Liver Segmental Anatomy and Analysis from Vessel and Tumor Segmentation


The invention is a novel graph-based method for the automated segmentation of tumors and major intra-hepatic blood vessels and identification of the liver anatomical segments. The method allows visualization and risk analysis for interventional planning involving the liver. The method avoids the shortcomings of the traditional graph cuts or intensity-based segmentation methods by including multi-phase enhancement modeling and shape likelihoods. The segmented vessels can be correctly classified into right, middle and left hepatic, and right and left portal veins using a hybrid process that incorporates anatomical information and competitive region growing. Tumors can be detected and segmented using their differential enhancement and shape with accuracy comparable to the reports from the Medical Image Computing and Computer Assisted Intervention (MICCAI) liver tumor segmentation competition. Furthermore, a vessel tracker allowed fitting planes to the major hepatic vasculature and identifying the liver segments according to the Couinaud atlas. The automated method can be used in conjunction with manual and automatic liver segmentations to perform enhanced visualization for diagnosis and planning of interventions.

Potential Commercial Applications: Competitive Advantages:
To assist in the visualization, diagnosis and planning of interventional procedures involving the liver. 
  • The method avoids the shortcomings of the traditional segmentation methods by including multi-phase enhancement modeling and shape likelihoods.
  • Tumors are segmented with accuracy comparable to the reports from MICCAI liver tumor segmentation competition.
  • Liver segments according to the Couinaud atlas are automatically identified.
  • The automated method allows the enhanced visualization of the liver for diagnosis and planning of interventions.


Development Stage:
The algorithm and software of the method are fully developed.

Inventors:

Bradford Wood (CC)  ➽ more inventions...

Marius Linguraru (CC)  ➽ more inventions...


Intellectual Property:
Software/Research Tool — Patent protection is not being pursued for this technology.

Collaboration Opportunity:

The NIH Clinical Center, Department of Radiology and Imaging Sciences, is seeking statements of capability or interest from parties interested in collaborative research to further develop, evaluate, or commercialize techniques for the enhanced visualization, diagnosis and image-based interventions of the liver. Please contact Ken Rose, Ph.D. at 301-435-3132 or rosek@mail.nih.gov for more information. Click here to view the NIH Clinical Center collaborative opportunity announcement.


Licensing Contact:
Admin. Licensing Specialist (ALS),
Email:
Phone:

OTT Reference No: E-178-2011/0
Updated: Aug 11, 2011