An Open-Source 7-Axis, Robotic Platform to Enable Dexterous Procedures within CT Scanners

Dimitri A. Schreiber, Daniel B. Shak, Alexander M. Norbash, Michael C. Yip

TL;DR

This paper introduces an open-source, 7-axis robotic platform designed for highly dexterous CT-guided needle biopsies, particularly for lung cancer. By employing a novel serial-link design with belt and cable drives, the robot achieves minimal backlash and a compact in-bore profile, enabling a wider range of approach angles and improved precision for targeting small lesions.

What Problem Does This Paper Address?

Percutaneous needle biopsy, a crucial step in diagnosing lung cancer, is currently performed manually under CT guidance. While CT imaging excels at localizing tumors, the manual insertion of needles is challenging, especially for small lesions (under 1cm). This often leads to multiple punctures, increased radiation exposure from repeated scans, and limitations in accessing lesions that require atypical approaches to avoid vital structures like blood vessels or bone. Existing robotic solutions often struggle with the confined space within the CT scanner, limited workspace, or introduce significant imaging artifacts. This paper aims to bridge this gap by providing a dexterous, low-profile, and open-source robotic system that enhances precision and accessibility for CT-guided biopsies.

Methods

The core of the system is a 7-Degree-of-Freedom (DoF) robotic arm. It combines a 3-DoF "exo-bore" motion stage (linear-linear-roll) with a 4-DoF "in-bore" actuator. A key innovation is the use of a serial-link design with a unique combination of belt and cable drives. This approach is crucial for achieving:

  • Zero Backlash and High Transparency: Unlike traditional geared systems, belt and cable drives minimize backlash, allowing for fine, incremental movements essential for precise needle manipulation. This also leads to high transparency, meaning the operator's input is directly translated to the robot's motion.
  • Compact In-Bore Profile: By placing motors outside the CT bore and utilizing a serial-link structure, the robot achieves a remarkably small cross-section (less than 16cm²), maximizing the usable workspace within the scanner.
  • Dexterity and Workspace: The 7 DoF, including redundant axes, allows for a wide range of approach angles to target lesions throughout the chest.
  • Radiolucency: The robot's construction prioritizes radiolucent materials like carbon-fiber composites and plastics to minimize CT imaging artifacts.

The system is teleoperated, with a radiologist controlling the robot using a 3DConnexion SpaceMouse. Visual feedback is provided through CT images and an integrated endoscopic camera for a first-person view of needle insertion. The control system utilizes PID controllers on an FPGA for low-level motor control and a TCP/IP web server for high-level remote control.

Datasets

The paper does not explicitly mention training datasets. However, it utilizes a custom-built lung phantom for experimental evaluation. This phantom consists of a synthetic foam sponge within an aluminum ribcage, containing 1cm diameter clay nodules. This benchmark is designed to simulate the challenges of lung biopsy procedures.

Key Insights and Ideas

  • Serial-link design with belt/cable drives is a powerful combination for precision robotics in confined spaces: The paper demonstrates that this approach effectively overcomes the backlash and stiction issues common in other robotic systems, leading to high transparency and fine motor control.
  • Minimizing in-bore profile is paramount for CT scanner robotics: The achieved 16cm² cross-section is a significant achievement, directly translating to increased workspace and maneuverability.
  • Open-source hardware can accelerate research in specialized robotics: By making the design and implementation details publicly available, the authors aim to lower the barrier for other researchers in surgical robotics.
  • Redundancy in robotic arms can be leveraged for collision avoidance: The 7 DoF provides an extra degree of freedom that can be used for intelligent path planning to avoid obstacles.

Key Results

  • Targeting Accuracy: In a teleoperated accuracy task, the system achieved a mean positional error of 0.73mm with a standard deviation of 0.30mm when puncturing bull's-eyes.
  • Repeatability: The robot demonstrated mechanical repeatability with a standard deviation of 2.43mm and 2.95 degrees for position and orientation, respectively.
  • Workspace: Simulations confirmed an expansive, collision-free workspace allowing access to targets across the entire chest.
  • CT Artifacts: The system's construction with radiolucent materials resulted in minimal shadowing and artifacts in CT images.
  • Biopsy Task: The system successfully reached targets in a custom lung phantom on the first puncture attempt during a simulated biopsy task.

Limitations and Future Work

  • Limited translational stage travel in hardware tests: The paper notes that the hardware setup used for experiments had shorter translational stage travels than the full design due to parts availability.
  • Future Work: The authors plan to develop MRI-compatible serial-link biopsy systems and a unified user interface for robot configuration and CT image integration. They also aim to explore further applications beyond lung biopsies.

Summary updated

Abstract

This paper describes the design, manufacture, and performance of a highly dexterous, low-profile, 7 Degree-of-Freedom (DoF) robotic arm for CT-guided percutaneous needle biopsy. Direct CT guidance allows physicians to localize tumours quickly; however, needle insertion is still performed by hand. This system is mounted to a fully active motion stage superior to the patient's head and teleoperated by a radiologist. Unlike other similar robots, this robot's fully serial-link approach uses a unique combination of belt and cable drives for high-transparency and minimal-backlash, allowing for an expansive working area and numerous approach angles to targets, all while maintaining a small in-bore cross-section of less than 16cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . Simulations verified the system's expansive collision free work-space and ability to hit targets across the entire chest, as required for lung cancer biopsy. Targeting error was on average <; 1mm on a teleoperated accuracy task, illustrating the system's sufficient accuracy to perform biopsy procedures. While this system was developed for lung biopsies, it can be easily modified for other CT-guided needle based procedures. Additionally, this system is open-hardware.
BibTeX
@inproceedings{schreiber2019open,
  author = {Dimitri A. Schreiber and Daniel B. Shak and Alexander M. Norbash and Michael C. Yip},
  title = {{An Open-Source 7-Axis, Robotic Platform to Enable Dexterous Procedures within CT Scanners}},
  booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems},
  year = {2019},
  url = {https://openreview.net/forum?id=ITFSjwk0K8},
}
Figure