube and an adjustable inner tube to achieve variable bending profiles and roll articulation, simplifying previous designs.
• Evaluated in anatomically realistic pulsatile phantom experiments, the robot safely navigated complex vessel geometries without device failure or vessel perforation.

The Modified COAST (COaxially Aligned STeerable) Guidewire Robot is a robotically steerable endovascular guidewire designed to enhance maneuverability and navigation precision in vascular interventions. Building on previous multi-tube steerable architectures, the Modified COAST robot simplifies mechanical complexity by employing a two-tube design with an active tip steering mechanism, enabling variable bending profiles, roll articulation, and operator-controlled insertion. The system has been evaluated under anatomically realistic pulsatile flow conditions using a teleoperated interface and C-arm fluoroscopic guidance, demonstrating the capability to navigate complex, tortuous vessel geometries without device failure or vessel perforation (Brumfiel et al., 15 Dec 2025).

1. Mechanical Structure and Actuation

The Modified COAST guidewire robot reduces the original three-tube-plus-tendon COAST design to a two-tube configuration with a steering tendon, while maintaining essential degrees of freedom. The mechanical composition is as follows:

• Outer tube: Nitinol, OD 0.35 mm, ID 0.30 mm. The distal ~5 mm segment is laser-micromachined with a unidirectional asymmetric notch (UAN) array that forms an active bending joint; proximal to it is a trapezoidal-notch pattern that increases stiffness and prevents kinking in curves.
• Inner tube: Nitinol, OD 0.20 mm, ID 0.10 mm, straight and unmachined, axially translatable to adjust the effective length of exposed notches, thus tuning the bending segment.
• Tendon: Nitinol wire, OD 0.076 mm, fixed at the distal tip using steel-reinforced epoxy.
• Sheath: 35 D durometer Pebax® (OD 0.66 mm, ID 0.51 mm), covers the assembly to protect from fluid ingress.

Steering is achieved by tensioning the tendon to induce bending in the UAN section, with the bending length variable via the inner tube's position. Rotational roll motion is imparted by rotating the entire outer tube. The outer tube, inner tube, and tendon are each anchored in separate modules of a 3D-printed, four-module gearbox (three modules used), each driven by DC motors via lead screws (for linear translation) or spur gears (for rotation). The full actuator assembly is mounted on a linear rail for translation into or out of the phantom vasculature (Brumfiel et al., 15 Dec 2025).

2. Kinematic and Dynamic Principles

The kinematic modeling for this class of guidewire robots relies on the segmentwise constant-curvature assumption. For each active bending segment, curvature $κ(s)$ is related to actuator displacements as

$κ(s) = \Delta l / L^2, \quad θ = κ \cdot L$

where $\Delta l$ is tendon displacement and $L$ is the actively bent length. The tip position in the base frame for a planar bend is

$$x = (1/κ)[1 - \cos(θ)], \quad y = (1/κ)\sin(θ), \quad z = 0$$

Full 3D orientation is parameterized with a roll angle $\varphi$. The Jacobian $J$ relates small actuator input changes $\Delta q$ (tendon translation, inner tube translation, tube rotation) to differential tip movement $\Delta x$ as

$\Delta x = J(q)\Delta q$

In the two-tube system, a single curvature and length parameter remain active, simplifying kinematic expressions compared to the prior three-tube configuration. No explicit re-derivation of dynamic or kinematic expressions is presented in the study; modeling and dynamic formulations for the three-tube system are detailed in previous works [(Brumfiel et al., 15 Dec 2025): Sarma et al.; Deaton et al.]. No equations for system dynamics, such as inertia, tendon friction, or damping, are specified in this evaluation.

3. Teleoperated Control

No closed-loop or autonomous control law is implemented in this version. Operation is exclusively teleoperated: the operator uses a wireless, game-style controller (Xbox-type) to individually command

• Outer-tube rotation (roll)
• Inner-tube translation (active bending length)
• Tendon translation (bending angle)
• Gross system insertion/retraction via linear stage

Feedback is limited to real-time 2D C-arm fluoroscopy, with no onboard tip force or shape sensing. No formal control diagrams or numerical specifications for controller tuning are provided. Future directions include the integration of model-based feedback (e.g., PID, model-predictive control) and embedded sensing hardware as proposed in previous studies (Brumfiel et al., 15 Dec 2025).

4. Anatomical Phantom Navigation Experiments

The system was evaluated using a compliant, multi-branch silicone phantom replicating an aortic bifurcation with internal iliac, renal, superior mesenteric, and celiac arteries (with hepatic, gastric, and splenic branches), subjected to pulsatile flow generated by a viscoelastic linear actuating pump (Vivitro Labs Inc.). C-arm imaging under a blackout cloth provided clinical-mimetic visualization. No electromagnetic tracking or position sensors were used. To mitigate buckling at the vessel entry, a human assistant applied manual support. Material properties of the phantom and numerical pulse parameters are unstated in the present report (Brumfiel et al., 15 Dec 2025).

5. Performance Outcomes and Observations

Performance evaluation is qualitative. No numerical performance metrics such as lateral navigation error $e_{err}$, minimum bending radius $R_{bend}$, or average insertion speed $v_{avg}$ are reported. Successful teleoperated navigation is demonstrated for the following paths:

• Aortic bifurcation to internal iliac artery
• Descending aorta to three left renal branches and the superior mesenteric artery with sub-branches
• Retrograde insertion into the celiac trunk, left gastric, and common hepatic branches

All traversals were conducted without device failure or vessel perforation. Quantitative measures such as completion times, navigation precision, or comparison to baseline/manual wire are not included (Brumfiel et al., 15 Dec 2025).

6. Technical Limitations and Proposed Improvements

The simplified two-tube design offers reduced complexity and cost, while retaining adjustable bending length, active tip steering, and roll in a compact envelope. However, observed limitations include:

• Suboptimal distal torque transmission: a perceptible lag exists between proximal roll input and distal wire rotation.
• Limited radiopacity of nitinol tubes/tendon reduces visualization under fluoroscopy.
• Slight reliance on manual assistance at the phantom entry point due to buckling behavior.

Future enhancements are proposed:

• Augment radiopacity (e.g., introduce tip markers)
• Optimize torque transmission (e.g., reintroduce a torque transfer element or optimize tendon path)
• Integrate closed-loop feedback and onboard force/shape sensors
• Quantitatively assess accuracy, speed, and safety in future studies (Brumfiel et al., 15 Dec 2025).

7. Clinical Impact and Research Outlook

The Modified COAST guidewire robot demonstrates effective navigation of anatomically complex vasculature under clinically relevant, fluoroscopy-only guidance in a pulsatile setting. The design's simplification maintains key functions with decreased mechanical burden. A plausible implication is that, pending further system improvements and quantitative validation, the approach could provide a foundation for teleoperated or semi-autonomous endovascular navigation platforms that reduce operator workload and fluoroscopy exposure. Ongoing and future research aims to address currently observed technical limitations, advance toward integrated sensing and closed-loop autonomy, and establish quantified performance baselines for pre-clinical and clinical translation (Brumfiel et al., 15 Dec 2025).