- Define requirements and system architecture for medical robotics software system based on 3D Slicer and Robot Operating System (ROS)
|
- Needs for 3D Slicer / ROS integration in ongoing research projects (presentations by participants) (See Breakout session)
- Axel (Autonomous Surgery using the KUKA LWR)
- Simon (dVRK?)
- Junichi (OpenIGTLink and medical robotics research)
- Tobias (OCT robot / Integration of KUKA robot and 3D Slicer
- Brainstorming 1: Requirements
- Applications (e.g. endoscopic surgery, percutaneous interventions, catheterization, etc.)
- Platforms -- Linux, Windows, MAC
- Roles for 3D Slicer -- visualization, image processing, etc.
- Roles for ROS -- vision, sensors, devices, etc.
- Brainstorming 2: Architecture for 3D Slicer-ROS integration
- Types of data exchanged between ROS and 3D Slicer
- Communication scheme between ROS and 3D Slicer
- Software package to provide
- Independent middleware?
- 3D Slicer plug-in modules
- ROS modules
- Brainstorming 3: Collaborative tools / teams
- Platforms to support (ROS mainly support Linux)
- Repository
|
- Considered the following clinical scenario:
- Obtain preoperative 3D image of the patient
- Create 3D surface model of the patient from the 3D image on 3D Slicer
- Set up the patient on the OR table
- Scan the patient with a surface scanner. The point cloud data is imported to ROS (NOTE: This can be 3D Slicer, and then transferred to ROS through OpenIGTLink)
- 3D Slicer send the 3D surface model to ROS through OpenIGTLink as POLYDATA
- Perform surface matching on ROS and send the result registration transform to 3D Slicer
- Define target on the original image (or the model) on 3D Slicer
- Send the target to ROS
- Move the robot to the targetj
- Worked on the proof-of-concept system using LEGO MINDSTORMS and ROS. Due to the limited resource at project week, we used:
- 2D drawing on the paper instead of 3D model
- web cam instead of 3D surface scanner
|
