2011 Winter Project Week:TubeTK VascularImageSegmentationAndAnalysis

• 

  Tortuosity for benign/malignant (Image from Dr. Bullitt, UNC)

• 

  Ultrasound to vessel registration

• 

  Spatial graphs of vasculature capture inter-population variations

• 

  Projects List

Key Investigators

• Kitware: Stephen Aylward, Danielle Pace
• SPL: Steve Pieper
• Luca Antiga, Daniel Haehn

Objective

TubeTK is a new open-source toolkit that hosts algorithms for applications involving images of tubes.

Two driving applications:

• Surgical guidance: registering pre-operative vascular models with intra-operative images (e.g., ultrasound)
• Characterizing vascular patters: using graph theory to distinguish clinical populations based on vascular patterns (e.g., benign -vs- malignant tumors via tortuosity)

History

• June 2001, UNC released the patent on vessel extraction method from [Aylward, Bullitt 1996...]
• TubeTK released under Apache 2.0 license: includes rights to patents

Approach, Plan

• Python module in Slicer 4 for centerline and radius estimation of vasculature in brain MRA
  • Workflow: brain envelop segmentation, seeding, extraction
• Integration with VMTK

Progress

• Skype meeting with VMTK team to learn design pattern to follow
• Extended TubeTK to include LDA methods for multi-echo MR segmentation
  • SWAN (susceptibility weighted angiography), T1, T2 data from U of Mississippi

Delivery Mechanism

This work will be delivered to the NA-MIC Kit as follows:

• All software written during the project week will be contributed to TubeTK, and algorithms will be incorporated into 3D Slicer as CLI applications.