Control and Clinical Qualification of a Robotic Laryngoscope

NOTE: Page still being constructed


robot_summary_picture.jpg Intra-airway surgery is a popular form of treatment for various conditions including throat cancer. However, the precise movements necessary in surgery are difficult for the surgeon to obtain. This project seeks to create a robotic scope manipulation system for laryngeal surgery.

  • Students: Liz Cha
  • Mentor(s): Kevin Olds, Dr. Russell Taylor
  • Other Collaborators: Dr. Jeremy Richmon

Background, Specific Aims, and Significance

Over 25,000 new cases of throat cancer are reported every year in the US, resulting in approximately 6,000 deaths per year 1). Since the throat area is fraught with complications for many cancer treatment techniques, such as radiation and chemotherapy, but relatively accessible to surgical techniques, a surgical approach is often taken. Two main surgical techniques are used: conventional surgery, and intra-airway surgery. Intra-airway surgery has many advantages over conventional surgery, including faster healing, less chance of infection, less visible scarring, and less risk of complications, but due to the confined space and poor visibility in the airway, such surgeries are often difficult and time consuming to perform. These difficulties can be partially overcome by using a flexible laryngoscope with a working port, but the problem of adequately controlling the scope with millimeter accuracy for sometimes hours at a time remains. This project seeks to solve this problem by using a robot to control the laryngoscope, thus allowing the surgeon to bypass the details of actuating the scope and focus on the higher level goals of the operation.

The main goal of this project is to create a robotic scope manipulation system for laryngeal surgery. The be clinically useful, such a system would need to:

  1. have equal or superior precision to manual scope manipulation
  2. pass all clinical regulatory requirements
  3. be rugged enough to withstand typical operating room conditions
  4. hold the scope steady even when the surgeon releases the system to attend to other tasks
  5. be easy and simple for users with little or no technical background to set up and use
  6. introduce no unnecessary risks into the procedure


Give minimum, expected, and maximum specific, measurable deliverables for the project. If more than one, give a numbered list. Something along the following lines.

  • Minimum:
    • Use of a robotized endoscope with:
      • Single hand operation for laser and scope, leaving the other hand free to use tissue manipulators
      • Precision movement
      • Laser and scope remain stationary when hands removed
      • Redundant Sensing
      • Emergency shutoffs and release
      • Small and Lightweight
      • “Operating-room-proof”
      • Built in software safety limits
      • GUI with variable speed adjustment and joint control
  • Expected:
    • Test on phantoms
    • Initial Cadaver experiments
    • Visualization of system on GUI
    • New and improved input device for surgeon
  • Maximum:
    • Extensive phantom/cadaver experiments demonstrating functionality of system
    • Clinical Approval

Technical Approach

The scope being used with the robot is a VNL-1570STK which is equipped with a 2.0 mm channel whose working port allows navigation around the “bends” in tissue. This endoscope also should allow for enough space for the surgeon to use an additional tool in his or her free hand. The scope will be attached to 3 motors which allow for the 3 degrees of freedom previously mentioned. These motors will allow for the movement of the endoscope tip, the turning of the endoscope as well as the movement of the endoscope in and out? The motors will have built in encoders to take digital position changes? as well as a set of potentiometers to measure the joints since absolute measurement is necessary. This second set of data will also allow the robot to check the measurements taken by the encoders as a safety mechanism in case of failure of the encoders. The combined digital and analog systems make overall system more robust due to their different failure points.

The motors will all be controlled using a Galil Motion Controller. The Galil Motion Controller will be enclosed in a NEMA enclosure (or box) along with a power supply. The Galil controller runs on 12 V and contains several analog and encoder inputs for each channel. The Galil controller will also use an ethernet interface to run to the Mil-spec laptop. The laptop will be drop resistant, spill resistant, high performance and long battery life. The laptop will use preexisting code from the CISST libraries in order to build an interface for the Galil Controller and the overall robot. This interface should also include compatibility with the CISST library software for the 3D space mouse.

Currently, we are using a 3D space mouse to interface with the surgeon. However, we hope to replace this device with another controller with 3 axes of motion that is more durable and sensitive to smaller movements. We plan to design and build our own. We plan to use USB protocol. This requires that a new wrapper be created to be used with the CISST libraries. If a suitable input device is found, a significant amount of work will still need to be done in order to make it compatible with the CISST libraries. We plan to follow a similar approach in code to that which exists for the Space Mouse currently.

The user interface will be created using QT. It should allow the surgeon to vary certain parameters of the robot including the range of velocities for each joint. It should also allow for mouse control of each of the joints. We hope to add a visualization debugging feature which should show us where each axis should be. It should also include a general message log and error lights which will alert the surgeon to various errors that may be occurring.

The system should also include various safety features in both hardware and software. We plan to use redundant sensing via linear potentiometers to detect possible position errors in the encoders or motion controller. Limit switches will stop any joint from being stressed to a dangerous point. In software, a “heartbeat” feature will be implement using the Galil Motion Controller and the laptop. The two systems will constantly send a “heartbeat” ping back and forth. If one system has not heard the heartbeat to send back within a certain time period, currently set at 50 μs, the entire system will cut powers to all motors.

Milestones and Progress

  1. Milestone name: Sentence or phrase explaining what it is. Include link(s) to additional pages or media as needed
    • Planned Date: xxxxxx
    • Expected Date: xxxxx
    • Status: xxxxx
  2. Milestone name: Sentence or phrase explaining what it is. Include link(s) to additional pages or media as needed
    • Planned Date: xxxxxx
    • Expected Date: xxxxx
    • Status: xxxxx
  3. etc.

Major changes or issues

  • bulletized summary of important changes to plan, unresolved dependencies, etc.


Give a short paragraph describing your results to date. This will most probably be an executive summary of the results reported at greater length in checkpoints or final reports. Note that you may update this from time to time. Also, remember that the more extensive discussions will be in the various reports.

In doing this, you may well want to create and refer to new pages with more detail and link to them with a suitable link. If you want to have a private link (not visible to outsiders), then see Prof Taylor. The basic link syntax for such a link would be description of private link. But we will need to set up some permissions and protections. Check with your mentor to see if this is needed.

Reports and presentations

NOTE: In uploading media, upload them to your own name space or to private sub-namespace, depending on whether the general public should be able to see them.

Project Bibliography

Here give numbered list of the bibliographic references for the project. Include links to pdf files wherever possible.

Other Resources and Project Files

Current code is kept in the LCSR repository under apps (Throat Robot) and can be found here.

Check out our research page here.

courses/446/2011/446-2011-10/laryngoscope_project.txt · Last modified: 2012/12/10 17:36 (external edit)