🏥 surgery

Haptic Portable Robotic Device for Automated Guidewire or Catheter Navigation in Endovascular Procedures

This paper presents a lightweight, portable robotic system that utilizes servo current-based haptic feedback and impedance control to autonomously navigate guidewires through endovascular lesions by classifying tissue stiffness and adapting force application, thereby enhancing safety and accessibility in resource-limited settings.

Original authors: Mohammadi, V., MacTaggart, J., Jadidi, M., Kamenskiy, A.

Published 2026-02-05

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Mohammadi, V., MacTaggart, J., Jadidi, M., Kamenskiy, A.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Idea: A "Smart" Robot for Heart Vessels

Imagine a doctor trying to thread a tiny, flexible wire (a guidewire) through a maze of blood vessels to fix a blockage. Usually, the doctor has to do this by hand, feeling for resistance with their fingers and watching X-ray screens. This is hard work, exposes them to radiation, and requires years of training.

This paper introduces a tiny, portable robot (about the size and weight of a smartphone) that can do this job automatically. Think of it as a "smart feeder" that pushes the wire through the body's plumbing, feeling for bumps and adjusting its grip just like a skilled human would, but without needing a human to hold it the whole time.

How the Robot "Feels" (The Tactile Analogy)

The robot doesn't have human fingers or expensive pressure sensors. Instead, it uses a clever trick: listening to the motor's effort.

The Analogy: Imagine you are pushing a heavy box across a floor. If the floor is smooth, you push easily. If you hit a rug or a wall, you have to push harder. Your muscles feel that extra effort.
The Robot's Version: The robot's motor is like your muscles. When the wire hits a blockage (like a plaque in an artery), the motor has to work harder to push it. The robot measures this extra electrical "effort" (current). If the motor suddenly works harder, the robot knows, "Oh, I hit something!"

The Robot's Brain: Two Ways of Thinking

The robot uses a two-step brainpower system to navigate:

1. The "Reflex" System (Finite-State Machine)
This is like a knee-jerk reaction.

The Action: The robot pushes the wire forward.
The Reaction: If the motor senses resistance (like hitting a wall), the robot immediately stops, pulls the wire back a little bit, and twists it 90 degrees.
The Goal: Just like a human doctor would, it tries to find a new angle to get around the bump. It keeps doing this until the path is clear.

2. The "Intelligence" System (Impedance Control)
This is where the robot gets smart about what it is hitting. Not all blockages are the same. Some are soft (like jelly), and some are hard (like a rock).

The Test: When the robot hits a bump, it tries to push the wire forward just 1 millimeter.
The Guess:
- If the wire moves easily but the motor doesn't work much harder, the robot guesses, "This is soft (like a clot)." It pushes gently.
- If the wire barely moves but the motor screams with effort, the robot guesses, "This is stiff (like a rock)." It pushes with more steady force.
- If it's somewhere in between, it guesses medium.
The Result: The robot adjusts its "personality" based on the bump. It becomes gentle with soft spots and firm with hard spots, preventing it from poking a hole in the blood vessel.

The Experiments: Testing the Robot

The researchers didn't test this on real people yet. Instead, they built a "training course" out of flexible plastic tubes.

The Obstacles: They put special 3D-printed inserts in the tubes to create blockages that were 20%, 40%, 60%, and even 80% blocked.
The Results:
- The robot successfully navigated all the blockages.
- The harder the blockage, the more times the robot had to pull back and try a different angle (just like a human would).
- The robot was very good at guessing the "hardness" of the blockage. It got the "stiff" ones right 95% of the time and the "soft" ones 85% of the time.

Why This Matters (According to the Paper)

The paper highlights three main advantages of this specific device:

It's Portable: It's small, light, and runs on a battery. It doesn't need to be bolted to a wall or connected to a giant X-ray machine. The authors suggest it could be used in ambulances, rural clinics, or even on space missions where big hospital equipment isn't available.
It's Safer: Because it reacts to resistance in milliseconds, it might stop before it pokes a hole in a blood vessel.
It's Independent: It can do the job without a highly skilled specialist standing over it the whole time, which could help in emergencies where experts aren't nearby.

In short: The paper describes a small, battery-powered robot that uses the "effort" of its motor to feel its way through blood vessels, automatically adjusting its push and twist to navigate blockages safely, mimicking the skills of a human doctor.

Technical Summary: Haptic Portable Robotic Device for Automated Guidewire Navigation

Problem Statement
Endovascular therapy offers significant advantages over open surgery, including minimally invasive access and faster recovery. However, current fluoroscopy-guided procedures rely heavily on the manual dexterity and experience of highly skilled operators. This dependence creates barriers in time-critical or resource-limited settings (e.g., prehospital trauma, rural clinics, or mass-casualty incidents) where trained specialists may be unavailable. Furthermore, existing robotic navigation systems (e.g., Corindus, Sensei) are typically large, tethered to hospital-grade imaging infrastructure, and lack true autonomy or tactile feedback. These limitations prevent their deployment in austere environments and leave operators vulnerable to radiation exposure and vessel injury due to the absence of real-time force sensing. There is a critical gap for a portable, intelligent system capable of autonomous guidewire navigation with haptic awareness.

Methodology
The authors developed a lightweight (400 g), portable robotic device designed to replicate the manual maneuvers of an interventionalist: linear advancement and axial rotation of a guidewire.

Mechanical Design: The device features a compact form factor (160 × 75 × 90 mm) constructed from 3D-printed resin and ABS plastic. It utilizes a two-axis Dynamixel 2XC430 servo for primary linear advancement and rotation, supported by three miniature KST X06 servos for gripping and fine "finger-like" twisting motions. A rack-and-pinion mechanism facilitates the twisting action, while silicone fingertip covers prevent slippage.
Control Architecture: The system is controlled by a Raspberry Pi 4 running Python-based software. It employs a two-layer control strategy:
1. Finite-State Machine (FSM): Mimics clinician behavior by monitoring motor current as a proxy for contact force. When current exceeds a threshold ( $F_{touch}$ ), indicating resistance, the FSM triggers a retraction and rotation routine (either 90° axial rotation or a human-like twist) to navigate around obstructions.
2. Impedance Control: To handle variable lesion stiffness, the system implements a mass-spring-damper model. Upon detecting contact, the robot performs a 1 mm probe to measure the change in motor current ( $\Delta I$ ) and displacement ( $\Delta x$ ). Based on these metrics, the obstruction is classified as Soft, Medium, or Stiff. The controller then dynamically adjusts virtual mass, damping, and stiffness gains to modulate push speed and applied force, ensuring compliant behavior in soft tissue and firmer control in calcified lesions.
Experimental Setup: Validation was conducted using a modular benchtop setup with flexible silicone tubing (15 mm internal diameter) simulating the aorta. Custom 3D-printed inserts created stenoses of 20%, 40%, 60%, and 80% severity. Experiments utilized both rigid and flexible (Formlabs Flexible Resin 80A) obstruction materials. Two detection sensitivity thresholds (75 mA and 94 mA) were tested across 160 total trials.

Key Results

Navigation Performance: The robot successfully navigated all simulated stenosis levels. As stenosis severity increased from 20% to 80%, the number of required retractions increased predictably, validating the force-based detection strategy. Higher detection sensitivity (75 mA) resulted in more frequent retractions across all stenosis levels compared to the baseline (94 mA).
Stiffness Classification: The system demonstrated high accuracy in classifying lesion stiffness. In a separate set of 60 trials using distinct material properties, the robot achieved F-scores of 0.95 for stiff, 0.83 for soft, and 0.77 for medium obstructions. Stiff lesions were identified with the highest precision due to sharp current increases and minimal displacement.
Material Interaction: Statistical analysis (three-way ANOVA) confirmed that retraction counts were significantly influenced by stenosis severity, detection sensitivity, and material stiffness. The system showed a stronger sensitivity to detection thresholds in rigid obstructions compared to flexible ones.

Significance and Claims
The paper claims that this prototype represents a significant step toward accessible, operator-independent endovascular therapy. Its primary contributions include:

Portability and Autonomy: Unlike existing hospital-bound systems, this device is battery-powered, untethered, and small enough for deployment in ambulances, field hospitals, or disaster zones.
Low-Cost Haptics: It achieves real-time tactile feedback and adaptive navigation using only internal motor current sensing, eliminating the need for expensive external force sensors or complex imaging infrastructure.
Safety and Adaptability: The impedance control strategy allows the robot to adapt its behavior to different tissue types, potentially reducing the risk of vessel perforation and hemorrhage.
Clinical Potential: The authors suggest the system could facilitate life-saving procedures like REBOA (Resuscitative Endovascular Balloon Occlusion of the Aorta) and TEVAR in prehospital settings, or serve as a pre-positioning tool for thrombectomy in acute stroke care, bridging the gap until definitive care is available.

The authors maintain a modest tone regarding limitations, acknowledging that current testing was restricted to static, single-lumen tubing models without blood flow or complex anatomical branching. They note that future work must validate the system in anatomically accurate flow models and in vivo settings before clinical translation.