Non-Linear Hysteresis Compensation of a Tendon-Sheath-Driven Robotic Manipulator Using Motor Current

Dong-Ho Lee, Young-Ho Kim, Jarrod Collins, Ankur Kapoor, Dong-Soo Kwon, Tommaso Mansi

TL;DR

This paper presents a novel method to compensate for the "slop"—specifically backlash hysteresis and dead zones—in flexible robotic catheters used in surgery. By using the robot's own motor current as a proxy for physical sensors, the researchers developed a simplified piecewise linear model that significantly improves steering accuracy without requiring expensive or bulky sensors at the catheter tip.

What Problem Does This Paper Address?

Tendon-sheath-driven manipulators (TSMs), such as those found in steerable catheters for heart surgery, are prized for being thin, flexible, and compliant. However, they are notoriously difficult to control precisely. Because the tendons rub against the sheath and have inherent slack, they exhibit non-linear behaviors:

  1. Backlash Hysteresis: A delay in movement when the motor reverses direction.
  2. Dead Zones: A range of input where the tip doesn't move at all because the tendons are just taking up slack.

While researchers have tried to model these effects before, most solutions require adding sensors (like EM trackers or vision systems) to the tip of the catheter. In a real clinical environment, these sensors add cost, increase the device's size, and complicate sterilization. This paper seeks an "open-loop" solution that achieves high precision using only the data already available in the robotic controller.

Methods

The authors propose a two-part solution: a new mathematical model and a clever way to "tune" it.

  • Piecewise Linear Model: Instead of using complex differential equations, the authors created a simplified model consisting of eight linear equations. This model maps the desired input to the actual output, accounting for the "flat" regions of dead zones and the "lag" regions of backlash.
  • Motor Current as a Sensor: The key innovation is using motor current to identify the model's parameters. They discovered that specific "signatures" in the current correspond to physical states:
  • A flat, constant current signal indicates the motor is moving through a dead zone.
  • A sharp current peak followed by a plateau indicates the end of a backlash interval.
  • Automated Calibration: By sweeping the catheter through its range of motion and analyzing the current spikes and plateaus (using Matlab functions like findchangepts), the system can automatically calibrate itself for a specific catheter before use.

Datasets

The researchers validated their approach using commercial medical hardware:

  • Hardware: Multiple Intra-cardiac Echocardiography (ICE) catheters (specifically the ACUSON AcuNav) and Transesophageal Echocardiography (TEE) probes from Siemens Healthineers.
  • Ground Truth: To verify their results, they used an Electromagnetic (EM) tracker (Northern Digital Inc.) attached to the catheter tip to measure actual position, though this sensor is not required for the final proposed control method.
  • Motion Profiles: The system was tested with both periodic (sinusoidal) and non-periodic (complex) steering motions.

Key Insights and Ideas

  • The "Free" Sensor: The most elegant idea is that the motor current is a "hidden" sensor. It reflects the tension in the tendons and the friction against the sheath, providing a window into the mechanical state of the device without adding a single wire to the catheter itself.
  • Symmetry and Shifts: They observed that while the dead zone is relatively stable, it "shifts" when you move the catheter in a second degree of freedom (e.g., bending left while already bent up). Their model accounts for this coupling.
  • Practicality over Complexity: By choosing a piecewise linear model over more "accurate" but computationally heavy models (like the Bouc-Wen model), they created a system that is easy to implement in real-time robotic controllers.

Key Results

The proposed compensation method led to dramatic improvements in accuracy across the board:

  • 1-DOF Precision: For single-axis movement, the Peak-to-Peak Error (PTPE) was reduced by 62% to 66%, and the Root Mean Squared Error (RMSE) dropped by 71% to 76%.
  • 2-DOF Precision: When moving in two axes simultaneously (a much harder task), the system still improved PTPE by roughly 40% and RMSE by 58%.
  • Qualitative Success: The "input vs. output" graphs, which normally look like a distorted loop (hysteresis), became nearly straight diagonal lines after compensation, meaning the catheter tip moved exactly where the surgeon intended.

Limitations and Future Work

  • Jerkiness: Because the model is "piecewise linear" (made of straight segments), the transitions between segments can sometimes cause a slight "jerk" in motion.
  • Mechanical Coupling: While the 2-DOF results were good, they weren't as perfect as the 1-DOF results. This suggests there are complex mechanical interactions between the four tendons that the current model doesn't fully capture.
  • Sheath Tortuosity: The experiments were mostly performed with the main body of the catheter held straight. In real surgery, the catheter might be snaked through curvy veins. The authors note that the "width" of the dead zone likely changes as the catheter bends, which is a primary target for future research.

Summary updated

Abstract

Tendon-sheath-driven manipulators (TSM) are widely used in minimally invasive surgical systems due to their long, thin shape, flexibility, and compliance making them easily steerable in narrow or tortuous environments. Many commercial TSM-based medical devices have non-linear phenomena resulting from their composition such as backlash hysteresis and dead zone, which lead to a considerable challenge for achieving precise control of the end effector pose. However, many recent works in the literature do not consider the combined effects and compensation of these phenomena, and less focus on practical ways to identify model parameters in realistic conditions. This paper proposes a simplified piecewise linear model to construct both backlash hysteresis and dead zone compensators together. Further, a practical method is introduced to identify model parameters using motor current from a robotic controller for the TSM. Our proposed methods are validated with multiple Intra-cardiac Echocardiography (ICE) catheters, which are typical commercial example of TSM, by periodic and non-periodic motions. Our results show that the errors from backlash hysteresis and dead zone are considerably reduced and therefore the accuracy of robotic control is improved when applying the presented methods.
BibTeX
@article{lee2021non,
  author = {Dong-Ho Lee and Young-Ho Kim and Jarrod Collins and Ankur Kapoor and Dong-Soo Kwon and Tommaso Mansi},
  title = {{Non-Linear Hysteresis Compensation of a Tendon-Sheath-Driven Robotic Manipulator Using Motor Current}},
  journal = {IEEE Robotics Autom. Lett.},
  year = {2021},
  url = {https://openreview.net/forum?id=zjQmkfjjihY},
}

Loading related content...

Figure