2013 Summer Project Week:Ultrasound Needle Detection

• 

  No needle.

• 

  First Needle Inserted

• 

  Second Needle Inserted

• 

  Third Needle Inserted

Key Investigators

• BWH: Alireza Mehrtash, Daniel Kostro, Matthew Toews, William Wells, Tina Kapur
• Queen's University: Tamas Ungi, Adam Rankin

Objective

The goal of this project is to a develop a needle detection method in tracked ultrasound for application in gynecological brachytherpay. We will try to develop code inside slicer to automatically detect and segment the gynecological brachytherapy needles using online tracked ultrasound images. Our system will use an electromagnetic tracking system and also Plus ultrasound library.

Approach, Plan

• Set-up and configure the hardware/software of the tracked ultrasound system using NDI Aurora, BK ultrasound, 3D Slicer and Plus Library.
• Perform all the prerequisite calibration and registration.
• Develop code for tracked ultrasound feature extraction, needle detection and segmentation.
  • 3D reconstruction of needle.
  • Online visualization and needle navigation.
• Embed to a Slicer module.
• Test the method on our gynecologic phantom, using BK trans-abdominal probe and gynecologic brachytherapy needles.

Progress

• With the help of the Queen's team we configured our ultrasound calibration system at AMIGO.
• We determined the procedure for calibration of BK trans-abdominal probe.
• Got the appropriate code from SlicerIGT for tracked-ultrasound recording and from SurfFeatures (Daniel and Mats) in order to process ultrasound video images.
• Next:
  • Record a needle insertion procedure inside a phantom with our calibrated tracked ultrasound setup.
  • Investigation on how we can modify our current needle detection methods (iGyne Slicer module) to detect needles in tracked ultrasound.

Delivery Mechanism

• Code and slicer module
• A writeup of developments/results for future publications.